Can I have a like and comment _老虎社区_美港股上老虎

7859c582
2021-03-16

Can I have a like and comment pls

How To Spot A Bubble<blockquote>如何发现气泡</blockquote>

Summary The defining feature of a bubble is inconsistency between expected returns based on price b

How To Spot A Bubble<blockquote>如何发现气泡</blockquote>

免责声明：上述内容仅代表发帖人个人观点，不构成本平台的任何投资建议。

推荐
最新

暂无评论

热议股票

{"i18n":{"language":"zh_CN"},"data":{"magic":2,"id":325193400,"tweetId":"325193400","gmtCreate":1615871774900,"gmtModify":1703494275779,"author":{"id":3574340444890290,"idStr":"3574340444890290","authorId":3574340444890290,"authorIdStr":"3574340444890290","name":"7859c582","avatar":"https://community-static.tradeup.com/news/default-avatar.jpg","vip":1,"userType":1,"introduction":"","boolIsFan":false,"boolIsHead":false,"crmLevel":2,"crmLevelSwitch":0,"individualDisplayBadges":[],"fanSize":3,"starInvestorFlag":false},"themes":[],"images":[],"coverImages":[],"extraTitle":"","html":"<html><head></head><body>Can I have a like and comment pls</body></html>","htmlText":"<html><head></head><body>Can I have a like and comment pls</body></html>","text":"Can I have a like and comment pls","highlighted":1,"essential":1,"paper":1,"likeSize":2,"commentSize":2,"repostSize":0,"favoriteSize":0,"link":"https://laohu8.com/post/325193400","repostId":1172271196,"repostType":4,"repost":{"id":"1172271196","kind":"news","pubTimestamp":1615871336,"share":"https://www.laohu8.com/m/news/1172271196?lang=zh_CN&edition=full","pubTime":"2021-03-16 13:08","market":"us","language":"en","title":"How To Spot A Bubble<blockquote>如何发现气泡</blockquote>","url":"https://stock-news.laohu8.com/highlight/detail?id=1172271196","media":"Hussman Funds","summary":"Summary\n\nThe defining feature of a bubble is inconsistency between expected returns based on price b","content":"Summary<blockquote>总结</blockquote> <ul> <li>The defining feature of a bubble is inconsistency between expected returns based on price behavior and expected returns based on valuations.</li> <li>If we compare our most reliable valuation measures with the valuation measures that one would obtain from a proper long-term discounted cash flow analysis, we find that they're nearly identical.</li> <li>One of the unfortunate bits of financial illiteracy that Wall Street has pushed into the heads of investors is the idea that extreme valuations are \"justified\" by low interest rates.</li> <li>It's undoubtedly true that profit margins, expected growth, and other factors have an effect on future deliverable cash flows and the valuations that investors place on stocks.</li> <li>To understand why extreme valuations imply high volatility, and require extremely long investment horizons, it's useful to consider the concept of \"duration.\"</li> </ul> I can show, really precisely, that there are two warranted prices for a share. The one I prefer is based on such fundamentals as earnings and growth rates, but the bubble is rational in a certain sense. The expectation of growth produces the growth, which confirms the expectation; people will buy it because it went up. But once you are convinced that it is not growing anymore, nobody wants to hold a stock because it is overvalued. Everybody wants to get out and it collapses, beyond the fundamentals.- Nobel Laureate Franco Modigliani, New York Times, March 30, 2000 The word \"bubble\" is tossed around quite a bit in the financial markets, but it's rarely used correctly. See, the thing that defines a bubble isn't that valuations are extremely high, or that expected returns are extremely low. Instead, what defines a bubble is that investors drive valuations higher without simultaneously adjusting expectations for returns lower. That is, investors extrapolate past returns based on price behavior, even though those expectations are inconsistent with the returns that would equate price with discounted cash flows.<blockquote><ul><li>泡沫的定义特征是基于价格行为的预期回报和基于估值的预期回报之间的不一致。</li><li>如果我们将最可靠的估值指标与从适当的长期贴现现金流分析中获得的估值指标进行比较，我们会发现它们几乎相同。</li><li>华尔街向投资者灌输的一个不幸的金融文盲是，低利率“证明”极端估值是合理的。</li><li>毫无疑问，利润率、预期增长和其他因素会影响未来可交付的现金流和投资者对股票的估值。</li><li>为了理解为什么极端估值意味着高波动性，并且需要极长的投资期限，考虑“久期”的概念是有用的。</li></ul>我可以非常精确地证明，一股有两个保证价格。我更喜欢的是基于盈利和增长率等基本面，但泡沫在某种意义上是理性的。对增长的预期产生了增长，增长证实了预期；人们会买它，因为它上涨了。但一旦你确信它不再增长，就没有人愿意持有一只股票，因为它被高估了。每个人都想退出，但它崩溃了，超出了基本面。——诺贝尔奖获得者佛朗哥·莫迪里阿尼，《纽约时报》，2000年3月30日“泡沫”一词在金融市场上经常被提及，但很少被正确使用。看，定义泡沫的不是估值极高，也不是预期回报极低。相反，泡沫的定义是投资者推高估值，而不同时降低回报预期。也就是说，投资者根据价格行为推断过去的回报，即使这些预期与将价格等同于贴现现金流的回报不一致。</blockquote> In March 2000, at the height of the technology bubble, I noted: \"Over time, price/revenue ratios come back in line. Currently, that would imply an 83% plunge in tech stocks. If you understand values and market history, you know we're not joking.\" The following month, I discussed Modigliani's quote above, and detailed the dynamics he was describing. The collapse of the 2000 bubble would ultimately erase half the value of the S&amp;P 500, and would take the tech-heavy Nasdaq 100 down an implausibly precise 83%.<blockquote>2000年3月，在科技泡沫最严重的时候，我指出：“随着时间的推移，市盈率会恢复正常。目前，这意味着科技股将暴跌83%。如果你了解价值观和市场历史，你就会知道我们不是在开玩笑。”接下来的一个月，我讨论了莫迪里阿尼在上面的引用，并详细描述了他所描述的动态。2000年泡沫的破裂最终将抹去标普500一半的价值，并使以科技股为主的纳斯达克100指数下跌83%，令人难以置信。</blockquote> The defining feature of a bubble is inconsistency between expected returns based on price behavior and expected returns based on valuations. If investors pay $150 today for a security that will deliver a single $100 payment a decade from now, but they also fully understand that they'll lose 4% annually on the deal, without extrapolating past gains into the future, then we might say the security is overvalued, and we might question why investors would accept that trade, but we can't call it a bubble.<blockquote>泡沫的定义特征是基于价格行为的预期回报和基于估值的预期回报之间的不一致。如果投资者今天支付150美元购买一种证券，而十年后只需支付100美元，但他们也完全明白，在不将过去的收益外推到未来的情况下，他们每年将损失4%，那么我们可能会说证券被高估了，我们可能会质疑为什么投资者会接受这种交易，但我们不能看涨期权它是泡沫。</blockquote> But if investors pay $150 today for that security, because they look back in the rear-view mirror, decide that it \"always goes up\" over time, and convince themselves that expected future returns are always positive, then you've got a bubble. Discounting the future $100 cash flow of the security using any positive expected return would produce a price less than $100. So the positive returns expected by investors are inconsistent with the returns that would equate price with discounted cash flows. The size of the bubble is the fraction of the market price that represents expectational \"hot air.\"<blockquote>但如果投资者今天为该证券支付150美元，因为他们回头看后视镜，认为它随着时间的推移“总是上涨”，并说服自己预期的未来回报总是正的，那么你就有了泡沫。使用任何正预期回报贴现证券未来100美元的现金流将产生低于100美元的价格。因此，投资者预期的正回报与将价格等同于贴现现金流的回报不一致。泡沫的大小是代表预期“热空气”的市场价格的分数。</blockquote> Likewise, the willingness of investors to embrace \"passive investments\" like ETFs and asset-backed securities based on past performance, with little concern about the valuations, yields, or credit risk of the securities inside, is the very soap from which bubbles repeatedly emerge. Amid the current enthusiasm for special purpose acquisition companies (SPACS), investors might recall the bubble in \"incubators\" at the 2000 peak, the \"conglomerates\" of the late-1960s Go-Go bubble, and even the South Sea Company in the early 1700s, along with similar companies formed at the time \"for carrying on an undertaking of great advantage, but nobody to know what it is.\"<blockquote>同样，投资者愿意接受基于过去表现的ETF和资产支持证券等“被动投资”，而很少关心其中证券的估值、收益率或信用风险，这正是泡沫反复出现的肥皂。在当前对特殊目的收购公司(SPAC)的热情中，投资者可能会回忆起2000年高峰期“孵化器”的泡沫、20世纪60年代末Go-Go泡沫的“企业集团”，甚至是20世纪初的南海公司。1700年代，以及当时成立的类似公司“是为了开展一项具有巨大优势的事业，但没有人知道它是什么”。</blockquote> If investors price the S&amp;P 500 at levels that are highly likely to produce negative returns for a decade, as they did in 1929 and 2000, and as I believe they are doing at present, yet investors continue to press stock prices higher on the expectation that they will provide historically normal levels of future return regardless of valuations, then you have the sort of inconsistency that defines a bubble.<blockquote>如果投资者将标普500定价在十年内极有可能产生负回报的水平，就像他们在1929年和2000年所做的那样，而且我相信他们目前正在做的那样，但投资者继续推高股价，因为他们期望无论估值如何，他们都将提供历史上正常的未来回报水平，那么就存在定义泡沫的不一致性。</blockquote> Likewise, if the expected return of a conventional passive investment mix is negative on a 10-12 year horizon (based on reliable valuation measures strongly correlated with actual subsequent returns over a century of market history), yet pension return assumptions remain locked near 7% annually, you've got a bubble, and most likely a future pension funding crisis, on your hands.<blockquote>同样，如果传统被动投资组合在10-12年内的预期回报为负（基于与一个世纪市场历史中实际后续回报密切相关的可靠估值指标），但养老金回报假设仍锁定在7%附近每年，你就面临着泡沫，很可能是未来的养老金融资危机。</blockquote> This is how very bad things have repeatedly happened in the financial markets across history, enabled by what Galbraith called \"the extreme brevity of the financial memory.\"<blockquote>这就是历史上金融市场上反复发生的非常糟糕的事情，加尔布雷斯所说的“金融记忆的极度短暂”促成了这一点。</blockquote> When Modigliani says that there are two \"warranted\" prices, he means that - at least in the short run - there are two ways that prices can fulfill the expectations of investors. In one case, investors have expectations about future returns, and those expectations are informed by the level of valuations. If prices rise, and expected cash flows haven't changed, investors recognize that future returns will be lower. In the \"bubble\" case, investors have high expectations about future returns, mainly based on past returns, and they act on those expectations by driving prices up further. So the expectation of additional price increases is simply reinforced. \"The expectation of growth produces the growth, which confirms the expectation.\"<blockquote>当莫迪里阿尼说有两种“合理”的价格时，他的意思是——至少在短期内——价格有两种方式可以满足投资者的预期。在一种情况下，投资者对未来回报有预期，而这些预期是由估值水平决定的。如果价格上涨，而预期现金流没有改变，投资者就会认识到未来的回报会更低。在“泡沫”案例中，投资者对未来回报有很高的预期，主要基于过去的回报，他们通过进一步推动价格上涨来实现这些预期。因此，进一步涨价的预期只是被强化了。“增长的预期产生了增长，增长证实了预期。”</blockquote> Only one of these prices is consistent, in that the rate of return expected by investors is also the rate of return that equates price with discounted future cash flows. The other price becomes increasingly detached from fundamentals, as a larger and larger fraction of the price represents hot air, and it ultimately collapses as the gap becomes untenably wide.<blockquote>这些价格中只有一个是一致的，因为投资者预期的回报率也是将价格等同于贴现的未来现金流的回报率。另一个价格变得越来越脱离基本面，因为价格中越来越大的部分代表着空话，随着差距变得难以维持，它最终会崩溃。</blockquote> During speculative segments of the market cycle, there's nothing that forces investors to recognize that higher valuations imply lower returns, or to change their expectation of high returns as far as the eye can see. That, of course, is why we use measures of market internals to gauge the inclination of investors toward speculation or risk-aversion. Valuation provides an enormous amount of information about likely long-term returns and potential market losses over the complete cycle. But valuation isn't a timing tool. In recent years, it hasn't even imposed a limit on speculative recklessness.<blockquote>在市场周期的投机阶段，没有什么能迫使投资者认识到更高的估值意味着更低的回报，或者改变他们对高回报的预期。当然，这就是为什么我们使用市场内部因素来衡量投资者的投机或避险倾向。估值提供了大量关于整个周期中可能的长期回报和潜在市场损失的信息。但估值不是择时工具。近年来，它甚至没有对投机鲁莽行为施加限制。</blockquote> Still, with each price advance, the actual long-term return implied by future cash flows - what investors will ultimately realize as those cash flows are delivered - collapses further, even while investors act on their delusion that long-term returns have nothing to do with price. Eventually, the bulk of the security price represents a bubble component, not the price that would actually need to exist in order for the long-term expectations of investors to be accurate.<blockquote>尽管如此，随着价格的每一次上涨，未来现金流隐含的实际长期回报——投资者最终将在这些现金流交付时意识到的回报——进一步崩溃，即使投资者按照长期回报与价格无关的错觉行事。最终，证券价格的大部分代表了泡沫成分，而不是为了使投资者的长期预期准确而实际需要存在的价格。</blockquote> As I wrote at the 2000 market peak:<blockquote>正如我在2000年市场高峰时所写的：</blockquote> \"As long as investors focus on year-to-year returns and not discounted cash flow calculations, the bubble can continue to grow in self-reinforcing fashion. Investors anticipate a high return, and the price behavior reinforces the expectation. The true long-term return becomes increasingly detached from the long-term return imagined by investors, and the bubble component accounts for an increasingly large proportion of the total price.\" By our most reliable measures, run-of-the-mill historical valuation norms are roughly 70% below current levels. I know you don't want to believe that.<blockquote>“只要投资者关注的是同比回报，而不是贴现现金流计算，泡沫就能以自我强化的方式持续增长。投资者预期高回报，价格行为强化了预期。真正的长期回报与投资者想象的长期回报越来越脱离，泡沫成分在总价中的占比越来越大。”根据我们最可靠的衡量标准，一般的历史估值标准比当前水平低约70%。我知道你不愿意相信。</blockquote> <img src=\"https://static.tigerbbs.com/493c338076df85e2a04ebf892af4762f\" tg-width=\"1280\" tg-height=\"713\"><blockquote></blockquote> The trap door quietly swings open when valuations are extreme and market internals begin to deteriorate. That's the situation we've observed in our measures in recent weeks, with the initial deterioration largely driven by debt securities, but with increasing divergences in equities as well.<blockquote>当估值极端且市场内部开始恶化时，活板门就会悄然打开。这就是我们最近几周在测量中观察到的情况，最初的恶化主要是由债务证券推动的，但股票的分歧也在扩大。</blockquote> Valuations and discounted cash flows<blockquote>估值及贴现现金流量</blockquote> Valuations measure the tradeoff between current prices and a very long-term stream of expected future cash flows. Every useful valuation ratio is just shorthand for that calculation. Every valuation ratio that fails that criterion is inferior, and you can show it in historical data. - John P. Hussman, Ph.D., The Meaning of Valuation, December 2019 I've often noted that the denominator of every good valuation measure is just shorthand for the decades and decades of cash flows that the security is likely to deliver in the future. In fact, we always test the validity of the valuation measures we use by examining:<blockquote>估值衡量当前价格与长期预期未来现金流之间的权衡。每一个有用的估值比率都只是该计算的简写。每一个不符合这个标准的估值比率都是劣质的，你可以在历史数据中显示出来。–John P.Hussman博士，《估值的意义》，2019年12月我经常注意到，每一个好的估值衡量标准的分母都只是证券未来可能提供的数十年现金流的简写。事实上，我们总是通过检查来测试我们使用的估值方法的有效性：</blockquote> a) how strongly the resulting valuation measures are correlated with actual subsequent total returns, and;<blockquote>a)由此产生的估值措施与实际后续总回报的相关性有多强；</blockquote> b) how closely they replicate an explicit discounted cash flow analysis.<blockquote>b)它们在多大程度上复制了明确的贴现现金流分析。</blockquote> Below, we'll examine a variety of valuation measures that offer some perspective on why I view the U.S. equity market as a bubble near the breaking point. Along the way, I'll point out some interesting features of valuations and their relationship with subsequent returns. If math gives you hives, feel free to skim over the small amount that I've included here and there.<blockquote>下面，我们将研究各种估值指标，这些指标可以解释为什么我认为美国股市是一个接近崩溃点的泡沫。在此过程中，我将指出估值的一些有趣特征及其与后续回报的关系。如果数学让你荨麻疹，请随意浏览我在这里和那里包含的少量内容。</blockquote> Consider first the relationship between valuations and subsequent returns. I'll state the following, which you can prove to yourself by toying around a bit with present value models: the logarithm of a good valuation measure should have an inverse and roughly linear relationship with the expected subsequent investment return.<blockquote>首先考虑估值和后续回报之间的关系。我将陈述以下内容，你可以通过玩弄现值模型来向自己证明这一点：一个好的估值衡量标准的对数应该与预期的后续投资回报呈反比且大致呈线性关系。</blockquote> Here's a simple example of what this looks like.<blockquote>这是一个简单的例子。</blockquote> <img src=\"https://static.tigerbbs.com/d1e5bdf4391c9c91a8b9a051f6635116\" tg-width=\"819\" tg-height=\"708\"><blockquote></blockquote> Here's what this looks like for MarketCap/GVA, our most reliable stock market valuation measure<blockquote>这是我们最可靠的股市估值指标MarketCap/GVA的情况</blockquote> <img src=\"https://static.tigerbbs.com/485141919c853722522159d2e9d05e85\" tg-width=\"1280\" tg-height=\"720\"><blockquote></blockquote> During the past three decades, we've studied and introduced a broad range of valuation measures. Most can be calculated back to 1947. Some can be evaluated over a century or more. Across history, even in recent decades, we find that the valuation measures that are best correlated with actual subsequent returns are those with muted sensitivity to cyclical fluctuations in profit margins, and that behave largely like broad, market-wide price/revenue ratios.<blockquote>在过去的三十年里，我们研究并引入了广泛的估值方法。大多数可以追溯到1947年。有些可以评估一个世纪或更长时间。纵观历史，甚至在最近几十年，我们发现与实际后续回报最相关的估值指标是那些对利润率周期性波动敏感性较低的估值指标，并且在很大程度上类似于广泛的全市场市盈率。</blockquote> If we compare our most reliable valuation measures with the valuation measures that one would obtain from a proper long-term discounted cash flow analysis, we find that they're nearly identical. Here's what this comparison looks like for the actual stream of dividends (including the impact of repurchases) delivered by the S&amp;P 500 since 1900, discounted at a fixed 10% rate (see the chart text for additional details).<blockquote>如果我们将最可靠的估值指标与从适当的长期贴现现金流分析中获得的估值指标进行比较，我们会发现它们几乎相同。以下是标普500自1900年以来实际股息流（包括回购的影响）的比较，按10%的固定利率贴现（有关更多详细信息，请参阅图表文本）。</blockquote> The reason we use a fixed rate of return is that a multiple of 1.0 is then, by definition, the level at which the S&amp;P 500 would have been priced for that particular level of expected return. Any deviation in the valuation multiple from 1.0 then gauges how far likely future returns are from that \"typical\" expected return. We're currently farther away from \"typical\" expected returns than at any moment in history, including the 1929 and 2000 market peaks.<blockquote>我们使用固定回报率的原因是，根据定义，1.0的倍数就是该特定预期回报水平的标普500定价水平。估值倍数与1.0的任何偏差都可以衡量未来回报与“典型”预期回报的可能性有多大。我们目前比历史上任何时候都更远离“典型”预期回报，包括1929年和2000年的市场峰值。</blockquote> <img src=\"https://static.tigerbbs.com/e6f61d2794095c2545d0d2c130320e45\" tg-width=\"1280\" tg-height=\"722\"><blockquote></blockquote> One of the unfortunate bits of financial illiteracy that Wall Street has pushed into the heads of investors is the idea that extreme valuations are \"justified\" by low interest rates. Now, it's certainly true that holding future cash flows constant, raising the price of an investment will lower the embedded rate of return, and vice versa. If you pay $32 today for $100 a decade from today, you can expect a 12% annual return. If you pay $82 for the same security, you can expect a 2% annual return. If you pay $100 today, you can expect nothing. So it's clearly true that holding future cash flows constant, a lower rate of return implies a higher level of valuation.<blockquote>华尔街向投资者灌输的一个不幸的金融文盲是，低利率“证明”极端估值是合理的。现在，保持未来现金流不变，提高投资价格会降低内含回报率，反之亦然，这是事实。如果你今天支付32美元，十年后支付100美元，你可以期待12%的年回报率。如果你为同样的证券支付82美元，你可以期待2%的年回报率。如果你今天付100美元，你什么也不能指望。因此，很明显，保持未来现金流不变，较低的回报率意味着较高的估值水平。</blockquote> The reason the statement \"low interest rates justify high valuations\" contributes to financial illiteracy is that the statement has been learned entirely out of context of the arithmetic. As a result, investors seem to imagine that, as long as high valuations can be \"justified,\" stocks can be expected to provide historically normal rates of return in the future. Likewise, investors seem to have no concept that if interest rates are low because growth rates are low, no valuation premium is \"justified\" for stocks, because the lower growth is already sufficient to bring future stock returns down to levels that are commensurate with the low level of interest rates.<blockquote>“低利率证明高估值是合理的”这句话导致金融文盲的原因是，这句话完全是在算术背景下学习的。因此，投资者似乎想象，只要高估值能够“合理”，股票就有望在未来提供历史上正常的回报率。同样，投资者似乎没有概念，如果利率低是因为增长率低，那么股票的估值溢价就“合理”，因为较低的增长已经足以将未来的股票回报降至与低利率水平相称的水平。</blockquote> The truth is simple but uncomfortable. If interest rates are low and expected growth is held constant, higher valuations imply lower long-term returns. If interest rates are low because expected future growth is also low, higher valuations are not required. Long-term returns will be lower anyway. A valuation premium just makes future returns even worse.<blockquote>道理很简单，却让人难受。如果利率较低且预期增长保持不变，较高的估值意味着较低的长期回报。如果利率较低，因为预期的未来增长也较低，则不需要更高的估值。无论如何，长期回报都会更低。估值溢价只会让未来的回报变得更糟。</blockquote> Saying that extremely low interest rates \"justify\" extremely high stock valuations is identical to saying that extremely low future returns on bonds \"justify\" extremely low future returns on stocks. I don't really think that's something Wall Street cares to clarify when it tells investors that stock market valuations are \"justified.\"<blockquote>说极低的利率“证明”极高的股票估值是合理的，就像说极低的债券未来回报率“证明”极低的股票未来回报率是一样的。当华尔街告诉投资者股市估值“合理”时，我真的不认为这是它想澄清的事情。</blockquote> Investment valuation is concerned with the relationship between three objects: the current price, the future cash flows, and the rate of return that connects the two like a string. The lower the current price and the higher the future cash flows, the steeper the string. The higher the current price and the lower the future cash flows, the flatter the string. Raise the current price above the future cash flows, and the string will point down instead of up. Given any two of these objects, you can calculate the third one.<blockquote>投资估值关注的是三个对象之间的关系：当前价格、未来现金流以及像绳子一样将两者连接起来的回报率。当前价格越低，未来现金流越高，弦越陡。当前价格越高，未来现金流越低，字符串就越扁平。将当前价格提高到未来现金流之上，字符串将指向下方而不是上方。给定这些对象中的任意两个，你可以计算第三个。</blockquote> For example, if you want to estimate expected long-term returns, you need two objects: a) the current price and b) the expected stream of future cash flows. A good valuation ratio is just shorthand for those two objects, so you can estimate the long-term return directly from the level of valuation. Then, if you like, you can compare it with the level of interest rates. That comparison can be useful, because even when investors realize that high stock market valuations imply low long-term returns, it's not at all clear that they realize how low long-term return prospects have been driven.<blockquote>例如，如果你想估计预期的长期回报，你需要两个对象：a）当前价格和b）未来现金流的预期流。一个好的估值比率只是那两个对象的速记，所以你可以直接从估值水平来估计长期回报。然后，如果你愿意，你可以把它和利率的水平进行比较。这种比较可能是有用的，因为即使投资者意识到高股市估值意味着低长期回报，他们也根本不清楚他们是否意识到长期回报前景有多低。