A perennial source of debate is whether the U.S. stock market is cheap or expensive. The biggest fights erupt between what I call the business-cycle school and the intrinsic-value school.
The business-cycle perspective is ably summarized by one of my favourite bloggers, Barry Ritholtz, who wrote, "looking at valuations involves two factors: price and profits. One is known, and one is unknown. That is where the economy comes in. Gains in corporate profits depend in large part on accelerating global economic growth."
Ritholtz endorses a model of stock-market pricing in which valuations are determined by near-term earnings, which in turn are set by the business cycle. Under this model, stocks are cheap or expensive based on where the market thinks we are in the business cycle versus where we actually are in the cycle. Valuation requires coming up with a view on the business cycle. This is Wall Street's preferred model.
Right now, many on Wall Street are saying U.S. equities are "reasonable." They cite a mildly above-average price/forward earnings ratio, low inflation, supportive flows into stocks, ultraloose monetary policy, and so forth, all justifying even higher P/E ratios and possibly more earnings growth.
Warren Buffett is an exemplar of the intrinsic-value mode of valuation. In the 
Berkshire Hathaway (BRK.B) owner's manual, he writes that intrinsic value "is the discounted value of the cash that can be taken out of a business during its remaining life." This doesn't sound all that different from the business-cycle model. After all, shouldn't intrinsic value rise during bull markets when earnings are rising and fall during bear markets when earnings are falling?
Perhaps, but not to the extent we see. Under reasonable growth and discount rate assumptions, the stock market's intrinsic value is mostly determined by cash flows more than 10 years out. Next year's earnings could take a huge hit, but given the mean-reverting nature of aggregate corporate earnings, intrinsic value should not be hurt nearly as much. Robert Shiller's seminal 1981 paper is a convincing demonstration that stock market prices are far too volatile in relation to changes in their intrinsic value to be explained by purely rational forces. Animal spirits rule the markets.
Intrinsic-value investors such as Jeremy Grantham and Seth Klarman see asset bubbles everywhere, citing cyclically adjusted valuation measures like Shiller P/E, which averages 10 years of inflation-adjusted earnings. These folks hate the Federal Reserve for "manipulating" or "distorting" market prices.
If there's a key difference between the two schools, it's that the business-cycle approach is just thinly veiled market-timing. It is John Maynard Keynes' beauty contest, where the winner is the best at figuring out what the average opinion of average opinion is. Indeed, pricing models that focus on the next few years are weakly tethered to the notion of intrinsic value. This is not to say that the model is wrong or necessarily inferior to the intrinsic-value approach.
People have different preferences. Investors who mostly care about discounted future cash flows, paying little heed to price movements, sentiment, and the like, are better suited for the intrinsic-value method. Investors who care a lot about relative performance in the short run rationally take the business-cycle approach. I think it's fair to say investors temperamentally suited to intrinsic-value investing are in the minority. Most investors care about how they're doing relative to their peers in the short run and will behave in ways at odds with their private assessments of the riskiness and magnitude of future cash flows in order to avoid sticking out. During the dot-com bubble, the estimable Barton Biggs said, "I'm 100% invested and scared as hell." Indeed, many are forced to care, because they're managing other people's money, and they risk their necks if they're both wrong and alone.
I don't mean to set up intrinsic-value investors as virtuous and business-cycle investors as sinful. While my sympathies lie with intrinsic-value investing, the problem with intrinsic value is, even with a crystal ball, one would not necessarily make money moving in and out of the market. Shiller's work shows the market can be over- or undervalued for many years. An investor in the mid-1970s who correctly foresaw the stock market's exceedingly bright future would have had to endure a grinding decade-long bear market before experiencing sweet vindication. Too bad he'd have lost his job or his nerve long before then.
The myopic business-cycle model is actually an excellent descriptor of how stock markets behave over the short and medium run: Stocks tend to do well when earnings surprises are pleasant and do poorly when they're not, even if earnings are likely to revert to the mean in the long run. An investor can make a killing by being faster and smarter than the crowd in figuring out when the cycle turns. Wall Street operates on this model, which is why one of its favourite ways to gauge market valuation is to compare the price/estimated forward earnings ratio, or P/FE, to its historical average.
