华尔街说股票很便宜，傻子才信

眼下，股市记录似乎每天都在刷新，但华尔街却向投资者发出的信息却是背道而驰。这些啦啦队长们称，从某些方面来看，股票实际上并不贵。但是，我们一定得小心那些利用远期盈利预测的投机商。

对于那些仍鼓励购买股票的市场策略师和专家来说，要证明眼下高高在上、甚至貌似过高的估值实际上并没有那么高，并非是件易事。远期市盈率也就应运而生。这一大受欢迎的指标的计算方法是：用未来4个季度预测的标普500的每股收益平均预估值，除以标普当前的价格。这些未来收益预估过于乐观，几乎不约而同地预测了利润的大幅提升。但综合来看，这一方法所得出的“远期”价格/收益市盈率（P/E）实际上总是比按照实际账面结果（也就是过去4个季度的实际营收）计算的市盈率低得多。

在12月27日接受CNBC采访时，瑞士信贷首席市场策略师乔纳森·格勒布曾预测，标普500将在2018年底达到3000点大关（较采访之日的价格将上涨12.2%），而他的依据便是远期市盈率。格勒布说：“让我们按照大多数投资者的方式，从远期的角度来看待市盈率。在税改法案出台之前，股票的市盈率是18.25。也就是说，公司每挣1美元，人们就会自愿掏18美元或更多的资金来购买该公司的股票。如果收益的增幅与我的预期一致，那么市盈率倍数就会降低，因此我觉得人们的购买价格其实相当于17倍的远期盈利。”

在大多数投资者看来，市盈率前景低于20会让他们感到欣慰不已。但是这一看似合理的远期数字颇具误导性。为了弄明白其中缘由，不妨看看以下两个因素。首先，让我们把如今的远期市盈率与它的同类——长期远期市盈率进行对比，而不是与过去的市盈率对比。此外，我们还需发问：哪种类型的收益预测必须低于20倍市盈率，这种预测合理吗？

价格创历史新高——过去或者远期

我们有必要强调，按照过去4个季度的GAAP（美国公认会计准则）收益计算，当前的股票市盈率达到了惊人的24，较1990年以来的平均值19高出了26%，比1888年以来的平均值16.7高出了44%。耶鲁大学经济学家罗伯特·席勒开发的CAPE市盈率更是达到了令人生畏的33.3，仅次于21世纪初的科技泡沫期间的股市估值。

在CNBC采访中，格勒布似乎根据运营收益来计算远期市盈率。然而，使用GAAP收益计算的远期预估值，虽然略低，但也采用了类似的方法。这又让我们回到了第一个问题：如果对比当今的远期估值与过去的预测，结果会如何？按照历史标准来看，远期市盈率是低了还是高了？

标普Global Market Intelligence称，分析师根据GAAP利润计算的四个季度（以1月8日周一为截点）远期市盈率平均预估值为19.25。从这一数字来看，股票的价格是十分合理的，不是吗？并不一定。自2003年初以来，每年1月8日的平均远期预测为15.7。因此当前19.25的估值高于16年以来的平均值。即便我们将2000-2002年（当时科技股的狂热仍在推动预测不断膨胀）的超高预测值计算在内，这一数字仍将高出平均值13%。

第二，回答以下问题会让我们得到启发：为了让远期市盈率低于20，那么利润预计应上涨多少？以2016年第三季度（标普500依据报告的营收的最后整个季度）为截点，过去12个月的GAAP每股收益总计达到了107.08美元。以2018年第四季度末为截点，分析师预测前四个季度每股盈利总计将达到136.75。涨幅接近28%。

别忘了，这一惊艳的业绩也只会将远期市盈率倍数降至19.25，仍然远高于该数据的平均值。这也无法让股票成为香饽饽。这种情况会发生吗？不大可能。截至第三季度，标普500的运营利润率为10.7%，创7年以来的新高，是标普历史上最辉煌的数字之一。即便美国去年年底已然通过了大幅减税法案，但28%的运营利润率增幅却需要空前绝后的业绩作为支撑。

为了判断股票价格是否过高还是处于合理区间，我们应该去看看股市的支撑点，例如过去的业绩，GAAP数据，或Shiller市盈率或CAPE这类更好的参考标准。追随远期市盈率无异于梦游幻境。（财富中文网）

译者：冯丰

审稿：夏林

At a time when the stock market seems to set new records every day, Wall Street is sending a counterintuitive message to investors: By some measures, the cheerleaders say, stocks are actually cheap. But beware of bulls bearing forward earnings projections.