</blockquote> The chart below shows our estimate of expected 12-year S&amp;P 500 total returns over-and-above Treasury bond yields, across a century of market history. Compare this with the nearly useless drivel that Wall Street passes off as the \"equity risk premium\" (typically quoted as the S&amp;P 500 forward earnings yield minus the 10-year Treasury yield). Yes, you're reading the chart correctly. Given current valuations, we expect the total return of the S&amp;P 500 to underperform the lowly yield on Treasury bonds by roughly -6% annually over the coming 12-year period.<blockquote>下图显示了我们对一个世纪的市场历史中预期12年期标普500总回报率高于国债收益率的估计。相比之下，华尔街称之为“股票风险溢价”（通常被引用为标普500远期收益收益率减去10年期国债收益率）的几乎毫无用处的废话。是的，你没看错图表。鉴于目前的估值，我们预计未来12年标普500的总回报率将低于国债的低收益率约-6%。</blockquote> <img src=\"https://static.tigerbbs.com/42561963df9d259a93fe0d4f2502c13e\" tg-width=\"1024\" tg-height=\"577\"><blockquote></blockquote> Many investors confuse the estimation of expected returns (which requires only expected cash flows and the observed price) with a different problem - the estimation of \"fair value.\" See, interest rates come into the picture when you have an expected stream of future cash flows and you want to calculate a \"fair\" current price. In that case, rather than picking an arbitrary rate of return from a hat, it's common to use the level of interest rates, plus some risk premium, as the expected long-term return (or discount rate, or capitalization rate). This sort of calculation can be super-sensitive to arbitrary choices.<blockquote>许多投资者将预期回报的估计（只需要预期现金流和观察到的价格）与另一个问题——“公允价值”的估计混为一谈。看，当你有一个预期的未来现金流，你想要计算一个“公平”的当前价格时，利率就会出现。在这种情况下，通常使用利率水平加上一些风险溢价作为预期的长期回报（或贴现率或资本化率），而不是从帽子中选择任意的回报率。这种计算对任意选择非常敏感。</blockquote> In particular, Wall Street loves to combine super-high growth rates, super-low discount rates, and super-long time horizons, which allows one to calculate a \"fair value\" that's as close to infinity as possible. The thing to remember is that whatever rate of return an analyst embeds into the fair value calculation is also the long-term rate of return you'll earn over time if you pay that price, assuming the future cash flows are delivered as expected.<blockquote>特别是，华尔街喜欢将超高的增长率、超低的贴现率和超长的时间范围结合起来，这使得人们可以计算出尽可能接近无穷大的“公允价值”。需要记住的是，无论分析师在公允价值计算中嵌入什么回报率，假设未来现金流按预期交付，如果您支付该价格，随着时间的推移您将获得的长期回报率。</blockquote> Among scores of measures we've evaluated or introduced over time, MarketCap/GVA (nonfinancial market capitalization to corporate gross value-added, including our estimate of foreign revenues) has the highest correlation with actual subsequent 10-12 year S&amp;P 500 total returns, followed by our Margin-Adjusted P/E (MAPE). I find it hilarious that the various valuation measures I've introduced over time are sometimes described as products of \"machine learning,\" \"data mining,\" and \"curve fitting\" when they are, in fact, just different versions of an apples-to-apples economy-wide price/revenue ratio.<blockquote>在我们随着时间的推移评估或引入的数十项指标中，MarketCap/GVA（非金融市值与企业总增加值之比，包括我们对国外收入的估计）与随后10-12年标普500实际总回报的相关性最高，其次是我们的利润率调整市盈率(MAPE)。我觉得好笑的是，随着时间的推移，我引入的各种估值指标有时被描述为“机器学习”、“数据挖掘”和“曲线拟合”的产物，而事实上，它们只是整个经济范围内市盈率的不同版本。</blockquote> <img src=\"https://static.tigerbbs.com/75dac426a3ad2f3830f9dce50f7c0a1c\" tg-width=\"1265\" tg-height=\"712\"><blockquote></blockquote> The S&amp;P 500 price/revenue ratio and nonfinancial market capitalization to GDP (the \"Buffett indicator\") also perform well, and better than earnings-based alternatives like price/forward earnings, the Fed Model, and even the Shiller CAPE. See, while earnings are necessary to generate long-term cash flows, they are also subject to fluctuations in profit margins (both cyclically and even from decade to decade) that turn out to be highly uninformative.<blockquote>标普500市盈率和非金融市值与GDP之比（“巴菲特指标”）也表现良好，优于基于收益的替代指标，如价格/远期收益、美联储模型，甚至席勒CAPE。看，虽然收益对于产生长期现金流是必要的，但它们也会受到利润率波动的影响（周期性地，甚至每十年一次），而事实证明，这些波动是非常缺乏信息的。</blockquote> Economically, fluctuations in profit margins are driven primarily by fluctuations in real unit labor costs. Because companies compete on the basis of realized after-tax profits rather than pre-tax profits, changes in tax policy also have far less durable impact on corporate profits than investors seem to imagine. While corporate profits got a tremendous boost last year from CARES spending (the deficit of one sector always emerges as the surplus of another), here's what the relationship between corporate profits and real unit labor costs looks like historically.<blockquote>在经济上，利润率的波动主要是由实际单位劳动力成本的波动驱动的。由于企业是基于已实现的税后利润而非税前利润进行竞争，税收政策的变化对企业利润的持久影响也远不如投资者似乎想象的那样。虽然去年企业利润从医疗保健支出中获得了巨大的提振（一个部门的赤字总是表现为另一个部门的盈余），但以下是企业利润和实际单位劳动力成本之间的历史关系。</blockquote> Real unit labor costs are presented on an inverted left scale. The upward pressure on labor costs (observed as a plunge in the blue line) isn't particularly auspicious for future profits. Still, there's so much distortion in recent quarters that I'd consider the jury to be out on how much of this will be sustained.<blockquote>实际单位劳动力成本以左倒比例呈现。劳动力成本的上行压力（观察为蓝线暴跌）对未来利润来说并不是特别有利。尽管如此，最近几个季度仍存在如此多的扭曲，以至于我认为这种情况会持续多久还没有定论。</blockquote> <img src=\"https://static.tigerbbs.com/3110e7d3c0468c39d343b20835566bcc\" tg-width=\"1157\" tg-height=\"652\"><blockquote></blockquote> It's undoubtedly true that profit margins, expected growth, and other factors have an effect on future deliverable cash flows and the valuations that investors place on stocks. But even for technology stocks, these assumptions should be made explicit and tested against history. You'll find that observable measures like price/revenue are still very serviceable. In fact, the extreme price/revenue multiples of technology stocks helped to inform my March 2000 projection of an 83% loss in that sector.<blockquote>毫无疑问，利润率、预期增长和其他因素会影响未来可交付的现金流和投资者对股票的估值。但即使对于科技股来说，这些假设也应该明确，并根据历史进行检验。你会发现像价格/收入这样的可观察指标仍然非常有用。事实上，科技股极端的市盈率帮助我在2000年3月预测该行业将损失83%。</blockquote> As Benjamin Graham wrote, \"The habit of relating what is paid to what is being offered is an invaluable trait in investment. The more dependent the valuation becomes on anticipations of the future - and the less it is tied to a figure demonstrated by past performance - the more vulnerable it becomes to possible miscalculation and serious error.\"<blockquote>正如本杰明·格雷厄姆（Benjamin Graham）所写的那样，“将支付的东西与提供的东西联系起来的习惯是投资中的一个无价特征。估值越依赖于对未来的预期——它与过去表现所证明的数字的联系越少——它就越容易受到可能的误判和严重错误的影响。”</blockquote> The current 5.19 price/revenue multiple for the Nasdaq 100, is the most extreme level since February 2001, which was followed by a 60% loss in the index (after it had already dropped in half). The situation is actually a bit worse than 2001 here. If one examines the largest components of the index, it becomes clear that their annual growth rates have declined substantially over time.<blockquote>纳斯达克100指数目前的市盈率为5.19，是自2001年2月以来的最极端水平，随后该指数下跌了60%（此前该指数已经下跌了一半）。这里的情况实际上比2001年还要糟糕一些。如果我们研究该指数的最大组成部分，就会发现它们的年增长率随着时间的推移大幅下降。</blockquote> As a result, a dollar of current revenues should arguably command a smaller multiple than a dollar of revenues might have during earlier periods of emerging growth. Yet even if one takes the Nasdaq 100 price/revenue ratio at face value, and even if one restricts attention to the bubble period since 2000, it's difficult to expect the Nasdaq to produce total returns over the coming decade much better than zero.<blockquote>因此，可以说，当前一美元收入的倍数应该低于新兴增长早期一美元收入的倍数。然而，即使人们从表面上看待纳斯达克100指数的市盈率，即使人们将注意力限制在2000年以来的泡沫时期，也很难指望纳斯达克在未来十年产生比零好得多的总回报。</blockquote> <img src=\"https://static.tigerbbs.com/48dd88f74eaed48d53b237b4d0209d8e\" tg-width=\"963\" tg-height=\"542\"><blockquote></blockquote> You don't really want to see what the same chart looks like for the S&amp;P 500.<blockquote>你不会真的想看到标普500的同样的图表是什么样子的。</blockquote> <img src=\"https://static.tigerbbs.com/24f70ab683592ef20a8c6c75fc9bd1fc\" tg-width=\"963\" tg-height=\"543\"><blockquote></blockquote> The same is true, unfortunately, for passive investment strategies. We presently estimate negative 12-year average annual total returns for a conventional passive investment mix invested 60% in the S&amp;P 500, 30% in Treasury bonds, and 10% in Treasury bills.<blockquote>不幸的是，被动投资策略也是如此。我们目前估计，60%投资于标普500、30%投资于国债、10%投资于国库券的传统被动投资组合的12年平均年总回报率为负。</blockquote> In a 2019&nbsp;white paper, I detailed an approach to estimate a \"value-focused asset allocation\" by jointly considering prevailing stock market valuations and interest rates. It specifies an investment allocation based on which asset class is estimated to have the highest average annual expected return, adjusted for risk, to each point in a long-term investment horizon. That allocation can then be modified by a risk-management component, to adjust the exposure during segments of the market cycle where risk-aversion or speculation among market participants may temporarily drive valuations to depressed or elevated levels. The white paper includes numerous charts showing how this value-focused asset allocation has changed across market history, particularly at important peaks and troughs in the stock and bond markets.<blockquote>在2019年的白皮书中，我详细介绍了一种通过共同考虑当前股市估值和利率来估计“以价值为中心的资产配置”的方法。它根据哪个资产类别估计具有最高的平均年预期回报（经风险调整）来指定长期投资范围内的每个点的投资分配。然后，风险管理组件可以修改这种分配，以在市场周期的各个部分调整风险敞口，在这些部分，市场参与者的风险规避或投机可能会暂时将估值推低或推高。该白皮书包括大量图表，显示了这种以价值为中心的资产配置在市场历史上是如何变化的，特别是在股票和债券市场的重要高峰和低谷。</blockquote> Along with those methods, I introduced our \"Endowment/spending multiple,\" which estimates the number of years of spending that a passive 60/30/10 investor requires up-front, in order to finance an expected 36-year stream of future inflation-adjusted spending. The idea here is that in a deeply undervalued market with high expected future returns, investors can finance a future stream of spending with far less than they require when valuations are extreme and prospective returns are low.<blockquote>除了这些方法，我还介绍了我们的“捐赠/支出倍数”，它估计了被动的60/30/10投资者预先需要的支出年数，以便为未来36年的通胀调整支出提供资金。这里的想法是，在一个预期未来回报很高的被严重低估的市场中，当估值极端且预期回报较低时，投资者可以用远低于他们所需的资金为未来的支出流提供资金。</blockquote> You know you're in a bubble when funding a 36-year stream of expected inflation-adjusted spending requires over 38 years of money up-front.<blockquote>当为36年的预期通胀调整支出提供资金需要超过38年的前期资金时，你就知道自己处于泡沫之中。</blockquote> <img src=\"https://static.tigerbbs.com/1adff1ad0dbaa2737f60cb7589d206fb\" tg-width=\"1230\" tg-height=\"693\"><blockquote></blockquote> In the chart below, the Endowment/spending multiple is presented on an inverted log scale (left), along with the actual subsequent average annual total return of a 60/30/10 portfolio mix (right scale). Needless to say, we adhere to investment disciplines that are intended to address problems like this.<blockquote>在下图中，捐赠/支出倍数以倒对数标度（左）以及60/30/10投资组合的实际后续平均年总回报（右标度）呈现。不用说，我们坚持旨在解决此类问题的投资纪律。</blockquote> <img src=\"https://static.tigerbbs.com/1480e881b4d04ace81a96f31f6368796\" tg-width=\"1229\" tg-height=\"692\"><blockquote></blockquote> As a testament to the breadth of this speculative episode, the median price/revenue ratio of S&amp;P 500 components now exceeds 3.3, easily a record, and extreme enough to provoke distress about the potential losses that innocent but poorly-informed investors may experience over the completion of this cycle.<blockquote>作为这一投机事件广度的证明，标普500成分股的市盈率中位数现在超过了3.3，很容易创下纪录，而且极端程度足以引发无辜但消息不灵通的投资者在完成这一周期后可能遭受的潜在损失。</blockquote> <img src=\"https://static.tigerbbs.com/ffe0cc33bdaf0695dacb7ae1d817edab\" tg-width=\"1110\" tg-height=\"625\"><blockquote></blockquote> The next chart shows the median price/revenue ratio of S&amp;P 500 components sorted into 10 deciles by valuation. The chart is presented on log scale to allow each line to be compared with its own history. Each segment on the vertical axis represents a doubling of valuations. Notice that every single decile of S&amp;P 500 components is at record valuation extremes. Investors now rely on a permanently high plateau in these extremes.<blockquote>下图显示了按估值分为10个十分位数的标普500组件的市盈率中位数。该图表以对数标度显示，以便将每条线与其自身的历史进行比较。纵轴上的每个细分市场代表估值翻倍。请注意，标普500成分股的每一个十分位数都处于创纪录的估值极端水平。投资者现在依赖于这些极端情况下的永久高位平台。</blockquote> <img src=\"https://static.tigerbbs.com/cb7c995e64e3c9cba85d6e0fc9124223\" tg-width=\"1280\" tg-height=\"721\"><blockquote></blockquote> It's interesting, but far less important, that the median price/earnings ratio of S&amp;P 500 components has also reached 32.4, the highest level in history. This compares with a median P/E of 19.4 at the 2000 market peak. The problem with P/E multiples, of course, is that they are substantially affected by earnings variability. In fact, prior to the current peak, the highest median P/E for the S&amp;P 500 was actually in March 2002, when the index was down 25% from the March 2000 bubble highs. That's because earnings were down far more by then, which boosted P/E ratios. Such is the danger of taking P/E multiples at face value.<blockquote>有趣的是，但远没有那么重要的是，标普500组件的市盈率中位数也达到了32.4，为历史最高水平。相比之下，2000年市场峰值时的市盈率中值为19.4。当然，市盈率倍数的问题在于它们很大程度上受到收益可变性的影响。事实上，在当前的峰值之前，标普500的最高市盈率中值实际上是在2002年3月，当时该指数比2000年3月的泡沫高点下跌了25%。这是因为当时盈利下降幅度更大，这提高了市盈率。这就是按面值计算市盈率倍数的危险。</blockquote> A crude but reasonably effective way to get around the cyclicality of earnings (but only for very broad indices), is to compare prices to the highest level of earnings achieved to-date. I introduced this metric back in 1998 as the price-to-peak-earnings ratio. The chart below shows a version of that. The blue line (left scale) shows the ratio of total U.S. equity market capitalization to GDP. The red line (right scale) shows the ratio of total U.S. equity market capitalization to the highest level of economy-wide U.S. profits to date. Investors have gotten themselves into trouble here.<blockquote>绕过收益周期性的一个粗略但相当有效的方法（但仅限于非常广泛的指数）是将价格与迄今为止达到的最高收益水平进行比较。我早在1998年就引入了这个指标，称为市盈率。下图显示了一个版本。蓝线（左标度）显示了美国股票总市值与GDP的比率。红线（右标度）显示了美国股票总市值与迄今为止美国整体经济利润最高水平的比率。投资者在这里遇到了麻烦。</blockquote> <img src=\"https://static.tigerbbs.com/a6630cf8302360e1b4645870051b21de\" tg-width=\"989\" tg-height=\"557\"><blockquote></blockquote> What if valuations remain extreme forever?<blockquote>如果估值永远保持极端怎么办？</blockquote> Probably the single most frequent question I've heard from investors over the past couple of years is \"What happens if valuations remain extreme forever?\" It's actually a version of the \"permanently high plateau\" that Irving Fisher disastrously projected in 1929. Still, recent years have produced enormous confidence among investors that the Federal Reserve's purchases of Treasury debt can permanently \"backstop\" the stock market.<blockquote>过去几年我从投资者那里听到的最常见的问题可能是“如果估值永远保持极端会发生什么？”这实际上是欧文·费希尔在1929年灾难性地预测的“永久高原”的一个版本。尽管如此，近年来投资者对美联储购买国债可以永久“支撑”股市充满信心。</blockquote> As I've detailed previously, quantitative easing supports the market only by creating zero interest hot potatoes that are uncomfortable for investors to hold (provided that they're inclined toward speculation), and that are impossible to get rid of in aggregate. Moreover, the Fed's purchases of corporate bonds during the pandemic were legally constrained to CARES funds provided by the Treasury, and ultimately amounted to $14 billion of bonds, in an economy with $11 trillion in corporate debt at $58 trillion in equity market capitalization.<blockquote>正如我之前详细介绍的，量化宽松只能通过创造零利息的烫手山芋来支持市场，这些烫手山芋让投资者持有起来不舒服（前提是他们倾向于投机），而且总体上不可能摆脱这些烫手山芋。此外，在一个拥有11万亿美元公司债务、58万亿美元股票市值的经济体中，美联储在疫情期间购买的公司债券在法律上仅限于财政部提供的CARES资金，最终债券金额为140亿美元。</blockquote> Suffice it to say that the \"Fed backstop\" is largely in the minds of investors, and relies almost exclusively on the psychological discomfort of holding low-yielding base money. Yet since perception can be indistinguishable from reality, particularly in the short run, it's important to entertain the question.<blockquote>可以说，“美联储支持”很大程度上存在于投资者的脑海中，几乎完全依赖于持有低收益基础货币的心理不适。然而，由于感知可能与现实难以区分，尤其是在短期内，所以考虑这个问题很重要。</blockquote> My answer is that if the Fed is indeed able to maintain valuations at the highest levels in history, forever, stock prices would likely grow at roughly the same rate as nominal GDP. So figure 1.6% real structural growth plus 2% inflation gets you 3.6%. Let's call it 4%, which would match nominal GDP growth in both the 10-year and 20-year periods ending at the Q4 2019 economic peak, just before the pandemic. The first casualty of rising inflation is stock valuations, so it's not at all clear that assuming higher inflation would help stocks until valuations were roughly normalized, which would require consumer prices to roughly triple.<blockquote>我的回答是，如果美联储确实能够永远将估值维持在历史最高水平，股价可能会以与名义GDP大致相同的速度增长。因此，1.6%的实际结构性增长加上2%的通胀率得到3.6%。让我们将其看涨期权为4%，这将与截至2019年第四季度经济峰值（就在大流行之前）的10年和20年期间的名义GDP增长相匹配。通胀上升的第一个受害者是股票估值，因此在估值大致正常化之前，假设通胀上升是否会对股票有所帮助，这一点还不清楚，这需要消费者价格大约上涨两倍。</blockquote> The chart below is a reminder of how structural real GDP growth has progressed over recent decades (driven by demographic labor force growth and trend productivity), and the basis for that 1.6% figure for structural real GDP growth.<blockquote>下图提醒我们近几十年来结构性实际GDP增长是如何进展的（由人口劳动力增长和趋势生产率驱动），以及结构性实际GDP增长1.6%数字的基础。</blockquote> <img src=\"https://static.tigerbbs.com/cf63b9b8f5f3b55136c8523d73e864f3\" tg-width=\"1131\" tg-height=\"637\"><blockquote></blockquote> Sticking with 4% nominal growth, and adding a 1.5% dividend yield, a \"permanently high plateau\" in market valuations would imply S&amp;P 500 total returns of about 5.5% annually. Again, this assumes that valuations never retreat from levels that presently stand at about 3.6 times their historical norms. Simply allow them to retreat to 2.4 times their historical norms a decade from now - which would still keep valuations among the highest 10% in U.S. history - and the resulting 10-year total return would drop to about 1.3%. I think this would actually be the best-case scenario even in a permanently overvalued world.<blockquote>坚持4%的名义增长率，加上1.5%的股息收益率，市场估值的“永久高位”将意味着标普500每年的总回报率约为5.5%。同样，这是假设估值永远不会从目前约为历史标准3.6倍的水平回落。只要让它们在十年后回落至历史正常水平的2.4倍——这仍将使估值保持在美国历史上最高的10%之内——由此产生的10年总回报率将降至1.3%左右。我认为，即使在一个永远被高估的世界里，这实际上也是最好的情况。</blockquote> At elevated valuations, even very small changes in expected return imply enormous changes in prices. So it's unlikely that a period of much higher average valuations will escape the prospect of relatively high volatility. Rather than a 70% market decline, which would presently be required for the S&amp;P 500 to simply touch historically run-of-the-mill valuation norms, investors could expect rather frequent market losses in the 20-35% range, which is essentially what we've seen even over the past few years.<blockquote>在高估值下，即使预期回报的微小变化也意味着价格的巨大变化。因此，平均估值高得多的时期不太可能逃脱相对较高波动性的前景。投资者可能预计市场会出现20-35%的频繁下跌，而不是70%的市场下跌，这是标普500目前简单地触及历史上的普通估值标准所必需的，这基本上是我们在过去几年中所看到的。</blockquote> All of that would be fine with us. We've adapted our discipline sufficiently (especially in late-2017) to tolerate the possibility of permanently sustained overvaluation. My impression is that the impact of those adaptations has become more evident as we've had greater opportunities to live into them. I can't say that I believe for a second that investors will actually be spared from a 50-70% loss in the S&amp;P 500 in the coming years, but again, it will be fine with us if the market never approaches historical valuation norms again. With the adaptations we introduced in late-2017, our discipline is flexible enough to navigate a bubble even without embracing its premise.<blockquote>所有这些对我们来说都没问题。我们已经充分调整了我们的纪律（尤其是在2017年末），以容忍永久持续高估的可能性。我的印象是，随着我们有更多的机会生活在其中，这些适应的影响变得更加明显。我不能说我一秒钟都相信投资者实际上不会在未来几年的标普500中遭受50-70%的损失，但同样，如果市场永远不会接近历史估值规范，我们也没问题。随着我们在2017年底引入的适应，我们的学科足够灵活，即使不接受泡沫的前提，也可以驾驭泡沫。</blockquote> An unusual overlap of high-risk conditions<blockquote>高风险状况的不寻常重叠</blockquote> Returning to Modigliani, my impression is that the advance of recent years to the most extreme valuations in history reflects exactly the bubble dynamics he described, and I expect that it will also end as he described (though not necessarily in one fell swoop): \"The expectation of growth produces the growth, which confirms the expectation; people will buy it because it went up. But once you are convinced that it is not growing anymore, nobody wants to hold a stock because it is overvalued. Everybody wants to get out and it collapses, beyond the fundamentals.\"<blockquote>回到莫迪里阿尼，我的印象是，近年来向历史上最极端估值的推进恰恰反映了他所描述的泡沫动态，我预计它也会像他所描述的那样结束（尽管不一定一蹴而就）：“对增长的预期产生了增长，增长证实了预期；人们会因为它上涨而买入。但一旦你确信它不再增长，就没有人愿意持有一只股票，因为它被高估了。每个人都想退出，但它崩溃了，超出了基本面。”</blockquote> One of our internal gauges tracks the correlation of market conditions with certain high-risk features that have preceded steep market collapses - a collection of measures capturing valuations, internals, sentiment, leverage, overextension, and yield pressures. Only a handful of instances in history overlap pre-crash conditions as well as they do at present. The current overlap is actually quite similar to August 1987. Meanwhile, the correlation of current conditions with features typically observed at market lows is the most negative in history.<blockquote>我们的一个内部指标跟踪市场状况与市场急剧崩溃之前的某些高风险特征的相关性——一系列衡量估值、内部因素、情绪、杠杆、过度扩张和收益率压力的指标。历史上只有少数情况像现在这样与崩溃前的情况重叠。目前的重叠实际上与1987年8月非常相似。与此同时，当前状况与市场低点时通常观察到的特征的相关性是历史上最负的。</blockquote> If you want my opinion, I suspect that a near-vertical market plunge on the order of 25-35% is coming, probably quite shortly, most likely out of the blue, as in 1987, driven by nothing more than the sudden concerted effort of overextended investors to sell, and the need for a large price adjustment in order to induce scarce buyers to take the other side.<blockquote>如果你想听听我的意见，我怀疑25-35%左右的近乎垂直的市场暴跌即将到来，可能很快，很可能是出乎意料的，就像1987年一样，无非是由过度扩张的投资者突然齐心协力抛售，以及需要大幅调整价格以诱导稀缺买家站在另一边。</blockquote> As usual, no forecasts are necessary. We'll align our investment stance in response to the valuations and market action that we observe at each point in time. Still, it's of particular concern that these overlaps are occurring in the context of the most extreme valuations in history, along with strikingly dysfunctional pockets of illiquidity in many individual issues. This dysfunctional behavior isn't about any particular video game retailer. I suspect it's actually about some sort of fragility or segmentation in order-flow mechanisms, possibly coupled with poorly managed derivatives exposure.<blockquote>像往常一样，不需要预测。