Forward earnings are pro-cyclical and optimistic, but they reflect the consensus view on near-term earnings and, by extension, the market's view of where we are in the business cycle. A historically low P/FE can be interpreted as the market being abnormally fearful in relation to near-term earnings prospects. This kind of myopic earnings model, while theoretically weak, seems to work in part because so many investors use it, knowingly or not.
Conventional wisdom holds that market-timing is impossible. However, well-done studies find evidence that some hedge funds and to a much lesser extent some mutual funds have market-timing skill. I'm not talking about being able to predict one-month market returns--studies have almost universally failed to find compelling evidence of that. Rather, certain managers seem to be able to know when the liquidity or volatility outlook is benign or hostile and shift their market exposures accordingly. Manager David Tepper used P/FE to call stock markets extremely cheap in late 2012.
He went as far as to say downside risk was low and the upside was high in part due to a global central bank put and muted inflation. He was right then, as he has been several times in the past, such as when he shifted to 40% cash in late 2007 and shifted back into risky assets in early 2009. The myopic earnings model plus a keen understanding of central-bank posture seems to be a potent combination for timing the market.
We do not see very many intrinsic-value investors timing the market. Why? Because intrinsic value is near-worthless for predicting market movements over horizons shorter than a decade. It is surprisingly hard to come up with a robust valuation-based market-timing rule that generates abnormal returns even in a back-test. On the flip side, it is very easy to find momentum-based timing signals that work in back-tests.
Anyone who makes a habit of leaping in and out of the market based on valuation is going lag the benchmarks for long periods, even if such efforts prove profitable in the end. The most successful intrinsic-value investors are stock-pickers who don't time the market, doing so perhaps once in a generation when valuations become so extreme they must revert within a reasonable window. I believe most successful market-timers play the Keynesian beauty contest. They cannot abide the notion of an idealized fair value toward which the market must gravitate.
Successful market-timers share some similarities:
1) They do not work for Wall Street banks or brokerages or make punditry their calling--they are hedge fund managers (and, rarely, mutual fund managers).
2) They believe central banks have enormous influence over asset prices and therefore never try to fight them without strong reasons.
3) They tend to chase momentum, or at least give it wide berth, all the while knowing that when market expectations are at such odds with near-term fundamentals, they should bet against the trend.
The last part distinguishes the George Soroses of the world from the wannabes.
So, what does this mean in terms of practical investing advice? Defenestrate buy-and-hold and take up macroeconomic forecasting? Hardly. Being a Michael Jordan is possible, but it ain't you. Market-timing can be profitable, but there are only so many market cycles in an investor's lifetime. Even if you successfully time the market on occasion, it's difficult to know whether it's due to luck or skill. Predicting where the market is heading is an exercise in supreme almost certainly unjustified self-confidence.
Most investors should not time the market but rather behave more like intrinsic-value stock-pickers: Hold on to equities like grim death, making adjustments over time against the market's gyrations. They should ignore all the forecasts generated by the Wall Street noise machine, because such forecasts are pro-cyclical and usually wrong over the long run. Moreover, those forecasts are designed for people who must run the myopic earnings surprise race. Individual investors saving for retirement have the luxury of taking the long view--if they can discipline themselves to not be envious of their neighbours.
Besides, for someone playing the short-term pricing game, taking Wall Street's forecasts seriously puts him behind the curve. After all, the Keynesian beauty contest is about being a superior judge of the consensus' view of the consensus, not simply being part of it. The main reason to consume public research from Wall Street is to identify consensus positions and see where they diverge most from one's private assessments.
Because the costs of market-timing are enormous and potentially fatal to one's retirement plans, investors who have substantial assets to lose should not play the game. Young investors with a long runway of lifetime earnings ahead of them can afford to play the game and learn over time whether they're any good at it without killing their retirements. They should, however, understand that they're likely to be on the losing end and therefore size their positions appropriately.