For market strategists and pundits arguing that stocks are still a buy, it’s a major challenge to prove that today’s lofty, even seemingly excessive, valuations aren’t really so lofty. That’s where forward earnings come in. This much-favored metric involves taking analysts’ consensus estimates for the S&P 500’s earnings-per-share, as predicted for the next four quarters, and dividing that number by the S&P’s current price. Those future-earnings estimates skew bluebird optimistic, almost invariably pointing to big increases in profits. Put it all together, and that methodology establishes a “forward” price-to-earnings multiple (P/E) that’s virtually always far lower, and a lot more fetching, than the P/E based on results already on the books—in other words, the actual earnings recorded over the past four quarters.

In an interview with CNBC on Dec. 27, Jonathan Golub, chief market strategist for Credit Suisse, cited the forward P/E to justify his prediction that the S&P 500 would reach 3000 by the close of 2018 (a 12.2% increase from where prices stood that day). “Let’s look at multiples the way most investors do on a forward basis,” declared Golub. “We were looking before the tax plan at an 18-and-a-quarter multiple for stocks. For every dollar you earn you’re willing to pay 18 or more dollars to buy that company. If earnings go up as much as I expect, the multiple gets cheaper, so I think that you’re probably on the actual earnings paying something like 17 forward.”

To the ears of most investors, the prospect of P/Es well below 20 are comforting. But the reasonable-sounding forward numbers are highly misleading. To understand why, let’s examine two factors. First, let’s look at how today’s forward P/E compares not with trailing P/Es, but with its own species, forward P/Es over long periods. And second, we should ask: What kind of earnings projections are required to arrive at a sub-20 P/E, and are they reasonable?

Historically high prices—backward or forward

It’s important to emphasize that the current stock multiple, based on trailing, four-quarter GAAP earnings, is a daunting 24. That’s 26% higher than the average of 19 since 1990, and 44% above the figure of 16.7 since 1888. The CAPE P/E developed by Yale economist Robert Shiller is even more forbidding at 33.3, a number surpassed only by valuations during the tech bubble of the early 2000s.

In the CNBC interview, Golub appears to be referring to forward P/Es based on operating earnings. But the forward estimates using GAAP earnings, though slightly lower, follow a similar pattern. Which brings us to the first point: How do today’s forward estimates compare with past projections? Is the forward P/E low or high by historical standards?

According to S&P Global Market Intelligence, the analysts’ consensus estimate of four-quarter forward P/E based on GAAP profits stood at 19.25 as of Monday, Jan. 8. Makes stocks sound reasonably priced, right? Not necessarily. Since early 2003, the average forward forecast for Jan. 8 of each year was 15.7. So the current estimate of 19.25 is 22% higher than the sixteen-year average. Even if we include the gigantic forecasts from 2000 to 2002, when the tech craze was still inflating projections, the current number is 13% above average.

Second, it’s instructive to answer this question: Just how much are profits expected to surge in order to push the forward P/E below 20? As of Q3 2016, the last full quarter of reported earnings for the S&P 500, GAAP trailing 12-month earnings per share (EPS), in the aggregate, totaled $107.08. By the end of Q4 2018, analysts predict total, four-quarter trailing EPS of $136.75. That would be a rise of almost 28%.

And remember, that epic result would only lower the forward P/E to 19.25, still much higher than the average for that figure. That hardly tags stocks as a bargain. Will it happen? Unlikely. As of Q3, S&P 500 operating margins stood at 10.7%, the highest level of seven years, and one of the most elevated figures in S&P history. Even in the wake of the huge tax cut passed late last year, boosting that operating-margin figure by 28% would require almost unprecedented performance.

To judge whether stocks are extremely expensive or reasonably valued, trust the stalwarts, like the trailing, GAAP numbers, or even better, the Shiller P/E or CAPE. Following forward multiples is a trip to fantasyland.