我们将根据每个时间点观察到的估值和市场行为来调整我们的投资立场。尽管如此，特别令人担忧的是，这些重叠发生在历史上最极端估值的背景下，以及许多个别问题中明显功能失调的流动性不足的背景下。这种不正常的行为与任何特定的视频游戏零售商无关。我怀疑这实际上是关于订单流机制的某种脆弱性或分割，可能与管理不善的衍生品敞口有关。</blockquote> As I used to teach my students, show me a financial debacle, and I'll show you someone who had a leveraged, mismatched position that they were suddenly forced to close into an illiquid market.<blockquote>正如我曾经教我的学生的那样，给我看一个金融崩溃，我会给你看一个拥有杠杆、不匹配头寸的人，他们突然被迫关闭到一个缺乏流动性的市场。</blockquote> Though my concerns run far beyond the amount of leverage in the system, it isn't helpful that the amount of leverage in the U.S. equity markets is now easily the highest in history. Some observers are inclined to bring this figure down by dividing instead by the market capitalization of equities. But here's some useful arithmetic:<blockquote>尽管我的担忧远远超出了系统中的杠杆金额，但美国股市的杠杆金额现在轻松达到历史最高水平并没有什么帮助。一些观察家倾向于通过除以股票市值来降低这一数字。但这里有一些有用的算术：</blockquote> Margin Debt/GDP = Margin Debt/Market Cap x Market Cap/GDP<blockquote>保证金债务/GDP=保证金债务/市值x市值/GDP</blockquote> To say that margin debt to GDP is at the highest level in history is to say not only that stocks are heavily owned on margin, but that those stocks are also breathtakingly overvalued. That combination is particularly worrisome.<blockquote>说保证金债务占GDP的比例处于历史最高水平，不仅是说保证金大量持有股票，而且这些股票的估值也高得惊人。这种组合尤其令人担忧。</blockquote> All crises have involved debt that, in one fashion or another, has become dangerously out of scale in relation to the underlying means of payment. - John Kenneth Galbraith, A Short History of Financial Euphoria <img src=\"https://static.tigerbbs.com/dc55dacbd5e01016cd9964422f941b0c\" tg-width=\"983\" tg-height=\"556\"><blockquote>所有危机都涉及债务，这些债务以这样或那样的方式与基本支付手段相比已经危险地超出了规模。——约翰·肯尼思·加尔布雷斯，《金融兴奋简史》</blockquote> The kind of event I'm suggesting would not bring valuations anywhere near historical valuation norms. Given current valuation extremes, it would be more like a palate cleanser. I have no particular expectation about what the next dish would be. In the event we do observe an abrupt market collapse, the Fed will undoubtedly respond with some new palliative. Whether or not it is effective will depend on the context of risk-aversion, inflation, credit risk, and other conditions at the time. The larger problem, as we've discussed, is that you can't \"save\" an overvalued asset by propping up its price. The value is in the future cash flows that will be delivered to investors over time. The elevated price only ensures that the long-term return between now and then will be dismal.<blockquote>我所建议的这种事件不会使估值接近历史估值规范。鉴于目前的极端估值，它更像是一种味觉清洁剂。我对下一道菜没有特别的期望。如果我们确实观察到市场突然崩溃，美联储无疑会采取一些新的姑息措施来应对。它是否有效将取决于当时的风险规避、通货膨胀、信用风险和其他条件。正如我们所讨论的，更大的问题是，你不能通过支撑价格来“拯救”被高估的资产。价值在于随着时间的推移将交付给投资者的未来现金流。价格上涨只会确保从现在到那时的长期回报将是惨淡的。</blockquote> Of course, nothing in our discipline relies on a market plunge. Given the combination of hypervaluation and divergent market internals (largely based on debt securities, but with increasing divergences in equities as well), I do believe that the stock market remains in a \"trap door\" situation. Still, that view will change as market conditions change. We'll refrain from adopting or amplifying a negative outlook if our measures of market internals improve. That discipline has served us well even amid record highs. Again, no forecasts are required, nor does this opinion drive our current investment stance. I just think the correlation with historical pre-crash conditions is worth noting.<blockquote>当然，我们的学科中没有任何东西依赖于市场暴跌。鉴于高估和市场内部分歧（主要基于债务证券，但股票分歧也在增加）的结合，我确实认为股市仍处于“活板门”的情况。尽管如此，这种观点将随着市场状况的变化而改变。如果我们对市场内部的衡量有所改善，我们将避免采用或放大负面前景。即使在创纪录的高点下，这种纪律也为我们提供了良好的服务。同样，不需要预测，这种观点也不会驱动我们当前的投资立场。我只是认为与历史碰撞前情况的相关性值得注意。</blockquote> How little, it will perhaps be agreed, was either original or otherwise remarkable about this history. Prices driven up on the expectation that they would go up, the expectation realized by the resulting purchases. Then the inevitable reversal of these expectations because of some seemingly damaging event or development or perhaps merely because the supply of intellectually vulnerable buyers was exhausted. Whatever the reason (and it is unimportant), the absolute certainty is that this world ends not with a whimper but with a bang. And so on to the moment of mass disillusion and the crash. This last, it will now be sufficiently evident, never comes gently. It is always accompanied by a desperate and largely unsuccessful effort to get out. - John Kenneth Galbraith on the 1929 collapse Valuations and investment duration<blockquote>也许人们会同意，关于这段历史，无论是原创的还是非凡的，都很少。价格上涨是基于价格会上涨的预期，这种预期通过由此产生的购买来实现。然后，由于一些看似破坏性的事件或发展，或者仅仅是因为智力脆弱的买家的供应耗尽，这些预期不可避免地逆转。不管原因是什么（这并不重要），绝对可以肯定的是，这个世界不是随着呜咽而结束，而是随着一声巨响而结束。等等，直到大规模幻灭和崩溃的时刻。这最后一点，现在已经足够明显了，从来不会轻易到来。它总是伴随着绝望且基本上不成功的脱身努力。——约翰·肯尼思·加尔布雷斯谈1929年的崩盘估值和投资期限</blockquote> To understand why extreme valuations imply high volatility, and require extremely long investment horizons, it's useful to consider the concept of \"duration.\"<blockquote>为了理解为什么极端估值意味着高波动性，并且需要极长的投资期限，考虑“久期”的概念是有用的。</blockquote> Every security is a claim on a stream of future cash flows that can be expected to be delivered to the investor over time. While the concept of \"duration\" is most commonly used for bonds, it's actually applicable to any security, no matter how \"lumpy\" the stream of cash flows might be.<blockquote>每种证券都是对未来现金流的债权，这些现金流预计将随着时间的推移交付给投资者。虽然“久期”的概念最常用于债券，但它实际上适用于任何证券，无论现金流有多“不稳定”。</blockquote> If you don't like math, feel free to skim over the various equations and just read the pull quotes. I've provided the details just for completeness.<blockquote>如果你不喜欢数学，可以随意浏览各种方程，只需阅读引号。为了完整起见，我提供了细节。</blockquote> The \"duration\" of a security can be defined in two ways.<blockquote>证券的“久期”可以通过两种方式定义。</blockquote> Investment horizon: the weighted average number of years it takes for the security to deliver its payments. For each period, you take the share of total present value represented by that year's payment, and multiply it by the number of years in the future the payment will be received. Add them all up. The result is the number of years from today (a weighted average) that the present value of your investment will be repaid. For example, the duration of a security that delivers a single payment a decade from now is simply 10 years.<blockquote>投资期限：证券支付款项所需的加权平均年数。对于每个期间，您将当年付款所代表的总现值份额乘以未来收到付款的年数。把它们加起来。结果是从今天起您的投资现值将得到偿还的年数（加权平均值）。例如，十年后一次性付款的证券的期限仅为10年。</blockquote> Elasticity: the percentage change in the security in response to a change in the underlying gross rate of return. Technically, elasticity is (-dP/P)/(dk/1+k). For example, suppose you own a security that will pay $100 a decade from now, and it's priced at $82.0348, for a 2% annual return. Now assume the expected return moves to 2.01%. The price would drop to $81.9544. Elasticity is (0.0804/82.0348)/(0.0001/1.02) = 10%<blockquote>弹性：证券响应基础总回报率变化的百分比变化。从技术上讲，弹性是(-dP/P)/(dk/1+k)。例如，假设您拥有一种证券，十年后将支付100美元，其定价为82.0348美元，年回报率为2%。现在假设预期回报率升至2.01%。价格将降至81.9544美元。弹性为(0.0804/82.0348)/(0.0001/1.02)=10%</blockquote> It turns out that the \"duration\" of a security in years is identical to its \"duration\" in terms of the percentage change of price in response to a 1% fluctuation in expected returns. Duration is also the holding period that \"immunizes\" the investor against changes in expected returns over time. In other words, assuming you reinvest your cash flows over time, duration gives you a good idea of how long you have to hold the security in order for your ending wealth to be largely independent of the fluctuations that the security experiences over that horizon.<blockquote>事实证明，就预期回报1%波动的价格变化百分比而言，证券以年为单位的“久期”与其“久期”相同。久期也是投资者对预期回报随时间变化“免疫”的持有期。换句话说，假设您随着时间的推移对现金流进行再投资，久期可以让您很好地了解您必须持有证券多长时间，以便您的最终财富在很大程度上独立于证券在该期间经历的波动。</blockquote> If you know differentiation, can prove to yourself that the \"modified duration\" of the S&amp;P 500 is essentially the inverse of the dividend yield. The modified duration is just (-dP/P)/dk or Macaulay duration/(1+k).<blockquote>如果你知道差异化，可以向自己证明标普500的“修正久期”本质上是股息收益率的倒数。修改后的持续时间只是（-dP/P）/dk或麦考利持续时间/（1+k）。</blockquote> Consider P = D/(k-g). Differentiating with respect to k, dP/dk = -P/(k-g), so duration (-dP/P)/dk = P/D.<blockquote>考虑P=D/（k-g）。相对于k微分，dP/dk=-P/（k-g），因此持续时间（-dP/P）/dk=P/D。</blockquote> Here's how to think about the link between valuations and duration. Presently, the dividend yield of the S&amp;P 500 is 1.48%. If the yield moves to 1.49%, holding dividends constant, prices drop by 1.48/1.49-1 = -0.67% on that 1 basis point move. Duration is just that sensitivity, defined for a 100 basis point move, which would be 67. It turns out that's also the weighted-average number of years from today that you'll receive your present value, if you invest today.<blockquote>以下是如何思考估值和久期之间的联系。目前，标普500的股息率为1.48%。如果收益率升至1.49%，保持股息不变，则价格在这1个基点的变动上下跌1.48/1.49-1=-0.67%。久期就是敏感度，定义为100个基点的变动，即67。事实证明，如果您今天投资，这也是从今天起您将获得现值的加权平均年数。</blockquote> Compare that to the typical situation over the past century, when the dividend yield of the S&amp;P 500 averaged about 3.7%. At normal valuations, a 1 basis point increase in the dividend yield would produce a price drop of 3.7/3.71-1 = 0.27%, implying a duration of 27 years.<blockquote>与过去一个世纪的典型情况相比，当时标普500的股息收益率平均约为3.7%。在正常估值下，股息收益率上升1个基点将导致价格下跌3.7/3.71-1=0.27%，这意味着持续时间为27年。</blockquote> It turns out that the 'duration' of a security in years is identical to its 'duration' in terms of the percentage change of price in response to a 1% fluctuation in expected returns. Duration is also the holding period that 'immunizes' the investor against changes in expected returns over time. Historically, investors wishing to match the duration of their investment portfolio to the duration of their investment horizon could be reasonably comfortable holding 100% of their assets in stocks, provided they had an investment horizon of about 25-30 years. Presently, these investors would need an investment horizon closer to 65-70 years. They are currently holding sippy cups.<blockquote>事实证明，就预期回报1%波动的价格变化百分比而言，以年为单位的证券的“久期”与其“久期”相同。久期也是使投资者免受预期回报随时间变化的影响的持有期。从历史上看，希望将投资组合的久期与投资期限相匹配的投资者可以相当放心地将100%的资产持有股票，前提是他们的投资期限约为25-30年。目前，这些投资者需要接近65-70年的投资期限。他们目前拿着吸管杯。</blockquote> Scarcity, usefulness, and value<blockquote>稀缺性、有用性和价值</blockquote> While we're on the subject of bubbles, I'll add a few comments on Bitcoin, just for fun. I'd write more, but my sides still hurt from laughing.<blockquote>当我们谈到泡沫的话题时，我会在比特币上补充一些评论，只是为了好玩。我会写更多，但我的身体仍然笑得很痛。</blockquote> Objects like tulip bulbs and Bitcoin differ from securities in that they do not deliver a stream of cash flows to the holder. Instead, what objects like tulips and currencies provide is a little stream of services over time, for example, as a perennial thing of beauty or as a means of payment. What people sometimes forget is that it is not just scarcity that defines the value of an object, but the stream of useful \"services\" that it provides (for some reason, nobody wants to buy my unique, limited edition, digitally-signed porcupine seat covers). The price of the object, and the stream of services it provides, should be commensurate.<blockquote>像郁金香球茎和比特币这样的物品与证券的不同之处在于，它们不会向持有者提供现金流。相反，像郁金香和货币这样的物品随着时间的推移提供了一点服务，例如，作为一种永恒的美丽事物或作为一种支付手段。人们有时会忘记的是，定义一件物品价值的不仅仅是稀缺，而是它提供的一系列有用的“服务”（出于某种原因，没有人想买我独特的限量版数字签名豪猪座套）。物品的价格和它提供的服务流应该是相称的。</blockquote> U.S. dollars, for example, have value primarily because they are tethered to the real economy by fiat (they legally must be accepted as a means of payment, as noted on the face of any dollar bill), and they represent the entire substrate of the banking system - nearly every payment that goes back and forth in the U.S. economy represents a transfer of base money. Base money (currency and bank reserves) provides billions of little \"services\" over time.<blockquote>例如，美元之所以有价值，主要是因为它们通过法定货币与实体经济联系在一起（正如任何美钞的正面所注明的那样，它们在法律上必须被接受为一种支付手段），并且它们代表了银行体系的整个基础——几乎在美国经济中来回的每一笔支付都代表着基础货币的转移。随着时间的推移，基础货币（货币和银行储备）提供了数十亿的小“服务”。</blockquote> With every transaction, reserves move electronically from bank to bank between one account holder and another. That combination of legal fiat and constant use as a substrate of the payments system is what gives money \"value.\" That value also means that the U.S. government essentially obtains revenue as \"seigniorage\" for producing the stuff. For those who imagine that governments are going to surrender that revenue in favor of using Bitcoin, I've got a non-fungible token to sell you.<blockquote>在每一笔交易中，准备金都会以电子方式在一个账户持有人和另一个账户持有人之间从一家银行转移到另一家银行。法定法令和作为支付系统基础的持续使用的结合赋予了货币“价值”。这一价值也意味着美国政府本质上通过生产这些东西获得了“铸币税”的收入。对于那些认为政府会放弃这些收入而使用比特币的人，我有一个不可替代的代币要卖给你。</blockquote> <img src=\"https://static.tigerbbs.com/747b2ba31a8d94d85d374d12b13b3c96\" tg-width=\"665\" tg-height=\"718\"><blockquote></blockquote> As I've noted before, blockchain is a brilliant algorithm, and I expect that it will have a great number of uses for secure transactions and inventory management. Bitcoin, however, is a token generated by an energy-inefficient, replicable blockchain app. Ultimately, its value rests on the capacity to provide transactions services, yet without fiat to require its use, and with strikingly narrow bandwidth - one block of roughly 2000 transactions every 10 minutes - that I expect will prove to be a wildly limiting feature. That's a problem in a world where speculators now value the stock of bitcoin at one-fifth the value of the entire U.S. monetary base.<blockquote>正如我之前提到的，区块链是一个出色的算法，我预计它将在安全交易和库存管理方面有大量用途。然而，比特币是一种由低能效、可复制的区块链应用程序生成的代币。最终，它的价值取决于提供交易服务的能力，但没有法定要求使用它，而且带宽非常窄——每10分钟大约有2000笔交易——我预计这将被证明是一个非常有限的功能。在投机者现在对比特币股票的估值是整个美国货币基础价值的五分之一的世界里，这是一个问题。</blockquote> If you think about how money is valued, it's clear that people accept it because they believe it will provide a claim on the future output of others. Of course, that expectation requires that future producers will also give away their output and accept the money, on the belief that yet other future producers will do the same. That expectation has to continue indefinitely. Like the question 'What holds up Atlas when Atlas holds up the world?' it's not enough to answer that he's standing on a turtle. It's got to be turtles all the way down. The value of money has an enormous psychological component.\" - John P. Hussman, Ph.D., Turtles All the Way Down, February 2019 Of course, Bitcoin may have a certain user base as a vehicle for money laundering and black market transactions, but that's an undesirable investment thesis. The vast majority of transactions are to exchange Bitcoin itself, though the New York Times did recently report that \"pornography, patio furniture, and an at-home coronavirus test are among the odd assortment of goods and services that people are purchasing with the cryptocurrency.\" So, basically, if your typical day consists of surfing porn on your patio while testing yourself for COVID, you're gonna want to look into Bitcoin.<blockquote>如果你想想货币是如何被估价的，很明显，人们接受它是因为他们相信它会提供对他人未来产出的要求。当然，这种期望要求未来的生产者也会放弃他们的产出并接受这笔钱，因为他们相信其他未来的生产者也会这样做。这种期望必须无限期地持续下去。就像这样的问题“当阿特拉斯支撑着世界时，是什么支撑着阿特拉斯？“仅仅回答他站在一只乌龟上是不够的。下面一定是海龟。金钱的价值具有巨大的心理成分。”–John P.Hussman，博士，海龟一路下跌，2019年2月当然，比特币可能拥有一定的用户群作为洗钱和黑市交易的工具，但这是一个不受欢迎的投资论点。绝大多数交易都是为了交换比特币本身，尽管《纽约时报》最近确实报道称，“色情、庭院家具和家庭冠状病毒测试都是人们用加密货币购买的各种奇怪的商品和服务。”所以，基本上，如果你典型的一天是在院子里浏览色情片，同时测试自己的COVID，你会想看看比特币。</blockquote> My largest concern is that people are actually forking over hard-earned savings in exchange for these tokens, which allows early \"miners\" to cash out. That's essentially the defining feature of a Ponzi scheme. Like all speculative bubbles that rely on increases in price, rather than cash flows generated by the production of value-added goods and services, Bitcoin isn't actually creating \"wealth.\" It's only creating the opportunity for wealth transfer, primarily from those who will end up holding the bag.<blockquote>我最担心的是，人们实际上是在用辛苦赚来的积蓄来换取这些代币，这使得早期的“矿工”可以套现。这本质上是庞氏骗局的定义特征。与所有依赖价格上涨而不是增值商品和服务生产产生的现金流的投机泡沫一样，比特币实际上并没有创造“财富”。这只是为财富转移创造了机会，主要是来自那些最终将掌管一切的人。</blockquote> Bitcoin has certain characteristics of base money in the sense that it's exchanged on an electronic ledger, but by design, transactions are limited to an average of about 2000 per block, with one block successfully validated, on average, every 10 minutes. In order to validate a transaction block, CPU farms across the world grind out terahashes of random SHA256 validation attempts in order to discover a sufficiently small binary that matches the cryptographic hash of the block.<blockquote>比特币具有基础货币的某些特征，因为它是在电子分类账上进行交易的，但根据设计，每个区块的交易平均限制在2000笔左右，平均每10分钟就有一个区块成功验证。为了验证一个事务块，世界各地的CPU场会研磨出随机SHA256验证尝试的万亿次哈希，以便发现与该块的加密哈希相匹配的足够小的二进制文件。</blockquote> All of this \"mining\" burns up about as much energy as it takes to run a modest-sized country. Validating a block of transactions produces a reward to the miner (and dilution of the coinbase) of 6.25 Bitcoin per block, which currently works out to nearly $200 per transaction. Yet the value of the median transaction in Bitcoin is only about $1000 in the first place.<blockquote>所有这些“采矿”消耗的能量大约相当于运行一个中等规模的国家所需的能量。验证一笔交易会给矿工带来每笔6.25比特币的奖励（并稀释coinbase），目前每笔交易的奖励接近200美元。然而，比特币的交易价值中值一开始只有1000美元左右。</blockquote> There's a rather primitive regression analysis floating around (tagged as \"sophisticated\" by some observers who apparently go numb at the word \"logarithm\") that attempts to relate the log price of bitcoin to the log \"stock/flow\" ratio, as if it represents some mechanistic supply-demand relationship. Aside from the fact that the correlation between two diagonal lines is always about 0.9-something, I find that one can obtain a better fit just by regressing the log price of Bitcoin on the log ratio of block difficulty/block reward, which is basically a measure of how much energy one needs to waste in order to mine a new bitcoin.<blockquote>有一种相当原始的回归分析（被一些显然对“对数”这个词麻木的观察者标记为“复杂”），试图将比特币的对数价格与对数“库存/流量”比率联系起来，好像它代表了某种机械的供需关系。除了两条对角线之间的相关性总是在0.9左右这一事实之外，我发现，只要将比特币的对数价格回归到区块难度/区块奖励的对数比率，就可以获得更好的拟合，这基本上是一种衡量一个人为了挖掘新的比特币需要浪费多少能量的方法。</blockquote> So the \"value\" of Bitcoin is partially linked to the backward-looking sunk cost of the energy wasted to mine these tokens. Still, I wouldn't dream of using this sort of \"model\" to trade an object whose \"value\" is primarily in the heads of speculators. Use it if you like. If you happen make money on it, feel free send me a check, preferably in U.S. dollars.<blockquote>因此，比特币的“价值”部分与开采这些代币所浪费的能源的向后看沉没成本有关。尽管如此，我不会梦想用这种“模型”来交易一个“价值”主要在投机者头脑中的对象。喜欢就用吧。如果你碰巧在上面赚钱，请随时给我寄一张支票，最好是美元。</blockquote> Undoubtedly, this view of Bitcoin will be unpopular among those who associate holding Bitcoin with superpowers like laser eyes and diamond hands. \"Not surprised Hussman doesn't get Bitcoin. Few do.\" M'kay. Look, there's certainly a case to be made that a speculative mindset creates its own reality, and while it does, there's an opportunity to obtain wealth transfers from frantic late-comers who can no longer tolerate missing out. Tulips gonna tulip. Not my gig, thanks.<blockquote>毫无疑问，这种对比特币的看法将在那些将持有比特币与激光眼和钻石手这样的超能力联系在一起的人中间不受欢迎。“赫斯曼不了解比特币并不奇怪。很少有人了解。”好吧。看，投机心态创造了自己的现实，当然有理由证明，虽然它创造了自己的现实，但也有机会从疯狂的后来者那里获得财富转移，他们再也不能容忍错过机会。郁金香会郁金香。不是我的工作，谢谢。</blockquote> In the short run, it will be said to be an attack, motivated by either deficient understanding or uncontrolled envy, of the wonderful process of enrichment. Those involved with the speculation are experiencing an increase in wealth - getting rich or being further enriched. No one wishes to believe that this is fortuitous or undeserved; all wish to think that it is the result of their own superior insight or intuition. As long as they are in, they have a strong pecuniary commitment to belief in the unique personal intelligence that tells them there will be yet more. Accordingly, possession must be associated with some special genius. Speculation buys up, in a very practical way, the intelligence of those involved. Only after the speculative collapse does the truth emerge. What was thought to be unusual acuity turns out to be only a fortuitous and unfortunate association with the assets. - John Kenneth Galbraith, A Brief History of Financial Euphoria <blockquote>从短期来看，人们会说这是一种攻击，其动机要么是缺乏理解，要么是无法控制的嫉妒，对丰富的奇妙过程。那些参与投机的人正在经历财富的增加——变得富有或进一步变得富有。没有人愿意相信这是偶然的或不应得的；所有人都希望认为这是他们自己卓越的洞察力或直觉的结果。只要他们在，他们就有一个强大的金钱承诺，相信独特的个人智慧告诉他们还会有更多。因此，附身必须与某种特殊的天才联系在一起。投机以一种非常实际的方式收买了相关人员的智力。只有在投机崩溃之后，真相才会浮出水面。被认为是不寻常的敏锐度被证明只是与资产的偶然和不幸的联系。——约翰·肯尼思·加尔布雷斯，《金融欣快简史》</blockquote>","source":"lsy1615871319183","collect":0,"html":"<!DOCTYPE html>\n<html>\n<head>\n<meta http-equiv=\"Content-Type\" content=\"text/html; charset=utf-8\" />\n<meta name=\"viewport\" content=\"width=device-width,initial-scale=1.0,minimum-scale=1.0,maximum-scale=1.0,user-scalable=no\"/>\n<meta name=\"format-detection\" content=\"telephone=no,email=no,address=no\" />\n<title>How To Spot A Bubble<blockquote>如何发现气泡</blockquote></title>\n<style type=\"text/css\">\na,abbr,acronym,address,applet,article,aside,audio,b,big,blockquote,body,canvas,caption,center,cite,code,dd,del,details,dfn,div,dl,dt,\nem,embed,fieldset,figcaption,figure,footer,form,h1,h2,h3,h4,h5,h6,header,hgroup,html,i,iframe,img,ins,kbd,label,legend,li,mark,menu,nav,\nobject,ol,output,p,pre,q,ruby,s,samp,section,small,span,strike,strong,sub,summary,sup,table,tbody,td,tfoot,th,thead,time,tr,tt,u,ul,var,video{ font:inherit;margin:0;padding:0;vertical-align:baseline;border:0 }\nbody{ font-size:16px; line-height:1.5; color:#999; background:transparent; }\n.wrapper{ overflow:hidden;word-break:break-all;padding:10px; }\nh1,h2{ font-weight:normal; line-height:1.35; margin-bottom:.6em; }\nh3,h4,h5,h6{ line-height:1.35; margin-bottom:1em; }\nh1{ font-size:24px; }\nh2{ font-size:20px; }\nh3{ font-size:18px; }\nh4{ font-size:16px; }\nh5{ font-size:14px; }\nh6{ font-size:12px; }\np,ul,ol,blockquote,dl,table{ margin:1.2em 0; }\nul,ol{ margin-left:2em; }\nul{ list-style:disc; }\nol{ list-style:decimal; }\nli,li p{ margin:10px 0;}\nimg{ max-width:100%;display:block;margin:0 auto 1em; }\nblockquote{ color:#B5B2B1; border-left:3px solid #aaa; padding:1em; }\nstrong,b{font-weight:bold;}\nem,i{font-style:italic;}\ntable{ width:100%;border-collapse:collapse;border-spacing:1px;margin:1em 0;font-size:.9em; }\nth,td{ padding:5px;text-align:left;border:1px solid #aaa; }\nth{ font-weight:bold;background:#5d5d5d; }\n.symbol-link{font-weight:bold;}\n/* header{ border-bottom:1px solid #494756; } */\n.title{ margin:0 0 8px;line-height:1.3;color:#ddd; }\n.meta {color:#5e5c6d;font-size:13px;margin:0 0 .5em; }\na{text-decoration:none; color:#2a4b87;}\n.meta .head { display: inline-block; overflow: hidden}\n.head .h-thumb { width: 30px; height: 30px; margin: 0; padding: 0; border-radius: 50%; float: left;}\n.head .h-content { margin: 0; padding: 0 0 0 9px; float: left;}\n.head .h-name {font-size: 13px; color: #eee; margin: 0;}\n.head .h-time {font-size: 12.5px; color: #7E829C; margin: 0;}\n.small {font-size: 12.5px; display: inline-block; transform: scale(0.9); -webkit-transform: scale(0.9); transform-origin: left; -webkit-transform-origin: left;}\n.smaller {font-size: 12.5px; display: inline-block; transform: scale(0.8); -webkit-transform: scale(0.8); transform-origin: left; -webkit-transform-origin: left;}\n.bt-text {font-size: 12px;margin: 1.5em 0 0 0}\n.bt-text p {margin: 0}\n</style>\n</head>\n<body>\n<div class=\"wrapper\">\n<header>\n<h2 class=\"title\">\nHow To Spot A Bubble<blockquote>如何发现气泡</blockquote>\n</h2>\n<h4 class=\"meta\">\n\nHussman Funds2021-03-16 13:08\n\n</h4>\n</header>\n<article>\nSummary<blockquote>总结</blockquote> <ul> <li>The defining feature of a bubble is inconsistency between expected returns based on price behavior and expected returns based on valuations.</li> <li>If we compare our most reliable valuation measures with the valuation measures that one would obtain from a proper long-term discounted cash flow analysis, we find that they're nearly identical.</li> <li>One of the unfortunate bits of financial illiteracy that Wall Street has pushed into the heads of investors is the idea that extreme valuations are \"justified\" by low interest rates.</li> <li>It's undoubtedly true that profit margins, expected growth, and other factors have an effect on future deliverable cash flows and the valuations that investors place on stocks.</li> <li>To understand why extreme valuations imply high volatility, and require extremely long investment horizons, it's useful to consider the concept of \"duration.\"</li> </ul> I can show, really precisely, that there are two warranted prices for a share. The one I prefer is based on such fundamentals as earnings and growth rates, but the bubble is rational in a certain sense. The expectation of growth produces the growth, which confirms the expectation; people will buy it because it went up. But once you are convinced that it is not growing anymore, nobody wants to hold a stock because it is overvalued. Everybody wants to get out and it collapses, beyond the fundamentals.- Nobel Laureate Franco Modigliani, New York Times, March 30, 2000 The word \"bubble\" is tossed around quite a bit in the financial markets, but it's rarely used correctly. See, the thing that defines a bubble isn't that valuations are extremely high, or that expected returns are extremely low. Instead, what defines a bubble is that investors drive valuations higher without simultaneously adjusting expectations for returns lower. That is, investors extrapolate past returns based on price behavior, even though those expectations are inconsistent with the returns that would equate price with discounted cash flows.<blockquote><ul><li>泡沫的定义特征是基于价格行为的预期回报和基于估值的预期回报之间的不一致。</li><li>如果我们将最可靠的估值指标与从适当的长期贴现现金流分析中获得的估值指标进行比较，我们会发现它们几乎相同。</li><li>华尔街向投资者灌输的一个不幸的金融文盲是，低利率“证明”极端估值是合理的。</li><li>毫无疑问，利润率、预期增长和其他因素会影响未来可交付的现金流和投资者对股票的估值。</li><li>为了理解为什么极端估值意味着高波动性，并且需要极长的投资期限，考虑“久期”的概念是有用的。</li></ul>我可以非常精确地证明，一股有两个保证价格。我更喜欢的是基于盈利和增长率等基本面，但泡沫在某种意义上是理性的。对增长的预期产生了增长，增长证实了预期；人们会买它，因为它上涨了。但一旦你确信它不再增长，就没有人愿意持有一只股票，因为它被高估了。每个人都想退出，但它崩溃了，超出了基本面。——诺贝尔奖获得者佛朗哥·莫迪里阿尼，《纽约时报》，2000年3月30日“泡沫”一词在金融市场上经常被提及，但很少被正确使用。看，定义泡沫的不是估值极高，也不是预期回报极低。相反，泡沫的定义是投资者推高估值，而不同时降低回报预期。也就是说，投资者根据价格行为推断过去的回报，即使这些预期与将价格等同于贴现现金流的回报不一致。</blockquote> In March 2000, at the height of the technology bubble, I noted: \"Over time, price/revenue ratios come back in line. Currently, that would imply an 83% plunge in tech stocks. If you understand values and market history, you know we're not joking.\" The following month, I discussed Modigliani's quote above, and detailed the dynamics he was describing. The collapse of the 2000 bubble would ultimately erase half the value of the S&amp;P 500, and would take the tech-heavy Nasdaq 100 down an implausibly precise 83%.<blockquote>2000年3月，在科技泡沫最严重的时候，我指出：“随着时间的推移，市盈率会恢复正常。目前，这意味着科技股将暴跌83%。如果你了解价值观和市场历史，你就会知道我们不是在开玩笑。”接下来的一个月，我讨论了莫迪里阿尼在上面的引用，并详细描述了他所描述的动态。2000年泡沫的破裂最终将抹去标普500一半的价值，并使以科技股为主的纳斯达克100指数下跌83%，令人难以置信。</blockquote> The defining feature of a bubble is inconsistency between expected returns based on price behavior and expected returns based on valuations. If investors pay $150 today for a security that will deliver a single $100 payment a decade from now, but they also fully understand that they'll lose 4% annually on the deal, without extrapolating past gains into the future, then we might say the security is overvalued, and we might question why investors would accept that trade, but we can't call it a bubble.<blockquote>泡沫的定义特征是基于价格行为的预期回报和基于估值的预期回报之间的不一致。如果投资者今天支付150美元购买一种证券，而十年后只需支付100美元，但他们也完全明白，在不将过去的收益外推到未来的情况下，他们每年将损失4%，那么我们可能会说证券被高估了，我们可能会质疑为什么投资者会接受这种交易，但我们不能看涨期权它是泡沫。</blockquote> But if investors pay $150 today for that security, because they look back in the rear-view mirror, decide that it \"always goes up\" over time, and convince themselves that expected future returns are always positive, then you've got a bubble. Discounting the future $100 cash flow of the security using any positive expected return would produce a price less than $100. So the positive returns expected by investors are inconsistent with the returns that would equate price with discounted cash flows. The size of the bubble is the fraction of the market price that represents expectational \"hot air.\"<blockquote>但如果投资者今天为该证券支付150美元，因为他们回头看后视镜，认为它随着时间的推移“总是上涨”，并说服自己预期的未来回报总是正的，那么你就有了泡沫。使用任何正预期回报贴现证券未来100美元的现金流将产生低于100美元的价格。因此，投资者预期的正回报与将价格等同于贴现现金流的回报不一致。泡沫的大小是代表预期“热空气”的市场价格的分数。</blockquote> Likewise, the willingness of investors to embrace \"passive investments\" like ETFs and asset-backed securities based on past performance, with little concern about the valuations, yields, or credit risk of the securities inside, is the very soap from which bubbles repeatedly emerge. Amid the current enthusiasm for special purpose acquisition companies (SPACS), investors might recall the bubble in \"incubators\" at the 2000 peak, the \"conglomerates\" of the late-1960s Go-Go bubble, and even the South Sea Company in the early 1700s, along with similar companies formed at the time \"for carrying on an undertaking of great advantage, but nobody to know what it is.\"<blockquote>同样，投资者愿意接受基于过去表现的ETF和资产支持证券等“被动投资”，而很少关心其中证券的估值、收益率或信用风险，这正是泡沫反复出现的肥皂。在当前对特殊目的收购公司(SPAC)的热情中，投资者可能会回忆起2000年高峰期“孵化器”的泡沫、20世纪60年代末Go-Go泡沫的“企业集团”，甚至是20世纪初的南海公司。1700年代，以及当时成立的类似公司“是为了开展一项具有巨大优势的事业，但没有人知道它是什么”。</blockquote> If investors price the S&amp;P 500 at levels that are highly likely to produce negative returns for a decade, as they did in 1929 and 2000, and as I believe they are doing at present, yet investors continue to press stock prices higher on the expectation that they will provide historically normal levels of future return regardless of valuations, then you have the sort of inconsistency that defines a bubble.<blockquote>如果投资者将标普500定价在十年内极有可能产生负回报的水平，就像他们在1929年和2000年所做的那样，而且我相信他们目前正在做的那样，但投资者继续推高股价，因为他们期望无论估值如何，他们都将提供历史上正常的未来回报水平，那么就存在定义泡沫的不一致性。</blockquote> Likewise, if the expected return of a conventional passive investment mix is negative on a 10-12 year horizon (based on reliable valuation measures strongly correlated with actual subsequent returns over a century of market history), yet pension return assumptions remain locked near 7% annually, you've got a bubble, and most likely a future pension funding crisis, on your hands.<blockquote>同样，如果传统被动投资组合在10-12年内的预期回报为负（基于与一个世纪市场历史中实际后续回报密切相关的可靠估值指标），但养老金回报假设仍锁定在7%附近每年，你就面临着泡沫，很可能是未来的养老金融资危机。</blockquote> This is how very bad things have repeatedly happened in the financial markets across history, enabled by what Galbraith called \"the extreme brevity of the financial memory.\"<blockquote>这就是历史上金融市场上反复发生的非常糟糕的事情，加尔布雷斯所说的“金融记忆的极度短暂”促成了这一点。</blockquote> When Modigliani says that there are two \"warranted\" prices, he means that - at least in the short run - there are two ways that prices can fulfill the expectations of investors. In one case, investors have expectations about future returns, and those expectations are informed by the level of valuations. If prices rise, and expected cash flows haven't changed, investors recognize that future returns will be lower. In the \"bubble\" case, investors have high expectations about future returns, mainly based on past returns, and they act on those expectations by driving prices up further. So the expectation of additional price increases is simply reinforced. \"The expectation of growth produces the growth, which confirms the expectation.\"<blockquote>当莫迪里阿尼说有两种“合理”的价格时，他的意思是——至少在短期内——价格有两种方式可以满足投资者的预期。在一种情况下，投资者对未来回报有预期，而这些预期是由估值水平决定的。如果价格上涨，而预期现金流没有改变，投资者就会认识到未来的回报会更低。在“泡沫”案例中，投资者对未来回报有很高的预期，主要基于过去的回报，他们通过进一步推动价格上涨来实现这些预期。因此，进一步涨价的预期只是被强化了。“增长的预期产生了增长，增长证实了预期。”</blockquote> Only one of these prices is consistent, in that the rate of return expected by investors is also the rate of return that equates price with discounted future cash flows. The other price becomes increasingly detached from fundamentals, as a larger and larger fraction of the price represents hot air, and it ultimately collapses as the gap becomes untenably wide.<blockquote>这些价格中只有一个是一致的，因为投资者预期的回报率也是将价格等同于贴现的未来现金流的回报率。另一个价格变得越来越脱离基本面，因为价格中越来越大的部分代表着空话，随着差距变得难以维持，它最终会崩溃。</blockquote> During speculative segments of the market cycle, there's nothing that forces investors to recognize that higher valuations imply lower returns, or to change their expectation of high returns as far as the eye can see. That, of course, is why we use measures of market internals to gauge the inclination of investors toward speculation or risk-aversion. Valuation provides an enormous amount of information about likely long-term returns and potential market losses over the complete cycle. But valuation isn't a timing tool. In recent years, it hasn't even imposed a limit on speculative recklessness.<blockquote>在市场周期的投机阶段，没有什么能迫使投资者认识到更高的估值意味着更低的回报，或者改变他们对高回报的预期。当然，这就是为什么我们使用市场内部因素来衡量投资者的投机或避险倾向。估值提供了大量关于整个周期中可能的长期回报和潜在市场损失的信息。但估值不是择时工具。近年来，它甚至没有对投机鲁莽行为施加限制。</blockquote> Still, with each price advance, the actual long-term return implied by future cash flows - what investors will ultimately realize as those cash flows are delivered - collapses further, even while investors act on their delusion that long-term returns have nothing to do with price. Eventually, the bulk of the security price represents a bubble component, not the price that would actually need to exist in order for the long-term expectations of investors to be accurate.<blockquote>尽管如此，随着价格的每一次上涨，未来现金流隐含的实际长期回报——投资者最终将在这些现金流交付时意识到的回报——进一步崩溃，即使投资者按照长期回报与价格无关的错觉行事。最终，证券价格的大部分代表了泡沫成分，而不是为了使投资者的长期预期准确而实际需要存在的价格。</blockquote> As I wrote at the 2000 market peak:<blockquote>正如我在2000年市场高峰时所写的：</blockquote> \"As long as investors focus on year-to-year returns and not discounted cash flow calculations, the bubble can continue to grow in self-reinforcing fashion. Investors anticipate a high return, and the price behavior reinforces the expectation. The true long-term return becomes increasingly detached from the long-term return imagined by investors, and the bubble component accounts for an increasingly large proportion of the total price.\" By our most reliable measures, run-of-the-mill historical valuation norms are roughly 70% below current levels. I know you don't want to believe that.<blockquote>“只要投资者关注的是同比回报，而不是贴现现金流计算，泡沫就能以自我强化的方式持续增长。投资者预期高回报，价格行为强化了预期。真正的长期回报与投资者想象的长期回报越来越脱离，泡沫成分在总价中的占比越来越大。”根据我们最可靠的衡量标准，一般的历史估值标准比当前水平低约70%。我知道你不愿意相信。</blockquote> <img src=\"https://static.tigerbbs.com/493c338076df85e2a04ebf892af4762f\" tg-width=\"1280\" tg-height=\"713\"><blockquote></blockquote> The trap door quietly swings open when valuations are extreme and market internals begin to deteriorate. That's the situation we've observed in our measures in recent weeks, with the initial deterioration largely driven by debt securities, but with increasing divergences in equities as well.<blockquote>当估值极端且市场内部开始恶化时，活板门就会悄然打开。这就是我们最近几周在测量中观察到的情况，最初的恶化主要是由债务证券推动的，但股票的分歧也在扩大。</blockquote> Valuations and discounted cash flows<blockquote>估值及贴现现金流量</blockquote> Valuations measure the tradeoff between current prices and a very long-term stream of expected future cash flows. Every useful valuation ratio is just shorthand for that calculation. Every valuation ratio that fails that criterion is inferior, and you can show it in historical data. - John P. Hussman, Ph.D., The Meaning of Valuation, December 2019 I've often noted that the denominator of every good valuation measure is just shorthand for the decades and decades of cash flows that the security is likely to deliver in the future. In fact, we always test the validity of the valuation measures we use by examining:<blockquote>估值衡量当前价格与长期预期未来现金流之间的权衡。每一个有用的估值比率都只是该计算的简写。每一个不符合这个标准的估值比率都是劣质的，你可以在历史数据中显示出来。–John P.Hussman博士，《估值的意义》，2019年12月我经常注意到，每一个好的估值衡量标准的分母都只是证券未来可能提供的数十年现金流的简写。事实上，我们总是通过检查来测试我们使用的估值方法的有效性：</blockquote> a) how strongly the resulting valuation measures are correlated with actual subsequent total returns, and;<blockquote>a)由此产生的估值措施与实际后续总回报的相关性有多强；</blockquote> b) how closely they replicate an explicit discounted cash flow analysis.<blockquote>b)它们在多大程度上复制了明确的贴现现金流分析。</blockquote> Below, we'll examine a variety of valuation measures that offer some perspective on why I view the U.S. equity market as a bubble near the breaking point. Along the way, I'll point out some interesting features of valuations and their relationship with subsequent returns. If math gives you hives, feel free to skim over the small amount that I've included here and there.<blockquote>下面，我们将研究各种估值指标，这些指标可以解释为什么我认为美国股市是一个接近崩溃点的泡沫。在此过程中，我将指出估值的一些有趣特征及其与后续回报的关系。如果数学让你荨麻疹，请随意浏览我在这里和那里包含的少量内容。</blockquote> Consider first the relationship between valuations and subsequent returns. I'll state the following, which you can prove to yourself by toying around a bit with present value models: the logarithm of a good valuation measure should have an inverse and roughly linear relationship with the expected subsequent investment return.<blockquote>首先考虑估值和后续回报之间的关系。我将陈述以下内容，你可以通过玩弄现值模型来向自己证明这一点：一个好的估值衡量标准的对数应该与预期的后续投资回报呈反比且大致呈线性关系。</blockquote> Here's a simple example of what this looks like.<blockquote>这是一个简单的例子。</blockquote> <img src=\"https://static.tigerbbs.com/d1e5bdf4391c9c91a8b9a051f6635116\" tg-width=\"819\" tg-height=\"708\"><blockquote></blockquote> Here's what this looks like for MarketCap/GVA, our most reliable stock market valuation measure<blockquote>这是我们最可靠的股市估值指标MarketCap/GVA的情况</blockquote> <img src=\"https://static.tigerbbs.com/485141919c853722522159d2e9d05e85\" tg-width=\"1280\" tg-height=\"720\"><blockquote></blockquote> During the past three decades, we've studied and introduced a broad range of valuation measures. Most can be calculated back to 1947. Some can be evaluated over a century or more. Across history, even in recent decades, we find that the valuation measures that are best correlated with actual subsequent returns are those with muted sensitivity to cyclical fluctuations in profit margins, and that behave largely like broad, market-wide price/revenue ratios.<blockquote>在过去的三十年里，我们研究并引入了广泛的估值方法。大多数可以追溯到1947年。有些可以评估一个世纪或更长时间。纵观历史，甚至在最近几十年，我们发现与实际后续回报最相关的估值指标是那些对利润率周期性波动敏感性较低的估值指标，并且在很大程度上类似于广泛的全市场市盈率。</blockquote> If we compare our most reliable valuation measures with the valuation measures that one would obtain from a proper long-term discounted cash flow analysis, we find that they're nearly identical. Here's what this comparison looks like for the actual stream of dividends (including the impact of repurchases) delivered by the S&amp;P 500 since 1900, discounted at a fixed 10% rate (see the chart text for additional details).<blockquote>如果我们将最可靠的估值指标与从适当的长期贴现现金流分析中获得的估值指标进行比较，我们会发现它们几乎相同。以下是标普500自1900年以来实际股息流（包括回购的影响）的比较，按10%的固定利率贴现（有关更多详细信息，请参阅图表文本）。</blockquote> The reason we use a fixed rate of return is that a multiple of 1.0 is then, by definition, the level at which the S&amp;P 500 would have been priced for that particular level of expected return. Any deviation in the valuation multiple from 1.0 then gauges how far likely future returns are from that \"typical\" expected return. We're currently farther away from \"typical\" expected returns than at any moment in history, including the 1929 and 2000 market peaks.<blockquote>我们使用固定回报率的原因是，根据定义，1.0的倍数就是该特定预期回报水平的标普500定价水平。估值倍数与1.0的任何偏差都可以衡量未来回报与“典型”预期回报的可能性有多大。我们目前比历史上任何时候都更远离“典型”预期回报，包括1929年和2000年的市场峰值。</blockquote> <img src=\"https://static.tigerbbs.com/e6f61d2794095c2545d0d2c130320e45\" tg-width=\"1280\" tg-height=\"722\"><blockquote></blockquote> One of the unfortunate bits of financial illiteracy that Wall Street has pushed into the heads of investors is the idea that extreme valuations are \"justified\" by low interest rates. Now, it's certainly true that holding future cash flows constant, raising the price of an investment will lower the embedded rate of return, and vice versa. If you pay $32 today for $100 a decade from today, you can expect a 12% annual return. If you pay $82 for the same security, you can expect a 2% annual return. If you pay $100 today, you can expect nothing. So it's clearly true that holding future cash flows constant, a lower rate of return implies a higher level of valuation.<blockquote>华尔街向投资者灌输的一个不幸的金融文盲是，低利率“证明”极端估值是合理的。现在，保持未来现金流不变，提高投资价格会降低内含回报率，反之亦然，这是事实。如果你今天支付32美元，十年后支付100美元，你可以期待12%的年回报率。如果你为同样的证券支付82美元，你可以期待2%的年回报率。如果你今天付100美元，你什么也不能指望。因此，很明显，保持未来现金流不变，较低的回报率意味着较高的估值水平。</blockquote> The reason the statement \"low interest rates justify high valuations\" contributes to financial illiteracy is that the statement has been learned entirely out of context of the arithmetic. As a result, investors seem to imagine that, as long as high valuations can be \"justified,\" stocks can be expected to provide historically normal rates of return in the future. Likewise, investors seem to have no concept that if interest rates are low because growth rates are low, no valuation premium is \"justified\" for stocks, because the lower growth is already sufficient to bring future stock returns down to levels that are commensurate with the low level of interest rates.<blockquote>“低利率证明高估值是合理的”这句话导致金融文盲的原因是，这句话完全是在算术背景下学习的。因此，投资者似乎想象，只要高估值能够“合理”，股票就有望在未来提供历史上正常的回报率。同样，投资者似乎没有概念，如果利率低是因为增长率低，那么股票的估值溢价就“合理”，因为较低的增长已经足以将未来的股票回报降至与低利率水平相称的水平。</blockquote> The truth is simple but uncomfortable. If interest rates are low and expected growth is held constant, higher valuations imply lower long-term returns. If interest rates are low because expected future growth is also low, higher valuations are not required. Long-term returns will be lower anyway. A valuation premium just makes future returns even worse.<blockquote>道理很简单，却让人难受。如果利率较低且预期增长保持不变，较高的估值意味着较低的长期回报。如果利率较低，因为预期的未来增长也较低，则不需要更高的估值。无论如何，长期回报都会更低。估值溢价只会让未来的回报变得更糟。</blockquote> Saying that extremely low interest rates \"justify\" extremely high stock valuations is identical to saying that extremely low future returns on bonds \"justify\" extremely low future returns on stocks. I don't really think that's something Wall Street cares to clarify when it tells investors that stock market valuations are \"justified.\"<blockquote>说极低的利率“证明”极高的股票估值是合理的，就像说极低的债券未来回报率“证明”极低的股票未来回报率是一样的。当华尔街告诉投资者股市估值“合理”时，我真的不认为这是它想澄清的事情。</blockquote> Investment valuation is concerned with the relationship between three objects: the current price, the future cash flows, and the rate of return that connects the two like a string. The lower the current price and the higher the future cash flows, the steeper the string. The higher the current price and the lower the future cash flows, the flatter the string. Raise the current price above the future cash flows, and the string will point down instead of up. Given any two of these objects, you can calculate the third one.<blockquote>投资估值关注的是三个对象之间的关系：当前价格、未来现金流以及像绳子一样将两者连接起来的回报率。当前价格越低，未来现金流越高，弦越陡。当前价格越高，未来现金流越低，字符串就越扁平。将当前价格提高到未来现金流之上，字符串将指向下方而不是上方。给定这些对象中的任意两个，你可以计算第三个。</blockquote> For example, if you want to estimate expected long-term returns, you need two objects: a) the current price and b) the expected stream of future cash flows. A good valuation ratio is just shorthand for those two objects, so you can estimate the long-term return directly from the level of valuation. Then, if you like, you can compare it with the level of interest rates. That comparison can be useful, because even when investors realize that high stock market valuations imply low long-term returns, it's not at all clear that they realize how low long-term return prospects have been driven.<blockquote>例如，如果你想估计预期的长期回报，你需要两个对象：a）当前价格和b）未来现金流的预期流。一个好的估值比率只是那两个对象的速记，所以你可以直接从估值水平来估计长期回报。然后，如果你愿意，你可以把它和利率的水平进行比较。这种比较可能是有用的，因为即使投资者意识到高股市估值意味着低长期回报，他们也根本不清楚他们是否意识到长期回报前景有多低。</blockquote> The chart below shows our estimate of expected 12-year S&amp;P 500 total returns over-and-above Treasury bond yields, across a century of market history. Compare this with the nearly useless drivel that Wall Street passes off as the \"equity risk premium\" (typically quoted as the S&amp;P 500 forward earnings yield minus the 10-year Treasury yield). Yes, you're reading the chart correctly. Given current valuations, we expect the total return of the S&amp;P 500 to underperform the lowly yield on Treasury bonds by roughly -6% annually over the coming 12-year period.<blockquote>下图显示了我们对一个世纪的市场历史中预期12年期标普500总回报率高于国债收益率的估计。相比之下，华尔街称之为“股票风险溢价”（通常被引用为标普500远期收益收益率减去10年期国债收益率）的几乎毫无用处的废话。是的，你没看错图表。鉴于目前的估值，我们预计未来12年标普500的总回报率将低于国债的低收益率约-6%。</blockquote> <img src=\"https://static.tigerbbs.com/42561963df9d259a93fe0d4f2502c13e\" tg-width=\"1024\" tg-height=\"577\"><blockquote></blockquote> Many investors confuse the estimation of expected returns (which requires only expected cash flows and the observed price) with a different problem - the estimation of \"fair value.\" See, interest rates come into the picture when you have an expected stream of future cash flows and you want to calculate a \"fair\" current price. In that case, rather than picking an arbitrary rate of return from a hat, it's common to use the level of interest rates, plus some risk premium, as the expected long-term return (or discount rate, or capitalization rate). This sort of calculation can be super-sensitive to arbitrary choices.<blockquote>许多投资者将预期回报的估计（只需要预期现金流和观察到的价格）与另一个问题——“公允价值”的估计混为一谈。看，当你有一个预期的未来现金流，你想要计算一个“公平”的当前价格时，利率就会出现。在这种情况下，通常使用利率水平加上一些风险溢价作为预期的长期回报（或贴现率或资本化率），而不是从帽子中选择任意的回报率。这种计算对任意选择非常敏感。</blockquote> In particular, Wall Street loves to combine super-high growth rates, super-low discount rates, and super-long time horizons, which allows one to calculate a \"fair value\" that's as close to infinity as possible. The thing to remember is that whatever rate of return an analyst embeds into the fair value calculation is also the long-term rate of return you'll earn over time if you pay that price, assuming the future cash flows are delivered as expected.<blockquote>特别是，华尔街喜欢将超高的增长率、超低的贴现率和超长的时间范围结合起来，这使得人们可以计算出尽可能接近无穷大的“公允价值”。需要记住的是，无论分析师在公允价值计算中嵌入什么回报率，假设未来现金流按预期交付，如果您支付该价格，随着时间的推移您将获得的长期回报率。</blockquote> Among scores of measures we've evaluated or introduced over time, MarketCap/GVA (nonfinancial market capitalization to corporate gross value-added, including our estimate of foreign revenues) has the highest correlation with actual subsequent 10-12 year S&amp;P 500 total returns, followed by our Margin-Adjusted P/E (MAPE). I find it hilarious that the various valuation measures I've introduced over time are sometimes described as products of \"machine learning,\" \"data mining,\" and \"curve fitting\" when they are, in fact, just different versions of an apples-to-apples economy-wide price/revenue ratio.<blockquote>在我们随着时间的推移评估或引入的数十项指标中，MarketCap/GVA（非金融市值与企业总增加值之比，包括我们对国外收入的估计）与随后10-12年标普500实际总回报的相关性最高，其次是我们的利润率调整市盈率(MAPE)。我觉得好笑的是，随着时间的推移，我引入的各种估值指标有时被描述为“机器学习”、“数据挖掘”和“曲线拟合”的产物，而事实上，它们只是整个经济范围内市盈率的不同版本。</blockquote> <img src=\"https://static.tigerbbs.com/75dac426a3ad2f3830f9dce50f7c0a1c\" tg-width=\"1265\" tg-height=\"712\"><blockquote></blockquote> The S&amp;P 500 price/revenue ratio and nonfinancial market capitalization to GDP (the \"Buffett indicator\") also perform well, and better than earnings-based alternatives like price/forward earnings, the Fed Model, and even the Shiller CAPE. See, while earnings are necessary to generate long-term cash flows, they are also subject to fluctuations in profit margins (both cyclically and even from decade to decade) that turn out to be highly uninformative.<blockquote>标普500市盈率和非金融市值与GDP之比（“巴菲特指标”）也表现良好，优于基于收益的替代指标，如价格/远期收益、美联储模型，甚至席勒CAPE。看，虽然收益对于产生长期现金流是必要的，但它们也会受到利润率波动的影响（周期性地，甚至每十年一次），而事实证明，这些波动是非常缺乏信息的。</blockquote> Economically, fluctuations in profit margins are driven primarily by fluctuations in real unit labor costs. Because companies compete on the basis of realized after-tax profits rather than pre-tax profits, changes in tax policy also have far less durable impact on corporate profits than investors seem to imagine. While corporate profits got a tremendous boost last year from CARES spending (the deficit of one sector always emerges as the surplus of another), here's what the relationship between corporate profits and real unit labor costs looks like historically.<blockquote>在经济上，利润率的波动主要是由实际单位劳动力成本的波动驱动的。由于企业是基于已实现的税后利润而非税前利润进行竞争，税收政策的变化对企业利润的持久影响也远不如投资者似乎想象的那样。虽然去年企业利润从医疗保健支出中获得了巨大的提振（一个部门的赤字总是表现为另一个部门的盈余），但以下是企业利润和实际单位劳动力成本之间的历史关系。</blockquote> Real unit labor costs are presented on an inverted left scale. The upward pressure on labor costs (observed as a plunge in the blue line) isn't particularly auspicious for future profits. Still, there's so much distortion in recent quarters that I'd consider the jury to be out on how much of this will be sustained.<blockquote>实际单位劳动力成本以左倒比例呈现。劳动力成本的上行压力（观察为蓝线暴跌）对未来利润来说并不是特别有利。尽管如此，最近几个季度仍存在如此多的扭曲，以至于我认为这种情况会持续多久还没有定论。</blockquote> <img src=\"https://static.tigerbbs.com/3110e7d3c0468c39d343b20835566bcc\" tg-width=\"1157\" tg-height=\"652\"><blockquote></blockquote> It's undoubtedly true that profit margins, expected growth, and other factors have an effect on future deliverable cash flows and the valuations that investors place on stocks. But even for technology stocks, these assumptions should be made explicit and tested against history. You'll find that observable measures like price/revenue are still very serviceable. In fact, the extreme price/revenue multiples of technology stocks helped to inform my March 2000 projection of an 83% loss in that sector.<blockquote>毫无疑问，利润率、预期增长和其他因素会影响未来可交付的现金流和投资者对股票的估值。但即使对于科技股来说，这些假设也应该明确，并根据历史进行检验。你会发现像价格/收入这样的可观察指标仍然非常有用。事实上，科技股极端的市盈率帮助我在2000年3月预测该行业将损失83%。</blockquote> As Benjamin Graham wrote, \"The habit of relating what is paid to what is being offered is an invaluable trait in investment. The more dependent the valuation becomes on anticipations of the future - and the less it is tied to a figure demonstrated by past performance - the more vulnerable it becomes to possible miscalculation and serious error.\"<blockquote>正如本杰明·格雷厄姆（Benjamin Graham）所写的那样，“将支付的东西与提供的东西联系起来的习惯是投资中的一个无价特征。估值越依赖于对未来的预期——它与过去表现所证明的数字的联系越少——它就越容易受到可能的误判和严重错误的影响。”</blockquote> The current 5.19 price/revenue multiple for the Nasdaq 100, is the most extreme level since February 2001, which was followed by a 60% loss in the index (after it had already dropped in half). The situation is actually a bit worse than 2001 here. If one examines the largest components of the index, it becomes clear that their annual growth rates have declined substantially over time.<blockquote>纳斯达克100指数目前的市盈率为5.19，是自2001年2月以来的最极端水平，随后该指数下跌了60%（此前该指数已经下跌了一半）。这里的情况实际上比2001年还要糟糕一些。如果我们研究该指数的最大组成部分，就会发现它们的年增长率随着时间的推移大幅下降。</blockquote> As a result, a dollar of current revenues should arguably command a smaller multiple than a dollar of revenues might have during earlier periods of emerging growth. Yet even if one takes the Nasdaq 100 price/revenue ratio at face value, and even if one restricts attention to the bubble period since 2000, it's difficult to expect the Nasdaq to produce total returns over the coming decade much better than zero.<blockquote>因此，可以说，当前一美元收入的倍数应该低于新兴增长早期一美元收入的倍数。然而，即使人们从表面上看待纳斯达克100指数的市盈率，即使人们将注意力限制在2000年以来的泡沫时期，也很难指望纳斯达克在未来十年产生比零好得多的总回报。</blockquote> <img src=\"https://static.tigerbbs.com/48dd88f74eaed48d53b237b4d0209d8e\" tg-width=\"963\" tg-height=\"542\"><blockquote></blockquote> You don't really want to see what the same chart looks like for the S&amp;P 500.<blockquote>你不会真的想看到标普500的同样的图表是什么样子的。</blockquote> <img src=\"https://static.tigerbbs.com/24f70ab683592ef20a8c6c75fc9bd1fc\" tg-width=\"963\" tg-height=\"543\"><blockquote></blockquote> The same is true, unfortunately, for passive investment strategies. We presently estimate negative 12-year average annual total returns for a conventional passive investment mix invested 60% in the S&amp;P 500, 30% in Treasury bonds, and 10% in Treasury bills.<blockquote>不幸的是，被动投资策略也是如此。我们目前估计，60%投资于标普500、30%投资于国债、10%投资于国库券的传统被动投资组合的12年平均年总回报率为负。</blockquote> In a 2019&nbsp;white paper, I detailed an approach to estimate a \"value-focused asset allocation\" by jointly considering prevailing stock market valuations and interest rates. It specifies an investment allocation based on which asset class is estimated to have the highest average annual expected return, adjusted for risk, to each point in a long-term investment horizon. That allocation can then be modified by a risk-management component, to adjust the exposure during segments of the market cycle where risk-aversion or speculation among market participants may temporarily drive valuations to depressed or elevated levels. The white paper includes numerous charts showing how this value-focused asset allocation has changed across market history, particularly at important peaks and troughs in the stock and bond markets.<blockquote>在2019年的白皮书中，我详细介绍了一种通过共同考虑当前股市估值和利率来估计“以价值为中心的资产配置”的方法。它根据哪个资产类别估计具有最高的平均年预期回报（经风险调整）来指定长期投资范围内的每个点的投资分配。然后，风险管理组件可以修改这种分配，以在市场周期的各个部分调整风险敞口，在这些部分，市场参与者的风险规避或投机可能会暂时将估值推低或推高。该白皮书包括大量图表，显示了这种以价值为中心的资产配置在市场历史上是如何变化的，特别是在股票和债券市场的重要高峰和低谷。</blockquote> Along with those methods, I introduced our \"Endowment/spending multiple,\" which estimates the number of years of spending that a passive 60/30/10 investor requires up-front, in order to finance an expected 36-year stream of future inflation-adjusted spending. The idea here is that in a deeply undervalued market with high expected future returns, investors can finance a future stream of spending with far less than they require when valuations are extreme and prospective returns are low.<blockquote>除了这些方法，我还介绍了我们的“捐赠/支出倍数”，它估计了被动的60/30/10投资者预先需要的支出年数，以便为未来36年的通胀调整支出提供资金。这里的想法是，在一个预期未来回报很高的被严重低估的市场中，当估值极端且预期回报较低时，投资者可以用远低于他们所需的资金为未来的支出流提供资金。</blockquote> You know you're in a bubble when funding a 36-year stream of expected inflation-adjusted spending requires over 38 years of money up-front.<blockquote>当为36年的预期通胀调整支出提供资金需要超过38年的前期资金时，你就知道自己处于泡沫之中。</blockquote> <img src=\"https://static.tigerbbs.com/1adff1ad0dbaa2737f60cb7589d206fb\" tg-width=\"1230\" tg-height=\"693\"><blockquote></blockquote> In the chart below, the Endowment/spending multiple is presented on an inverted log scale (left), along with the actual subsequent average annual total return of a 60/30/10 portfolio mix (right scale). Needless to say, we adhere to investment disciplines that are intended to address problems like this.<blockquote>在下图中，捐赠/支出倍数以倒对数标度（左）以及60/30/10投资组合的实际后续平均年总回报（右标度）呈现。不用说，我们坚持旨在解决此类问题的投资纪律。</blockquote> <img src=\"https://static.tigerbbs.com/1480e881b4d04ace81a96f31f6368796\" tg-width=\"1229\" tg-height=\"692\"><blockquote></blockquote> As a testament to the breadth of this speculative episode, the median price/revenue ratio of S&amp;P 500 components now exceeds 3.3, easily a record, and extreme enough to provoke distress about the potential losses that innocent but poorly-informed investors may experience over the completion of this cycle.<blockquote>作为这一投机事件广度的证明，标普500成分股的市盈率中位数现在超过了3.3，很容易创下纪录，而且极端程度足以引发无辜但消息不灵通的投资者在完成这一周期后可能遭受的潜在损失。</blockquote> <img src=\"https://static.tigerbbs.com/ffe0cc33bdaf0695dacb7ae1d817edab\" tg-width=\"1110\" tg-height=\"625\"><blockquote></blockquote> The next chart shows the median price/revenue ratio of S&amp;P 500 components sorted into 10 deciles by valuation. The chart is presented on log scale to allow each line to be compared with its own history. Each segment on the vertical axis represents a doubling of valuations. Notice that every single decile of S&amp;P 500 components is at record valuation extremes. Investors now rely on a permanently high plateau in these extremes.<blockquote>下图显示了按估值分为10个十分位数的标普500组件的市盈率中位数。该图表以对数标度显示，以便将每条线与其自身的历史进行比较。纵轴上的每个细分市场代表估值翻倍。请注意，标普500成分股的每一个十分位数都处于创纪录的估值极端水平。投资者现在依赖于这些极端情况下的永久高位平台。</blockquote> <img src=\"https://static.tigerbbs.com/cb7c995e64e3c9cba85d6e0fc9124223\" tg-width=\"1280\" tg-height=\"721\"><blockquote></blockquote> It's interesting, but far less important, that the median price/earnings ratio of S&amp;P 500 components has also reached 32.4, the highest level in history. This compares with a median P/E of 19.4 at the 2000 market peak. The problem with P/E multiples, of course, is that they are substantially affected by earnings variability. In fact, prior to the current peak, the highest median P/E for the S&amp;P 500 was actually in March 2002, when the index was down 25% from the March 2000 bubble highs. That's because earnings were down far more by then, which boosted P/E ratios. Such is the danger of taking P/E multiples at face value.<blockquote>有趣的是，但远没有那么重要的是，标普500组件的市盈率中位数也达到了32.4，为历史最高水平。相比之下，2000年市场峰值时的市盈率中值为19.4。当然，市盈率倍数的问题在于它们很大程度上受到收益可变性的影响。事实上，在当前的峰值之前，标普500的最高市盈率中值实际上是在2002年3月，当时该指数比2000年3月的泡沫高点下跌了25%。这是因为当时盈利下降幅度更大，这提高了市盈率。这就是按面值计算市盈率倍数的危险。</blockquote> A crude but reasonably effective way to get around the cyclicality of earnings (but only for very broad indices), is to compare prices to the highest level of earnings achieved to-date. I introduced this metric back in 1998 as the price-to-peak-earnings ratio. The chart below shows a version of that. The blue line (left scale) shows the ratio of total U.S. equity market capitalization to GDP. The red line (right scale) shows the ratio of total U.S. equity market capitalization to the highest level of economy-wide U.S. profits to date. Investors have gotten themselves into trouble here.<blockquote>绕过收益周期性的一个粗略但相当有效的方法（但仅限于非常广泛的指数）是将价格与迄今为止达到的最高收益水平进行比较。我早在1998年就引入了这个指标，称为市盈率。下图显示了一个版本。蓝线（左标度）显示了美国股票总市值与GDP的比率。红线（右标度）显示了美国股票总市值与迄今为止美国整体经济利润最高水平的比率。投资者在这里遇到了麻烦。</blockquote> <img src=\"https://static.tigerbbs.com/a6630cf8302360e1b4645870051b21de\" tg-width=\"989\" tg-height=\"557\"><blockquote></blockquote> What if valuations remain extreme forever?<blockquote>如果估值永远保持极端怎么办？</blockquote> Probably the single most frequent question I've heard from investors over the past couple of years is \"What happens if valuations remain extreme forever?\" It's actually a version of the \"permanently high plateau\" that Irving Fisher disastrously projected in 1929. Still, recent years have produced enormous confidence among investors that the Federal Reserve's purchases of Treasury debt can permanently \"backstop\" the stock market.<blockquote>过去几年我从投资者那里听到的最常见的问题可能是“如果估值永远保持极端会发生什么？”这实际上是欧文·费希尔在1929年灾难性地预测的“永久高原”的一个版本。尽管如此，近年来投资者对美联储购买国债可以永久“支撑”股市充满信心。</blockquote> As I've detailed previously, quantitative easing supports the market only by creating zero interest hot potatoes that are uncomfortable for investors to hold (provided that they're inclined toward speculation), and that are impossible to get rid of in aggregate. Moreover, the Fed's purchases of corporate bonds during the pandemic were legally constrained to CARES funds provided by the Treasury, and ultimately amounted to $14 billion of bonds, in an economy with $11 trillion in corporate debt at $58 trillion in equity market capitalization.<blockquote>正如我之前详细介绍的，量化宽松只能通过创造零利息的烫手山芋来支持市场，这些烫手山芋让投资者持有起来不舒服（前提是他们倾向于投机），而且总体上不可能摆脱这些烫手山芋。此外，在一个拥有11万亿美元公司债务、58万亿美元股票市值的经济体中，美联储在疫情期间购买的公司债券在法律上仅限于财政部提供的CARES资金，最终债券金额为140亿美元。</blockquote> Suffice it to say that the \"Fed backstop\" is largely in the minds of investors, and relies almost exclusively on the psychological discomfort of holding low-yielding base money. Yet since perception can be indistinguishable from reality, particularly in the short run, it's important to entertain the question.<blockquote>可以说，“美联储支持”很大程度上存在于投资者的脑海中，几乎完全依赖于持有低收益基础货币的心理不适。然而，由于感知可能与现实难以区分，尤其是在短期内，所以考虑这个问题很重要。</blockquote> My answer is that if the Fed is indeed able to maintain valuations at the highest levels in history, forever, stock prices would likely grow at roughly the same rate as nominal GDP. So figure 1.6% real structural growth plus 2% inflation gets you 3.6%. Let's call it 4%, which would match nominal GDP growth in both the 10-year and 20-year periods ending at the Q4 2019 economic peak, just before the pandemic. The first casualty of rising inflation is stock valuations, so it's not at all clear that assuming higher inflation would help stocks until valuations were roughly normalized, which would require consumer prices to roughly triple.<blockquote>我的回答是，如果美联储确实能够永远将估值维持在历史最高水平，股价可能会以与名义GDP大致相同的速度增长。因此，1.6%的实际结构性增长加上2%的通胀率得到3.6%。让我们将其看涨期权为4%，这将与截至2019年第四季度经济峰值（就在大流行之前）的10年和20年期间的名义GDP增长相匹配。通胀上升的第一个受害者是股票估值，因此在估值大致正常化之前，假设通胀上升是否会对股票有所帮助，这一点还不清楚，这需要消费者价格大约上涨两倍。</blockquote> The chart below is a reminder of how structural real GDP growth has progressed over recent decades (driven by demographic labor force growth and trend productivity), and the basis for that 1.6% figure for structural real GDP growth.<blockquote>下图提醒我们近几十年来结构性实际GDP增长是如何进展的（由人口劳动力增长和趋势生产率驱动），以及结构性实际GDP增长1.6%数字的基础。</blockquote> <img src=\"https://static.tigerbbs.com/cf63b9b8f5f3b55136c8523d73e864f3\" tg-width=\"1131\" tg-height=\"637\"><blockquote></blockquote> Sticking with 4% nominal growth, and adding a 1.5% dividend yield, a \"permanently high plateau\" in market valuations would imply S&amp;P 500 total returns of about 5.5% annually. Again, this assumes that valuations never retreat from levels that presently stand at about 3.6 times their historical norms. Simply allow them to retreat to 2.4 times their historical norms a decade from now - which would still keep valuations among the highest 10% in U.S. history - and the resulting 10-year total return would drop to about 1.3%. I think this would actually be the best-case scenario even in a permanently overvalued world.<blockquote>坚持4%的名义增长率，加上1.5%的股息收益率，市场估值的“永久高位”将意味着标普500每年的总回报率约为5.5%。同样，这是假设估值永远不会从目前约为历史标准3.6倍的水平回落。只要让它们在十年后回落至历史正常水平的2.4倍——这仍将使估值保持在美国历史上最高的10%之内——由此产生的10年总回报率将降至1.3%左右。我认为，即使在一个永远被高估的世界里，这实际上也是最好的情况。</blockquote> At elevated valuations, even very small changes in expected return imply enormous changes in prices. So it's unlikely that a period of much higher average valuations will escape the prospect of relatively high volatility. Rather than a 70% market decline, which would presently be required for the S&amp;P 500 to simply touch historically run-of-the-mill valuation norms, investors could expect rather frequent market losses in the 20-35% range, which is essentially what we've seen even over the past few years.<blockquote>在高估值下，即使预期回报的微小变化也意味着价格的巨大变化。因此，平均估值高得多的时期不太可能逃脱相对较高波动性的前景。投资者可能预计市场会出现20-35%的频繁下跌，而不是70%的市场下跌，这是标普500目前简单地触及历史上的普通估值标准所必需的，这基本上是我们在过去几年中所看到的。</blockquote> All of that would be fine with us. We've adapted our discipline sufficiently (especially in late-2017) to tolerate the possibility of permanently sustained overvaluation. My impression is that the impact of those adaptations has become more evident as we've had greater opportunities to live into them. I can't say that I believe for a second that investors will actually be spared from a 50-70% loss in the S&amp;P 500 in the coming years, but again, it will be fine with us if the market never approaches historical valuation norms again. With the adaptations we introduced in late-2017, our discipline is flexible enough to navigate a bubble even without embracing its premise.<blockquote>所有这些对我们来说都没问题。我们已经充分调整了我们的纪律（尤其是在2017年末），以容忍永久持续高估的可能性。我的印象是，随着我们有更多的机会生活在其中，这些适应的影响变得更加明显。我不能说我一秒钟都相信投资者实际上不会在未来几年的标普500中遭受50-70%的损失，但同样，如果市场永远不会接近历史估值规范，我们也没问题。随着我们在2017年底引入的适应，我们的学科足够灵活，即使不接受泡沫的前提，也可以驾驭泡沫。</blockquote> An unusual overlap of high-risk conditions<blockquote>高风险状况的不寻常重叠</blockquote> Returning to Modigliani, my impression is that the advance of recent years to the most extreme valuations in history reflects exactly the bubble dynamics he described, and I expect that it will also end as he described (though not necessarily in one fell swoop): \"The expectation of growth produces the growth, which confirms the expectation; people will buy it because it went up. But once you are convinced that it is not growing anymore, nobody wants to hold a stock because it is overvalued. Everybody wants to get out and it collapses, beyond the fundamentals.\"<blockquote>回到莫迪里阿尼，我的印象是，近年来向历史上最极端估值的推进恰恰反映了他所描述的泡沫动态，我预计它也会像他所描述的那样结束（尽管不一定一蹴而就）：“对增长的预期产生了增长，增长证实了预期；人们会因为它上涨而买入。但一旦你确信它不再增长，就没有人愿意持有一只股票，因为它被高估了。每个人都想退出，但它崩溃了，超出了基本面。”</blockquote> One of our internal gauges tracks the correlation of market conditions with certain high-risk features that have preceded steep market collapses - a collection of measures capturing valuations, internals, sentiment, leverage, overextension, and yield pressures. Only a handful of instances in history overlap pre-crash conditions as well as they do at present. The current overlap is actually quite similar to August 1987. Meanwhile, the correlation of current conditions with features typically observed at market lows is the most negative in history.<blockquote>我们的一个内部指标跟踪市场状况与市场急剧崩溃之前的某些高风险特征的相关性——一系列衡量估值、内部因素、情绪、杠杆、过度扩张和收益率压力的指标。历史上只有少数情况像现在这样与崩溃前的情况重叠。目前的重叠实际上与1987年8月非常相似。与此同时，当前状况与市场低点时通常观察到的特征的相关性是历史上最负的。</blockquote> If you want my opinion, I suspect that a near-vertical market plunge on the order of 25-35% is coming, probably quite shortly, most likely out of the blue, as in 1987, driven by nothing more than the sudden concerted effort of overextended investors to sell, and the need for a large price adjustment in order to induce scarce buyers to take the other side.<blockquote>如果你想听听我的意见，我怀疑25-35%左右的近乎垂直的市场暴跌即将到来，可能很快，很可能是出乎意料的，就像1987年一样，无非是由过度扩张的投资者突然齐心协力抛售，以及需要大幅调整价格以诱导稀缺买家站在另一边。</blockquote> As usual, no forecasts are necessary. We'll align our investment stance in response to the valuations and market action that we observe at each point in time. Still, it's of particular concern that these overlaps are occurring in the context of the most extreme valuations in history, along with strikingly dysfunctional pockets of illiquidity in many individual issues. This dysfunctional behavior isn't about any particular video game retailer. I suspect it's actually about some sort of fragility or segmentation in order-flow mechanisms, possibly coupled with poorly managed derivatives exposure.<blockquote>像往常一样，不需要预测。我们将根据每个时间点观察到的估值和市场行为来调整我们的投资立场。尽管如此，特别令人担忧的是，这些重叠发生在历史上最极端估值的背景下，以及许多个别问题中明显功能失调的流动性不足的背景下。这种不正常的行为与任何特定的视频游戏零售商无关。我怀疑这实际上是关于订单流机制的某种脆弱性或分割，可能与管理不善的衍生品敞口有关。</blockquote> As I used to teach my students, show me a financial debacle, and I'll show you someone who had a leveraged, mismatched position that they were suddenly forced to close into an illiquid market.<blockquote>正如我曾经教我的学生的那样，给我看一个金融崩溃，我会给你看一个拥有杠杆、不匹配头寸的人，他们突然被迫关闭到一个缺乏流动性的市场。</blockquote> Though my concerns run far beyond the amount of leverage in the system, it isn't helpful that the amount of leverage in the U.S. equity markets is now easily the highest in history. Some observers are inclined to bring this figure down by dividing instead by the market capitalization of equities. But here's some useful arithmetic:<blockquote>尽管我的担忧远远超出了系统中的杠杆金额，但美国股市的杠杆金额现在轻松达到历史最高水平并没有什么帮助。一些观察家倾向于通过除以股票市值来降低这一数字。但这里有一些有用的算术：</blockquote> Margin Debt/GDP = Margin Debt/Market Cap x Market Cap/GDP<blockquote>保证金债务/GDP=保证金债务/市值x市值/GDP</blockquote> To say that margin debt to GDP is at the highest level in history is to say not only that stocks are heavily owned on margin, but that those stocks are also breathtakingly overvalued. That combination is particularly worrisome.<blockquote>说保证金债务占GDP的比例处于历史最高水平，不仅是说保证金大量持有股票，而且这些股票的估值也高得惊人。这种组合尤其令人担忧。</blockquote> All crises have involved debt that, in one fashion or another, has become dangerously out of scale in relation to the underlying means of payment. - John Kenneth Galbraith, A Short History of Financial Euphoria <img src=\"https://static.tigerbbs.com/dc55dacbd5e01016cd9964422f941b0c\" tg-width=\"983\" tg-height=\"556\"><blockquote>所有危机都涉及债务，这些债务以这样或那样的方式与基本支付手段相比已经危险地超出了规模。——约翰·肯尼思·加尔布雷斯，《金融兴奋简史》</blockquote> The kind of event I'm suggesting would not bring valuations anywhere near historical valuation norms. Given current valuation extremes, it would be more like a palate cleanser. I have no particular expectation about what the next dish would be. In the event we do observe an abrupt market collapse, the Fed will undoubtedly respond with some new palliative. Whether or not it is effective will depend on the context of risk-aversion, inflation, credit risk, and other conditions at the time. The larger problem, as we've discussed, is that you can't \"save\" an overvalued asset by propping up its price. The value is in the future cash flows that will be delivered to investors over time. The elevated price only ensures that the long-term return between now and then will be dismal.<blockquote>我所建议的这种事件不会使估值接近历史估值规范。鉴于目前的极端估值，它更像是一种味觉清洁剂。我对下一道菜没有特别的期望。如果我们确实观察到市场突然崩溃，美联储无疑会采取一些新的姑息措施来应对。它是否有效将取决于当时的风险规避、通货膨胀、信用风险和其他条件。正如我们所讨论的，更大的问题是，你不能通过支撑价格来“拯救”被高估的资产。价值在于随着时间的推移将交付给投资者的未来现金流。价格上涨只会确保从现在到那时的长期回报将是惨淡的。</blockquote> Of course, nothing in our discipline relies on a market plunge. Given the combination of hypervaluation and divergent market internals (largely based on debt securities, but with increasing divergences in equities as well), I do believe that the stock market remains in a \"trap door\" situation. Still, that view will change as market conditions change. We'll refrain from adopting or amplifying a negative outlook if our measures of market internals improve. That discipline has served us well even amid record highs. Again, no forecasts are required, nor does this opinion drive our current investment stance. I just think the correlation with historical pre-crash conditions is worth noting.<blockquote>当然，我们的学科中没有任何东西依赖于市场暴跌。鉴于高估和市场内部分歧（主要基于债务证券，但股票分歧也在增加）的结合，我确实认为股市仍处于“活板门”的情况。尽管如此，这种观点将随着市场状况的变化而改变。如果我们对市场内部的衡量有所改善，我们将避免采用或放大负面前景。即使在创纪录的高点下，这种纪律也为我们提供了良好的服务。同样，不需要预测，这种观点也不会驱动我们当前的投资立场。我只是认为与历史碰撞前情况的相关性值得注意。</blockquote> How little, it will perhaps be agreed, was either original or otherwise remarkable about this history. Prices driven up on the expectation that they would go up, the expectation realized by the resulting purchases. Then the inevitable reversal of these expectations because of some seemingly damaging event or development or perhaps merely because the supply of intellectually vulnerable buyers was exhausted. Whatever the reason (and it is unimportant), the absolute certainty is that this world ends not with a whimper but with a bang. And so on to the moment of mass disillusion and the crash. This last, it will now be sufficiently evident, never comes gently. It is always accompanied by a desperate and largely unsuccessful effort to get out. - John Kenneth Galbraith on the 1929 collapse Valuations and investment duration<blockquote>也许人们会同意，关于这段历史，无论是原创的还是非凡的，都很少。价格上涨是基于价格会上涨的预期，这种预期通过由此产生的购买来实现。然后，由于一些看似破坏性的事件或发展，或者仅仅是因为智力脆弱的买家的供应耗尽，这些预期不可避免地逆转。不管原因是什么（这并不重要），绝对可以肯定的是，这个世界不是随着呜咽而结束，而是随着一声巨响而结束。等等，直到大规模幻灭和崩溃的时刻。这最后一点，现在已经足够明显了，从来不会轻易到来。它总是伴随着绝望且基本上不成功的脱身努力。——约翰·肯尼思·加尔布雷斯谈1929年的崩盘估值和投资期限</blockquote> To understand why extreme valuations imply high volatility, and require extremely long investment horizons, it's useful to consider the concept of \"duration.\"<blockquote>为了理解为什么极端估值意味着高波动性，并且需要极长的投资期限，考虑“久期”的概念是有用的。</blockquote> Every security is a claim on a stream of future cash flows that can be expected to be delivered to the investor over time. While the concept of \"duration\" is most commonly used for bonds, it's actually applicable to any security, no matter how \"lumpy\" the stream of cash flows might be.<blockquote>每种证券都是对未来现金流的债权，这些现金流预计将随着时间的推移交付给投资者。虽然“久期”的概念最常用于债券，但它实际上适用于任何证券，无论现金流有多“不稳定”。</blockquote> If you don't like math, feel free to skim over the various equations and just read the pull quotes. I've provided the details just for completeness.<blockquote>如果你不喜欢数学，可以随意浏览各种方程，只需阅读引号。为了完整起见，我提供了细节。</blockquote> The \"duration\" of a security can be defined in two ways.<blockquote>证券的“久期”可以通过两种方式定义。</blockquote> Investment horizon: the weighted average number of years it takes for the security to deliver its payments. For each period, you take the share of total present value represented by that year's payment, and multiply it by the number of years in the future the payment will be received. Add them all up. The result is the number of years from today (a weighted average) that the present value of your investment will be repaid. For example, the duration of a security that delivers a single payment a decade from now is simply 10 years.<blockquote>投资期限：证券支付款项所需的加权平均年数。对于每个期间，您将当年付款所代表的总现值份额乘以未来收到付款的年数。把它们加起来。结果是从今天起您的投资现值将得到偿还的年数（加权平均值）。例如，十年后一次性付款的证券的期限仅为10年。</blockquote> Elasticity: the percentage change in the security in response to a change in the underlying gross rate of return. Technically, elasticity is (-dP/P)/(dk/1+k). For example, suppose you own a security that will pay $100 a decade from now, and it's priced at $82.0348, for a 2% annual return. Now assume the expected return moves to 2.01%. The price would drop to $81.9544. Elasticity is (0.0804/82.0348)/(0.0001/1.02) = 10%<blockquote>弹性：证券响应基础总回报率变化的百分比变化。从技术上讲，弹性是(-dP/P)/(dk/1+k)。例如，假设您拥有一种证券，十年后将支付100美元，其定价为82.0348美元，年回报率为2%。现在假设预期回报率升至2.01%。价格将降至81.9544美元。弹性为(0.0804/82.0348)/(0.0001/1.02)=10%</blockquote> It turns out that the \"duration\" of a security in years is identical to its \"duration\" in terms of the percentage change of price in response to a 1% fluctuation in expected returns. Duration is also the holding period that \"immunizes\" the investor against changes in expected returns over time. In other words, assuming you reinvest your cash flows over time, duration gives you a good idea of how long you have to hold the security in order for your ending wealth to be largely independent of the fluctuations that the security experiences over that horizon.<blockquote>事实证明，就预期回报1%波动的价格变化百分比而言，证券以年为单位的“久期”与其“久期”相同。久期也是投资者对预期回报随时间变化“免疫”的持有期。换句话说，假设您随着时间的推移对现金流进行再投资，久期可以让您很好地了解您必须持有证券多长时间，以便您的最终财富在很大程度上独立于证券在该期间经历的波动。</blockquote> If you know differentiation, can prove to yourself that the \"modified duration\" of the S&amp;P 500 is essentially the inverse of the dividend yield. The modified duration is just (-dP/P)/dk or Macaulay duration/(1+k).<blockquote>如果你知道差异化，可以向自己证明标普500的“修正久期”本质上是股息收益率的倒数。修改后的持续时间只是（-dP/P）/dk或麦考利持续时间/（1+k）。</blockquote> Consider P = D/(k-g). Differentiating with respect to k, dP/dk = -P/(k-g), so duration (-dP/P)/dk = P/D.<blockquote>考虑P=D/（k-g）。相对于k微分，dP/dk=-P/（k-g），因此持续时间（-dP/P）/dk=P/D。</blockquote> Here's how to think about the link between valuations and duration. Presently, the dividend yield of the S&amp;P 500 is 1.48%. If the yield moves to 1.49%, holding dividends constant, prices drop by 1.48/1.49-1 = -0.67% on that 1 basis point move. Duration is just that sensitivity, defined for a 100 basis point move, which would be 67. It turns out that's also the weighted-average number of years from today that you'll receive your present value, if you invest today.<blockquote>以下是如何思考估值和久期之间的联系。目前，标普500的股息率为1.48%。如果收益率升至1.49%，保持股息不变，则价格在这1个基点的变动上下跌1.48/1.49-1=-0.67%。久期就是敏感度，定义为100个基点的变动，即67。事实证明，如果您今天投资，这也是从今天起您将获得现值的加权平均年数。</blockquote> Compare that to the typical situation over the past century, when the dividend yield of the S&amp;P 500 averaged about 3.7%. At normal valuations, a 1 basis point increase in the dividend yield would produce a price drop of 3.7/3.71-1 = 0.27%, implying a duration of 27 years.<blockquote>与过去一个世纪的典型情况相比，当时标普500的股息收益率平均约为3.7%。在正常估值下，股息收益率上升1个基点将导致价格下跌3.7/3.71-1=0.27%，这意味着持续时间为27年。</blockquote> It turns out that the 'duration' of a security in years is identical to its 'duration' in terms of the percentage change of price in response to a 1% fluctuation in expected returns. Duration is also the holding period that 'immunizes' the investor against changes in expected returns over time. Historically, investors wishing to match the duration of their investment portfolio to the duration of their investment horizon could be reasonably comfortable holding 100% of their assets in stocks, provided they had an investment horizon of about 25-30 years. Presently, these investors would need an investment horizon closer to 65-70 years. They are currently holding sippy cups.<blockquote>事实证明，就预期回报1%波动的价格变化百分比而言，以年为单位的证券的“久期”与其“久期”相同。久期也是使投资者免受预期回报随时间变化的影响的持有期。从历史上看，希望将投资组合的久期与投资期限相匹配的投资者可以相当放心地将100%的资产持有股票，前提是他们的投资期限约为25-30年。目前，这些投资者需要接近65-70年的投资期限。他们目前拿着吸管杯。</blockquote> Scarcity, usefulness, and value<blockquote>稀缺性、有用性和价值</blockquote> While we're on the subject of bubbles, I'll add a few comments on Bitcoin, just for fun. I'd write more, but my sides still hurt from laughing.<blockquote>当我们谈到泡沫的话题时，我会在比特币上补充一些评论，只是为了好玩。我会写更多，但我的身体仍然笑得很痛。</blockquote> Objects like tulip bulbs and Bitcoin differ from securities in that they do not deliver a stream of cash flows to the holder. Instead, what objects like tulips and currencies provide is a little stream of services over time, for example, as a perennial thing of beauty or as a means of payment. What people sometimes forget is that it is not just scarcity that defines the value of an object, but the stream of useful \"services\" that it provides (for some reason, nobody wants to buy my unique, limited edition, digitally-signed porcupine seat covers). The price of the object, and the stream of services it provides, should be commensurate.<blockquote>像郁金香球茎和比特币这样的物品与证券的不同之处在于，它们不会向持有者提供现金流。相反，像郁金香和货币这样的物品随着时间的推移提供了一点服务，例如，作为一种永恒的美丽事物或作为一种支付手段。人们有时会忘记的是，定义一件物品价值的不仅仅是稀缺，而是它提供的一系列有用的“服务”（出于某种原因，没有人想买我独特的限量版数字签名豪猪座套）。物品的价格和它提供的服务流应该是相称的。</blockquote> U.S. dollars, for example, have value primarily because they are tethered to the real economy by fiat (they legally must be accepted as a means of payment, as noted on the face of any dollar bill), and they represent the entire substrate of the banking system - nearly every payment that goes back and forth in the U.S. economy represents a transfer of base money. Base money (currency and bank reserves) provides billions of little \"services\" over time.<blockquote>例如，美元之所以有价值，主要是因为它们通过法定货币与实体经济联系在一起（正如任何美钞的正面所注明的那样，它们在法律上必须被接受为一种支付手段），并且它们代表了银行体系的整个基础——几乎在美国经济中来回的每一笔支付都代表着基础货币的转移。随着时间的推移，基础货币（货币和银行储备）提供了数十亿的小“服务”。</blockquote> With every transaction, reserves move electronically from bank to bank between one account holder and another. That combination of legal fiat and constant use as a substrate of the payments system is what gives money \"value.\" That value also means that the U.S. government essentially obtains revenue as \"seigniorage\" for producing the stuff. For those who imagine that governments are going to surrender that revenue in favor of using Bitcoin, I've got a non-fungible token to sell you.<blockquote>在每一笔交易中，准备金都会以电子方式在一个账户持有人和另一个账户持有人之间从一家银行转移到另一家银行。法定法令和作为支付系统基础的持续使用的结合赋予了货币“价值”。这一价值也意味着美国政府本质上通过生产这些东西获得了“铸币税”的收入。对于那些认为政府会放弃这些收入而使用比特币的人，我有一个不可替代的代币要卖给你。</blockquote> <img src=\"https://static.tigerbbs.com/747b2ba31a8d94d85d374d12b13b3c96\" tg-width=\"665\" tg-height=\"718\"><blockquote></blockquote> As I've noted before, blockchain is a brilliant algorithm, and I expect that it will have a great number of uses for secure transactions and inventory management. Bitcoin, however, is a token generated by an energy-inefficient, replicable blockchain app. Ultimately, its value rests on the capacity to provide transactions services, yet without fiat to require its use, and with strikingly narrow bandwidth - one block of roughly 2000 transactions every 10 minutes - that I expect will prove to be a wildly limiting feature. That's a problem in a world where speculators now value the stock of bitcoin at one-fifth the value of the entire U.S. monetary base.<blockquote>正如我之前提到的，区块链是一个出色的算法，我预计它将在安全交易和库存管理方面有大量用途。然而，比特币是一种由低能效、可复制的区块链应用程序生成的代币。最终，它的价值取决于提供交易服务的能力，但没有法定要求使用它，而且带宽非常窄——每10分钟大约有2000笔交易——我预计这将被证明是一个非常有限的功能。在投机者现在对比特币股票的估值是整个美国货币基础价值的五分之一的世界里，这是一个问题。</blockquote> If you think about how money is valued, it's clear that people accept it because they believe it will provide a claim on the future output of others. Of course, that expectation requires that future producers will also give away their output and accept the money, on the belief that yet other future producers will do the same. That expectation has to continue indefinitely. Like the question 'What holds up Atlas when Atlas holds up the world?' it's not enough to answer that he's standing on a turtle. It's got to be turtles all the way down. The value of money has an enormous psychological component.\" - John P. Hussman, Ph.D., Turtles All the Way Down, February 2019 Of course, Bitcoin may have a certain user base as a vehicle for money laundering and black market transactions, but that's an undesirable investment thesis. The vast majority of transactions are to exchange Bitcoin itself, though the New York Times did recently report that \"pornography, patio furniture, and an at-home coronavirus test are among the odd assortment of goods and services that people are purchasing with the cryptocurrency.\" So, basically, if your typical day consists of surfing porn on your patio while testing yourself for COVID, you're gonna want to look into Bitcoin.<blockquote>如果你想想货币是如何被估价的，很明显，人们接受它是因为他们相信它会提供对他人未来产出的要求。当然，这种期望要求未来的生产者也会放弃他们的产出并接受这笔钱，因为他们相信其他未来的生产者也会这样做。这种期望必须无限期地持续下去。就像这样的问题“当阿特拉斯支撑着世界时，是什么支撑着阿特拉斯？“仅仅回答他站在一只乌龟上是不够的。下面一定是海龟。金钱的价值具有巨大的心理成分。”–John P.Hussman，博士，海龟一路下跌，2019年2月当然，比特币可能拥有一定的用户群作为洗钱和黑市交易的工具，但这是一个不受欢迎的投资论点。绝大多数交易都是为了交换比特币本身，尽管《纽约时报》最近确实报道称，“色情、庭院家具和家庭冠状病毒测试都是人们用加密货币购买的各种奇怪的商品和服务。”所以，基本上，如果你典型的一天是在院子里浏览色情片，同时测试自己的COVID，你会想看看比特币。</blockquote> My largest concern is that people are actually forking over hard-earned savings in exchange for these tokens, which allows early \"miners\" to cash out. That's essentially the defining feature of a Ponzi scheme. Like all speculative bubbles that rely on increases in price, rather than cash flows generated by the production of value-added goods and services, Bitcoin isn't actually creating \"wealth.\" It's only creating the opportunity for wealth transfer, primarily from those who will end up holding the bag.<blockquote>我最担心的是，人们实际上是在用辛苦赚来的积蓄来换取这些代币，这使得早期的“矿工”可以套现。这本质上是庞氏骗局的定义特征。与所有依赖价格上涨而不是增值商品和服务生产产生的现金流的投机泡沫一样，比特币实际上并没有创造“财富”。这只是为财富转移创造了机会，主要是来自那些最终将掌管一切的人。</blockquote> Bitcoin has certain characteristics of base money in the sense that it's exchanged on an electronic ledger, but by design, transactions are limited to an average of about 2000 per block, with one block successfully validated, on average, every 10 minutes. In order to validate a transaction block, CPU farms across the world grind out terahashes of random SHA256 validation attempts in order to discover a sufficiently small binary that matches the cryptographic hash of the block.<blockquote>比特币具有基础货币的某些特征，因为它是在电子分类账上进行交易的，但根据设计，每个区块的交易平均限制在2000笔左右，平均每10分钟就有一个区块成功验证。为了验证一个事务块，世界各地的CPU场会研磨出随机SHA256验证尝试的万亿次哈希，以便发现与该块的加密哈希相匹配的足够小的二进制文件。</blockquote> All of this \"mining\" burns up about as much energy as it takes to run a modest-sized country. Validating a block of transactions produces a reward to the miner (and dilution of the coinbase) of 6.25 Bitcoin per block, which currently works out to nearly $200 per transaction. Yet the value of the median transaction in Bitcoin is only about $1000 in the first place.<blockquote>所有这些“采矿”消耗的能量大约相当于运行一个中等规模的国家所需的能量。验证一笔交易会给矿工带来每笔6.25比特币的奖励（并稀释coinbase），目前每笔交易的奖励接近200美元。然而，比特币的交易价值中值一开始只有1000美元左右。</blockquote> There's a rather primitive regression analysis floating around (tagged as \"sophisticated\" by some observers who apparently go numb at the word \"logarithm\") that attempts to relate the log price of bitcoin to the log \"stock/flow\" ratio, as if it represents some mechanistic supply-demand relationship. Aside from the fact that the correlation between two diagonal lines is always about 0.9-something, I find that one can obtain a better fit just by regressing the log price of Bitcoin on the log ratio of block difficulty/block reward, which is basically a measure of how much energy one needs to waste in order to mine a new bitcoin.<blockquote>有一种相当原始的回归分析（被一些显然对“对数”这个词麻木的观察者标记为“复杂”），试图将比特币的对数价格与对数“库存/流量”比率联系起来，好像它代表了某种机械的供需关系。除了两条对角线之间的相关性总是在0.9左右这一事实之外，我发现，只要将比特币的对数价格回归到区块难度/区块奖励的对数比率，就可以获得更好的拟合，这基本上是一种衡量一个人为了挖掘新的比特币需要浪费多少能量的方法。</blockquote> So the \"value\" of Bitcoin is partially linked to the backward-looking sunk cost of the energy wasted to mine these tokens. Still, I wouldn't dream of using this sort of \"model\" to trade an object whose \"value\" is primarily in the heads of speculators. Use it if you like. If you happen make money on it, feel free send me a check, preferably in U.S. dollars.<blockquote>因此，比特币的“价值”部分与开采这些代币所浪费的能源的向后看沉没成本有关。尽管如此，我不会梦想用这种“模型”来交易一个“价值”主要在投机者头脑中的对象。喜欢就用吧。如果你碰巧在上面赚钱，请随时给我寄一张支票，最好是美元。</blockquote> Undoubtedly, this view of Bitcoin will be unpopular among those who associate holding Bitcoin with superpowers like laser eyes and diamond hands. \"Not surprised Hussman doesn't get Bitcoin. Few do.\" M'kay. Look, there's certainly a case to be made that a speculative mindset creates its own reality, and while it does, there's an opportunity to obtain wealth transfers from frantic late-comers who can no longer tolerate missing out. Tulips gonna tulip. Not my gig, thanks.<blockquote>毫无疑问，这种对比特币的看法将在那些将持有比特币与激光眼和钻石手这样的超能力联系在一起的人中间不受欢迎。“赫斯曼不了解比特币并不奇怪。很少有人了解。”好吧。看，投机心态创造了自己的现实，当然有理由证明，虽然它创造了自己的现实，但也有机会从疯狂的后来者那里获得财富转移，他们再也不能容忍错过机会。郁金香会郁金香。不是我的工作，谢谢。</blockquote> In the short run, it will be said to be an attack, motivated by either deficient understanding or uncontrolled envy, of the wonderful process of enrichment. Those involved with the speculation are experiencing an increase in wealth - getting rich or being further enriched. No one wishes to believe that this is fortuitous or undeserved; all wish to think that it is the result of their own superior insight or intuition. As long as they are in, they have a strong pecuniary commitment to belief in the unique personal intelligence that tells them there will be yet more. Accordingly, possession must be associated with some special genius. Speculation buys up, in a very practical way, the intelligence of those involved. Only after the speculative collapse does the truth emerge. What was thought to be unusual acuity turns out to be only a fortuitous and unfortunate association with the assets. - John Kenneth Galbraith, A Brief History of Financial Euphoria <blockquote>从短期来看，人们会说这是一种攻击，其动机要么是缺乏理解，要么是无法控制的嫉妒，对丰富的奇妙过程。那些参与投机的人正在经历财富的增加——变得富有或进一步变得富有。没有人愿意相信这是偶然的或不应得的；所有人都希望认为这是他们自己卓越的洞察力或直觉的结果。只要他们在，他们就有一个强大的金钱承诺，相信独特的个人智慧告诉他们还会有更多。因此，附身必须与某种特殊的天才联系在一起。投机以一种非常实际的方式收买了相关人员的智力。只有在投机崩溃之后，真相才会浮出水面。被认为是不寻常的敏锐度被证明只是与资产的偶然和不幸的联系。——约翰·肯尼思·加尔布雷斯，《金融欣快简史》</blockquote>\n<div class=\"bt-text\">\n\n\n 来源：<a href=\"https://www.hussmanfunds.com/comment/mc210315/\">Hussman Funds</a>\n为提升您的阅读体验，我们对本页面进行了排版优化\n\n\n</div>\n</article>\n</div>\n</body>\n</html>\n","type":0,"thumbnail":"","relate_stocks":{".SPX":"S&P 500 Index",".IXIC":"NASDAQ Composite",".DJI":"道琼斯"},"source_url":"https://www.hussmanfunds.com/comment/mc210315/","is_english":true,"share_image_url":"https://static.laohu8.com/e9f99090a1c2ed51c021029395664489","article_id":"1172271196","content_text":"Summary\n\nThe defining feature of a bubble is inconsistency between expected returns based on price behavior and expected returns based on valuations.\nIf we compare our most reliable valuation measures with the valuation measures that one would obtain from a proper long-term discounted cash flow analysis, we find that they're nearly identical.\nOne of the unfortunate bits of financial illiteracy that Wall Street has pushed into the heads of investors is the idea that extreme valuations are \"justified\" by low interest rates.\nIt's undoubtedly true that profit margins, expected growth, and other factors have an effect on future deliverable cash flows and the valuations that investors place on stocks.\nTo understand why extreme valuations imply high volatility, and require extremely long investment horizons, it's useful to consider the concept of \"duration.\"\n\n\n I can show, really precisely, that there are two warranted prices for a share. The one I prefer is based on such fundamentals as earnings and growth rates, but the bubble is rational in a certain sense. The expectation of growth produces the growth, which confirms the expectation; people will buy it because it went up. But once you are convinced that it is not growing anymore, nobody wants to hold a stock because it is overvalued. Everybody wants to get out and it collapses, beyond the fundamentals.- Nobel Laureate Franco Modigliani, New York Times, March 30, 2000\n\nThe word \"bubble\" is tossed around quite a bit in the financial markets, but it's rarely used correctly. See, the thing that defines a bubble isn't that valuations are extremely high, or that expected returns are extremely low. Instead, what defines a bubble is that investors drive valuations higher without simultaneously adjusting expectations for returns lower. That is, investors extrapolate past returns based on price behavior, even though those expectations are inconsistent with the returns that would equate price with discounted cash flows.\nIn March 2000, at the height of the technology bubble, I noted: \"Over time, price/revenue ratios come back in line. Currently, that would imply an 83% plunge in tech stocks. If you understand values and market history, you know we're not joking.\" The following month, I discussed Modigliani's quote above, and detailed the dynamics he was describing. The collapse of the 2000 bubble would ultimately erase half the value of the S&P 500, and would take the tech-heavy Nasdaq 100 down an implausibly precise 83%.\nThe defining feature of a bubble is inconsistency between expected returns based on price behavior and expected returns based on valuations. If investors pay $150 today for a security that will deliver a single $100 payment a decade from now, but they also fully understand that they'll lose 4% annually on the deal, without extrapolating past gains into the future, then we might say the security is overvalued, and we might question why investors would accept that trade, but we can't call it a bubble.\nBut if investors pay $150 today for that security, because they look back in the rear-view mirror, decide that it \"always goes up\" over time, and convince themselves that expected future returns are always positive, then you've got a bubble. Discounting the future $100 cash flow of the security using any positive expected return would produce a price less than $100. So the positive returns expected by investors are inconsistent with the returns that would equate price with discounted cash flows. The size of the bubble is the fraction of the market price that represents expectational \"hot air.\"\nLikewise, the willingness of investors to embrace \"passive investments\" like ETFs and asset-backed securities based on past performance, with little concern about the valuations, yields, or credit risk of the securities inside, is the very soap from which bubbles repeatedly emerge. Amid the current enthusiasm for special purpose acquisition companies (SPACS), investors might recall the bubble in \"incubators\" at the 2000 peak, the \"conglomerates\" of the late-1960s Go-Go bubble, and even the South Sea Company in the early 1700s, along with similar companies formed at the time \"for carrying on an undertaking of great advantage, but nobody to know what it is.\"\nIf investors price the S&P 500 at levels that are highly likely to produce negative returns for a decade, as they did in 1929 and 2000, and as I believe they are doing at present, yet investors continue to press stock prices higher on the expectation that they will provide historically normal levels of future return regardless of valuations, then you have the sort of inconsistency that defines a bubble.\nLikewise, if the expected return of a conventional passive investment mix is negative on a 10-12 year horizon (based on reliable valuation measures strongly correlated with actual subsequent returns over a century of market history), yet pension return assumptions remain locked near 7% annually, you've got a bubble, and most likely a future pension funding crisis, on your hands.\nThis is how very bad things have repeatedly happened in the financial markets across history, enabled by what Galbraith called \"the extreme brevity of the financial memory.\"\nWhen Modigliani says that there are two \"warranted\" prices, he means that - at least in the short run - there are two ways that prices can fulfill the expectations of investors. In one case, investors have expectations about future returns, and those expectations are informed by the level of valuations. If prices rise, and expected cash flows haven't changed, investors recognize that future returns will be lower. In the \"bubble\" case, investors have high expectations about future returns, mainly based on past returns, and they act on those expectations by driving prices up further. So the expectation of additional price increases is simply reinforced. \"The expectation of growth produces the growth, which confirms the expectation.\"\nOnly one of these prices is consistent, in that the rate of return expected by investors is also the rate of return that equates price with discounted future cash flows. The other price becomes increasingly detached from fundamentals, as a larger and larger fraction of the price represents hot air, and it ultimately collapses as the gap becomes untenably wide.\nDuring speculative segments of the market cycle, there's nothing that forces investors to recognize that higher valuations imply lower returns, or to change their expectation of high returns as far as the eye can see. That, of course, is why we use measures of market internals to gauge the inclination of investors toward speculation or risk-aversion. Valuation provides an enormous amount of information about likely long-term returns and potential market losses over the complete cycle. But valuation isn't a timing tool. In recent years, it hasn't even imposed a limit on speculative recklessness.\nStill, with each price advance, the actual long-term return implied by future cash flows - what investors will ultimately realize as those cash flows are delivered - collapses further, even while investors act on their delusion that long-term returns have nothing to do with price. Eventually, the bulk of the security price represents a bubble component, not the price that would actually need to exist in order for the long-term expectations of investors to be accurate.\nAs I wrote at the 2000 market peak:\n\n \"As long as investors focus on year-to-year returns and not discounted cash flow calculations, the bubble can continue to grow in self-reinforcing fashion. Investors anticipate a high return, and the price behavior reinforces the expectation. The true long-term return becomes increasingly detached from the long-term return imagined by investors, and the bubble component accounts for an increasingly large proportion of the total price.\"\n\nBy our most reliable measures, run-of-the-mill historical valuation norms are roughly 70% below current levels. I know you don't want to believe that.\n\nThe trap door quietly swings open when valuations are extreme and market internals begin to deteriorate. That's the situation we've observed in our measures in recent weeks, with the initial deterioration largely driven by debt securities, but with increasing divergences in equities as well.\nValuations and discounted cash flows\n\n Valuations measure the tradeoff between current prices and a very long-term stream of expected future cash flows. Every useful valuation ratio is just shorthand for that calculation. Every valuation ratio that fails that criterion is inferior, and you can show it in historical data.\n\n\n - John P. Hussman, Ph.D., The Meaning of Valuation, December 2019\n\nI've often noted that the denominator of every good valuation measure is just shorthand for the decades and decades of cash flows that the security is likely to deliver in the future. In fact, we always test the validity of the valuation measures we use by examining:\na) how strongly the resulting valuation measures are correlated with actual subsequent total returns, and;\nb) how closely they replicate an explicit discounted cash flow analysis.\nBelow, we'll examine a variety of valuation measures that offer some perspective on why I view the U.S. equity market as a bubble near the breaking point. Along the way, I'll point out some interesting features of valuations and their relationship with subsequent returns. If math gives you hives, feel free to skim over the small amount that I've included here and there.\nConsider first the relationship between valuations and subsequent returns. I'll state the following, which you can prove to yourself by toying around a bit with present value models: the logarithm of a good valuation measure should have an inverse and roughly linear relationship with the expected subsequent investment return.\nHere's a simple example of what this looks like.\n\nHere's what this looks like for MarketCap/GVA, our most reliable stock market valuation measure\n\nDuring the past three decades, we've studied and introduced a broad range of valuation measures. Most can be calculated back to 1947. Some can be evaluated over a century or more. Across history, even in recent decades, we find that the valuation measures that are best correlated with actual subsequent returns are those with muted sensitivity to cyclical fluctuations in profit margins, and that behave largely like broad, market-wide price/revenue ratios.\nIf we compare our most reliable valuation measures with the valuation measures that one would obtain from a proper long-term discounted cash flow analysis, we find that they're nearly identical. Here's what this comparison looks like for the actual stream of dividends (including the impact of repurchases) delivered by the S&P 500 since 1900, discounted at a fixed 10% rate (see the chart text for additional details).\nThe reason we use a fixed rate of return is that a multiple of 1.0 is then, by definition, the level at which the S&P 500 would have been priced for that particular level of expected return. Any deviation in the valuation multiple from 1.0 then gauges how far likely future returns are from that \"typical\" expected return. We're currently farther away from \"typical\" expected returns than at any moment in history, including the 1929 and 2000 market peaks.\n\nOne of the unfortunate bits of financial illiteracy that Wall Street has pushed into the heads of investors is the idea that extreme valuations are \"justified\" by low interest rates. Now, it's certainly true that holding future cash flows constant, raising the price of an investment will lower the embedded rate of return, and vice versa. If you pay $32 today for $100 a decade from today, you can expect a 12% annual return. If you pay $82 for the same security, you can expect a 2% annual return. If you pay $100 today, you can expect nothing. So it's clearly true that holding future cash flows constant, a lower rate of return implies a higher level of valuation.\nThe reason the statement \"low interest rates justify high valuations\" contributes to financial illiteracy is that the statement has been learned entirely out of context of the arithmetic. As a result, investors seem to imagine that, as long as high valuations can be \"justified,\" stocks can be expected to provide historically normal rates of return in the future. Likewise, investors seem to have no concept that if interest rates are low because growth rates are low, no valuation premium is \"justified\" for stocks, because the lower growth is already sufficient to bring future stock returns down to levels that are commensurate with the low level of interest rates.\nThe truth is simple but uncomfortable. If interest rates are low and expected growth is held constant, higher valuations imply lower long-term returns. If interest rates are low because expected future growth is also low, higher valuations are not required. Long-term returns will be lower anyway. A valuation premium just makes future returns even worse.\nSaying that extremely low interest rates \"justify\" extremely high stock valuations is identical to saying that extremely low future returns on bonds \"justify\" extremely low future returns on stocks. I don't really think that's something Wall Street cares to clarify when it tells investors that stock market valuations are \"justified.\"\nInvestment valuation is concerned with the relationship between three objects: the current price, the future cash flows, and the rate of return that connects the two like a string. The lower the current price and the higher the future cash flows, the steeper the string. The higher the current price and the lower the future cash flows, the flatter the string. Raise the current price above the future cash flows, and the string will point down instead of up. Given any two of these objects, you can calculate the third one.\nFor example, if you want to estimate expected long-term returns, you need two objects: a) the current price and b) the expected stream of future cash flows. A good valuation ratio is just shorthand for those two objects, so you can estimate the long-term return directly from the level of valuation. Then, if you like, you can compare it with the level of interest rates. That comparison can be useful, because even when investors realize that high stock market valuations imply low long-term returns, it's not at all clear that they realize how low long-term return prospects have been driven.\nThe chart below shows our estimate of expected 12-year S&P 500 total returns over-and-above Treasury bond yields, across a century of market history. Compare this with the nearly useless drivel that Wall Street passes off as the \"equity risk premium\" (typically quoted as the S&P 500 forward earnings yield minus the 10-year Treasury yield). Yes, you're reading the chart correctly. Given current valuations, we expect the total return of the S&P 500 to underperform the lowly yield on Treasury bonds by roughly -6% annually over the coming 12-year period.\n\nMany investors confuse the estimation of expected returns (which requires only expected cash flows and the observed price) with a different problem - the estimation of \"fair value.\" See, interest rates come into the picture when you have an expected stream of future cash flows and you want to calculate a \"fair\" current price. In that case, rather than picking an arbitrary rate of return from a hat, it's common to use the level of interest rates, plus some risk premium, as the expected long-term return (or discount rate, or capitalization rate). This sort of calculation can be super-sensitive to arbitrary choices.\nIn particular, Wall Street loves to combine super-high growth rates, super-low discount rates, and super-long time horizons, which allows one to calculate a \"fair value\" that's as close to infinity as possible. The thing to remember is that whatever rate of return an analyst embeds into the fair value calculation is also the long-term rate of return you'll earn over time if you pay that price, assuming the future cash flows are delivered as expected.\nAmong scores of measures we've evaluated or introduced over time, MarketCap/GVA (nonfinancial market capitalization to corporate gross value-added, including our estimate of foreign revenues) has the highest correlation with actual subsequent 10-12 year S&P 500 total returns, followed by our Margin-Adjusted P/E (MAPE). I find it hilarious that the various valuation measures I've introduced over time are sometimes described as products of \"machine learning,\" \"data mining,\" and \"curve fitting\" when they are, in fact, just different versions of an apples-to-apples economy-wide price/revenue ratio.\n\nThe S&P 500 price/revenue ratio and nonfinancial market capitalization to GDP (the \"Buffett indicator\") also perform well, and better than earnings-based alternatives like price/forward earnings, the Fed Model, and even the Shiller CAPE. See, while earnings are necessary to generate long-term cash flows, they are also subject to fluctuations in profit margins (both cyclically and even from decade to decade) that turn out to be highly uninformative.\nEconomically, fluctuations in profit margins are driven primarily by fluctuations in real unit labor costs. Because companies compete on the basis of realized after-tax profits rather than pre-tax profits, changes in tax policy also have far less durable impact on corporate profits than investors seem to imagine. While corporate profits got a tremendous boost last year from CARES spending (the deficit of one sector always emerges as the surplus of another), here's what the relationship between corporate profits and real unit labor costs looks like historically.\nReal unit labor costs are presented on an inverted left scale. The upward pressure on labor costs (observed as a plunge in the blue line) isn't particularly auspicious for future profits. Still, there's so much distortion in recent quarters that I'd consider the jury to be out on how much of this will be sustained.\n\nIt's undoubtedly true that profit margins, expected growth, and other factors have an effect on future deliverable cash flows and the valuations that investors place on stocks. But even for technology stocks, these assumptions should be made explicit and tested against history. You'll find that observable measures like price/revenue are still very serviceable. In fact, the extreme price/revenue multiples of technology stocks helped to inform my March 2000 projection of an 83% loss in that sector.\nAs Benjamin Graham wrote, \"The habit of relating what is paid to what is being offered is an invaluable trait in investment. The more dependent the valuation becomes on anticipations of the future - and the less it is tied to a figure demonstrated by past performance - the more vulnerable it becomes to possible miscalculation and serious error.\"\nThe current 5.19 price/revenue multiple for the Nasdaq 100, is the most extreme level since February 2001, which was followed by a 60% loss in the index (after it had already dropped in half). The situation is actually a bit worse than 2001 here. If one examines the largest components of the index, it becomes clear that their annual growth rates have declined substantially over time.\nAs a result, a dollar of current revenues should arguably command a smaller multiple than a dollar of revenues might have during earlier periods of emerging growth. Yet even if one takes the Nasdaq 100 price/revenue ratio at face value, and even if one restricts attention to the bubble period since 2000, it's difficult to expect the Nasdaq to produce total returns over the coming decade much better than zero.\n\nYou don't really want to see what the same chart looks like for the S&P 500.\n\nThe same is true, unfortunately, for passive investment strategies. We presently estimate negative 12-year average annual total returns for a conventional passive investment mix invested 60% in the S&P 500, 30% in Treasury bonds, and 10% in Treasury bills.\nIn a 2019 white paper, I detailed an approach to estimate a \"value-focused asset allocation\" by jointly considering prevailing stock market valuations and interest rates. It specifies an investment allocation based on which asset class is estimated to have the highest average annual expected return, adjusted for risk, to each point in a long-term investment horizon. That allocation can then be modified by a risk-management component, to adjust the exposure during segments of the market cycle where risk-aversion or speculation among market participants may temporarily drive valuations to depressed or elevated levels. The white paper includes numerous charts showing how this value-focused asset allocation has changed across market history, particularly at important peaks and troughs in the stock and bond markets.\nAlong with those methods, I introduced our \"Endowment/spending multiple,\" which estimates the number of years of spending that a passive 60/30/10 investor requires up-front, in order to finance an expected 36-year stream of future inflation-adjusted spending. The idea here is that in a deeply undervalued market with high expected future returns, investors can finance a future stream of spending with far less than they require when valuations are extreme and prospective returns are low.\nYou know you're in a bubble when funding a 36-year stream of expected inflation-adjusted spending requires over 38 years of money up-front.\n\nIn the chart below, the Endowment/spending multiple is presented on an inverted log scale (left), along with the actual subsequent average annual total return of a 60/30/10 portfolio mix (right scale). Needless to say, we adhere to investment disciplines that are intended to address problems like this.\n\nAs a testament to the breadth of this speculative episode, the median price/revenue ratio of S&P 500 components now exceeds 3.3, easily a record, and extreme enough to provoke distress about the potential losses that innocent but poorly-informed investors may experience over the completion of this cycle.\n\nThe next chart shows the median price/revenue ratio of S&P 500 components sorted into 10 deciles by valuation. The chart is presented on log scale to allow each line to be compared with its own history. Each segment on the vertical axis represents a doubling of valuations. Notice that every single decile of S&P 500 components is at record valuation extremes. Investors now rely on a permanently high plateau in these extremes.\n\nIt's interesting, but far less important, that the median price/earnings ratio of S&P 500 components has also reached 32.4, the highest level in history. This compares with a median P/E of 19.4 at the 2000 market peak. The problem with P/E multiples, of course, is that they are substantially affected by earnings variability. In fact, prior to the current peak, the highest median P/E for the S&P 500 was actually in March 2002, when the index was down 25% from the March 2000 bubble highs. That's because earnings were down far more by then, which boosted P/E ratios. Such is the danger of taking P/E multiples at face value.\nA crude but reasonably effective way to get around the cyclicality of earnings (but only for very broad indices), is to compare prices to the highest level of earnings achieved to-date. I introduced this metric back in 1998 as the price-to-peak-earnings ratio. The chart below shows a version of that. The blue line (left scale) shows the ratio of total U.S. equity market capitalization to GDP. The red line (right scale) shows the ratio of total U.S. equity market capitalization to the highest level of economy-wide U.S. profits to date. Investors have gotten themselves into trouble here.\n\nWhat if valuations remain extreme forever?\nProbably the single most frequent question I've heard from investors over the past couple of years is \"What happens if valuations remain extreme forever?\" It's actually a version of the \"permanently high plateau\" that Irving Fisher disastrously projected in 1929. Still, recent years have produced enormous confidence among investors that the Federal Reserve's purchases of Treasury debt can permanently \"backstop\" the stock market.\nAs I've detailed previously, quantitative easing supports the market only by creating zero interest hot potatoes that are uncomfortable for investors to hold (provided that they're inclined toward speculation), and that are impossible to get rid of in aggregate. Moreover, the Fed's purchases of corporate bonds during the pandemic were legally constrained to CARES funds provided by the Treasury, and ultimately amounted to $14 billion of bonds, in an economy with $11 trillion in corporate debt at $58 trillion in equity market capitalization.\nSuffice it to say that the \"Fed backstop\" is largely in the minds of investors, and relies almost exclusively on the psychological discomfort of holding low-yielding base money. Yet since perception can be indistinguishable from reality, particularly in the short run, it's important to entertain the question.\nMy answer is that if the Fed is indeed able to maintain valuations at the highest levels in history, forever, stock prices would likely grow at roughly the same rate as nominal GDP. So figure 1.6% real structural growth plus 2% inflation gets you 3.6%. Let's call it 4%, which would match nominal GDP growth in both the 10-year and 20-year periods ending at the Q4 2019 economic peak, just before the pandemic. The first casualty of rising inflation is stock valuations, so it's not at all clear that assuming higher inflation would help stocks until valuations were roughly normalized, which would require consumer prices to roughly triple.\nThe chart below is a reminder of how structural real GDP growth has progressed over recent decades (driven by demographic labor force growth and trend productivity), and the basis for that 1.6% figure for structural real GDP growth.\n\nSticking with 4% nominal growth, and adding a 1.5% dividend yield, a \"permanently high plateau\" in market valuations would imply S&P 500 total returns of about 5.5% annually. Again, this assumes that valuations never retreat from levels that presently stand at about 3.6 times their historical norms. Simply allow them to retreat to 2.4 times their historical norms a decade from now - which would still keep valuations among the highest 10% in U.S. history - and the resulting 10-year total return would drop to about 1.3%. I think this would actually be the best-case scenario even in a permanently overvalued world.\nAt elevated valuations, even very small changes in expected return imply enormous changes in prices. So it's unlikely that a period of much higher average valuations will escape the prospect of relatively high volatility. Rather than a 70% market decline, which would presently be required for the S&P 500 to simply touch historically run-of-the-mill valuation norms, investors could expect rather frequent market losses in the 20-35% range, which is essentially what we've seen even over the past few years.\nAll of that would be fine with us. We've adapted our discipline sufficiently (especially in late-2017) to tolerate the possibility of permanently sustained overvaluation. My impression is that the impact of those adaptations has become more evident as we've had greater opportunities to live into them. I can't say that I believe for a second that investors will actually be spared from a 50-70% loss in the S&P 500 in the coming years, but again, it will be fine with us if the market never approaches historical valuation norms again. With the adaptations we introduced in late-2017, our discipline is flexible enough to navigate a bubble even without embracing its premise.\nAn unusual overlap of high-risk conditions\nReturning to Modigliani, my impression is that the advance of recent years to the most extreme valuations in history reflects exactly the bubble dynamics he described, and I expect that it will also end as he described (though not necessarily in one fell swoop): \"The expectation of growth produces the growth, which confirms the expectation; people will buy it because it went up. But once you are convinced that it is not growing anymore, nobody wants to hold a stock because it is overvalued. Everybody wants to get out and it collapses, beyond the fundamentals.\"\nOne of our internal gauges tracks the correlation of market conditions with certain high-risk features that have preceded steep market collapses - a collection of measures capturing valuations, internals, sentiment, leverage, overextension, and yield pressures. Only a handful of instances in history overlap pre-crash conditions as well as they do at present. The current overlap is actually quite similar to August 1987. Meanwhile, the correlation of current conditions with features typically observed at market lows is the most negative in history.\nIf you want my opinion, I suspect that a near-vertical market plunge on the order of 25-35% is coming, probably quite shortly, most likely out of the blue, as in 1987, driven by nothing more than the sudden concerted effort of overextended investors to sell, and the need for a large price adjustment in order to induce scarce buyers to take the other side.\nAs usual, no forecasts are necessary. We'll align our investment stance in response to the valuations and market action that we observe at each point in time. Still, it's of particular concern that these overlaps are occurring in the context of the most extreme valuations in history, along with strikingly dysfunctional pockets of illiquidity in many individual issues. This dysfunctional behavior isn't about any particular video game retailer. I suspect it's actually about some sort of fragility or segmentation in order-flow mechanisms, possibly coupled with poorly managed derivatives exposure.\nAs I used to teach my students, show me a financial debacle, and I'll show you someone who had a leveraged, mismatched position that they were suddenly forced to close into an illiquid market.\nThough my concerns run far beyond the amount of leverage in the system, it isn't helpful that the amount of leverage in the U.S. equity markets is now easily the highest in history. Some observers are inclined to bring this figure down by dividing instead by the market capitalization of equities. But here's some useful arithmetic:\nMargin Debt/GDP = Margin Debt/Market Cap x Market Cap/GDP\nTo say that margin debt to GDP is at the highest level in history is to say not only that stocks are heavily owned on margin, but that those stocks are also breathtakingly overvalued. That combination is particularly worrisome.\n\n All crises have involved debt that, in one fashion or another, has become dangerously out of scale in relation to the underlying means of payment.\n\n\n - John Kenneth Galbraith, A Short History of Financial Euphoria\n\n\nThe kind of event I'm suggesting would not bring valuations anywhere near historical valuation norms. Given current valuation extremes, it would be more like a palate cleanser. I have no particular expectation about what the next dish would be. In the event we do observe an abrupt market collapse, the Fed will undoubtedly respond with some new palliative. Whether or not it is effective will depend on the context of risk-aversion, inflation, credit risk, and other conditions at the time. The larger problem, as we've discussed, is that you can't \"save\" an overvalued asset by propping up its price. The value is in the future cash flows that will be delivered to investors over time. The elevated price only ensures that the long-term return between now and then will be dismal.\nOf course, nothing in our discipline relies on a market plunge. Given the combination of hypervaluation and divergent market internals (largely based on debt securities, but with increasing divergences in equities as well), I do believe that the stock market remains in a \"trap door\" situation. Still, that view will change as market conditions change. We'll refrain from adopting or amplifying a negative outlook if our measures of market internals improve. That discipline has served us well even amid record highs. Again, no forecasts are required, nor does this opinion drive our current investment stance. I just think the correlation with historical pre-crash conditions is worth noting.\n\n How little, it will perhaps be agreed, was either original or otherwise remarkable about this history. Prices driven up on the expectation that they would go up, the expectation realized by the resulting purchases. Then the inevitable reversal of these expectations because of some seemingly damaging event or development or perhaps merely because the supply of intellectually vulnerable buyers was exhausted. Whatever the reason (and it is unimportant), the absolute certainty is that this world ends not with a whimper but with a bang. And so on to the moment of mass disillusion and the crash. This last, it will now be sufficiently evident, never comes gently. It is always accompanied by a desperate and largely unsuccessful effort to get out.\n\n\n - John Kenneth Galbraith on the 1929 collapse\n\nValuations and investment duration\nTo understand why extreme valuations imply high volatility, and require extremely long investment horizons, it's useful to consider the concept of \"duration.\"\nEvery security is a claim on a stream of future cash flows that can be expected to be delivered to the investor over time. While the concept of \"duration\" is most commonly used for bonds, it's actually applicable to any security, no matter how \"lumpy\" the stream of cash flows might be.\nIf you don't like math, feel free to skim over the various equations and just read the pull quotes. I've provided the details just for completeness.\nThe \"duration\" of a security can be defined in two ways.\nInvestment horizon: the weighted average number of years it takes for the security to deliver its payments. For each period, you take the share of total present value represented by that year's payment, and multiply it by the number of years in the future the payment will be received. Add them all up. The result is the number of years from today (a weighted average) that the present value of your investment will be repaid. For example, the duration of a security that delivers a single payment a decade from now is simply 10 years.\nElasticity: the percentage change in the security in response to a change in the underlying gross rate of return. Technically, elasticity is (-dP/P)/(dk/1+k). For example, suppose you own a security that will pay $100 a decade from now, and it's priced at $82.0348, for a 2% annual return. Now assume the expected return moves to 2.01%. The price would drop to $81.9544. Elasticity is (0.0804/82.0348)/(0.0001/1.02) = 10%\nIt turns out that the \"duration\" of a security in years is identical to its \"duration\" in terms of the percentage change of price in response to a 1% fluctuation in expected returns. Duration is also the holding period that \"immunizes\" the investor against changes in expected returns over time. In other words, assuming you reinvest your cash flows over time, duration gives you a good idea of how long you have to hold the security in order for your ending wealth to be largely independent of the fluctuations that the security experiences over that horizon.\nIf you know differentiation, can prove to yourself that the \"modified duration\" of the S&P 500 is essentially the inverse of the dividend yield. The modified duration is just (-dP/P)/dk or Macaulay duration/(1+k).\nConsider P = D/(k-g). Differentiating with respect to k, dP/dk = -P/(k-g), so duration (-dP/P)/dk = P/D.\nHere's how to think about the link between valuations and duration. Presently, the dividend yield of the S&P 500 is 1.48%. If the yield moves to 1.49%, holding dividends constant, prices drop by 1.48/1.49-1 = -0.67% on that 1 basis point move. Duration is just that sensitivity, defined for a 100 basis point move, which would be 67. It turns out that's also the weighted-average number of years from today that you'll receive your present value, if you invest today.\nCompare that to the typical situation over the past century, when the dividend yield of the S&P 500 averaged about 3.7%. At normal valuations, a 1 basis point increase in the dividend yield would produce a price drop of 3.7/3.71-1 = 0.27%, implying a duration of 27 years.\n\n It turns out that the 'duration' of a security in years is identical to its 'duration' in terms of the percentage change of price in response to a 1% fluctuation in expected returns. Duration is also the holding period that 'immunizes' the investor against changes in expected returns over time.\n\nHistorically, investors wishing to match the duration of their investment portfolio to the duration of their investment horizon could be reasonably comfortable holding 100% of their assets in stocks, provided they had an investment horizon of about 25-30 years. Presently, these investors would need an investment horizon closer to 65-70 years. They are currently holding sippy cups.\nScarcity, usefulness, and value\nWhile we're on the subject of bubbles, I'll add a few comments on Bitcoin, just for fun. I'd write more, but my sides still hurt from laughing.\nObjects like tulip bulbs and Bitcoin differ from securities in that they do not deliver a stream of cash flows to the holder. Instead, what objects like tulips and currencies provide is a little stream of services over time, for example, as a perennial thing of beauty or as a means of payment. What people sometimes forget is that it is not just scarcity that defines the value of an object, but the stream of useful \"services\" that it provides (for some reason, nobody wants to buy my unique, limited edition, digitally-signed porcupine seat covers). The price of the object, and the stream of services it provides, should be commensurate.\nU.S. dollars, for example, have value primarily because they are tethered to the real economy by fiat (they legally must be accepted as a means of payment, as noted on the face of any dollar bill), and they represent the entire substrate of the banking system - nearly every payment that goes back and forth in the U.S. economy represents a transfer of base money. Base money (currency and bank reserves) provides billions of little \"services\" over time.\nWith every transaction, reserves move electronically from bank to bank between one account holder and another. That combination of legal fiat and constant use as a substrate of the payments system is what gives money \"value.\" That value also means that the U.S. government essentially obtains revenue as \"seigniorage\" for producing the stuff. For those who imagine that governments are going to surrender that revenue in favor of using Bitcoin, I've got a non-fungible token to sell you.\n\nAs I've noted before, blockchain is a brilliant algorithm, and I expect that it will have a great number of uses for secure transactions and inventory management. Bitcoin, however, is a token generated by an energy-inefficient, replicable blockchain app. Ultimately, its value rests on the capacity to provide transactions services, yet without fiat to require its use, and with strikingly narrow bandwidth - one block of roughly 2000 transactions every 10 minutes - that I expect will prove to be a wildly limiting feature. That's a problem in a world where speculators now value the stock of bitcoin at one-fifth the value of the entire U.S. monetary base.\n\n If you think about how money is valued, it's clear that people accept it because they believe it will provide a claim on the future output of others. Of course, that expectation requires that future producers will also give away their output and accept the money, on the belief that yet other future producers will do the same. That expectation has to continue indefinitely. Like the question 'What holds up Atlas when Atlas holds up the world?' it's not enough to answer that he's standing on a turtle. It's got to be turtles all the way down. The value of money has an enormous psychological component.\"\n\n\n - John P. Hussman, Ph.D., Turtles All the Way Down, February 2019\n\nOf course, Bitcoin may have a certain user base as a vehicle for money laundering and black market transactions, but that's an undesirable investment thesis. The vast majority of transactions are to exchange Bitcoin itself, though the New York Times did recently report that \"pornography, patio furniture, and an at-home coronavirus test are among the odd assortment of goods and services that people are purchasing with the cryptocurrency.\" So, basically, if your typical day consists of surfing porn on your patio while testing yourself for COVID, you're gonna want to look into Bitcoin.\nMy largest concern is that people are actually forking over hard-earned savings in exchange for these tokens, which allows early \"miners\" to cash out. That's essentially the defining feature of a Ponzi scheme. Like all speculative bubbles that rely on increases in price, rather than cash flows generated by the production of value-added goods and services, Bitcoin isn't actually creating \"wealth.\" It's only creating the opportunity for wealth transfer, primarily from those who will end up holding the bag.\nBitcoin has certain characteristics of base money in the sense that it's exchanged on an electronic ledger, but by design, transactions are limited to an average of about 2000 per block, with one block successfully validated, on average, every 10 minutes. In order to validate a transaction block, CPU farms across the world grind out terahashes of random SHA256 validation attempts in order to discover a sufficiently small binary that matches the cryptographic hash of the block.\nAll of this \"mining\" burns up about as much energy as it takes to run a modest-sized country. Validating a block of transactions produces a reward to the miner (and dilution of the coinbase) of 6.25 Bitcoin per block, which currently works out to nearly $200 per transaction. Yet the value of the median transaction in Bitcoin is only about $1000 in the first place.\nThere's a rather primitive regression analysis floating around (tagged as \"sophisticated\" by some observers who apparently go numb at the word \"logarithm\") that attempts to relate the log price of bitcoin to the log \"stock/flow\" ratio, as if it represents some mechanistic supply-demand relationship. Aside from the fact that the correlation between two diagonal lines is always about 0.9-something, I find that one can obtain a better fit just by regressing the log price of Bitcoin on the log ratio of block difficulty/block reward, which is basically a measure of how much energy one needs to waste in order to mine a new bitcoin.\nSo the \"value\" of Bitcoin is partially linked to the backward-looking sunk cost of the energy wasted to mine these tokens. Still, I wouldn't dream of using this sort of \"model\" to trade an object whose \"value\" is primarily in the heads of speculators. Use it if you like. If you happen make money on it, feel free send me a check, preferably in U.S. dollars.\nUndoubtedly, this view of Bitcoin will be unpopular among those who associate holding Bitcoin with superpowers like laser eyes and diamond hands. \"Not surprised Hussman doesn't get Bitcoin. Few do.\" M'kay. Look, there's certainly a case to be made that a speculative mindset creates its own reality, and while it does, there's an opportunity to obtain wealth transfers from frantic late-comers who can no longer tolerate missing out. Tulips gonna tulip. Not my gig, thanks.\n\n In the short run, it will be said to be an attack, motivated by either deficient understanding or uncontrolled envy, of the wonderful process of enrichment. Those involved with the speculation are experiencing an increase in wealth - getting rich or being further enriched. No one wishes to believe that this is fortuitous or undeserved; all wish to think that it is the result of their own superior insight or intuition. As long as they are in, they have a strong pecuniary commitment to belief in the unique personal intelligence that tells them there will be yet more. Accordingly, possession must be associated with some special genius. Speculation buys up, in a very practical way, the intelligence of those involved. Only after the speculative collapse does the truth emerge. What was thought to be unusual acuity turns out to be only a fortuitous and unfortunate association with the assets.\n\n\n - John Kenneth Galbraith, A Brief History of Financial Euphoria","news_type":1,"symbols_score_info":{".IXIC":0.9,".SPX":0.9,".DJI":0.9}},"isVote":1,"tweetType":1,"viewCount":2400,"commentLimit":10,"likeStatus":false,"favoriteStatus":false,"reportStatus":false,"symbols":[],"verified":2,"subType":0,"readableState":1,"langContent":"EN","currentLanguage":"EN","warmUpFlag":false,"orderFlag":false,"shareable":true,"causeOfNotShareable":"","featuresForAnalytics":[],"commentAndTweetFlag":false,"andRepostAutoSelectedFlag":false,"upFlag":false,"length":26,"xxTargetLangEnum":"ORIG"},"commentList":[],"hasMoreComment":false,"orderType":2}

How To Spot A Bubble<blockquote>如何发现气泡</blockquote>

评论（2）

热议股票