Start with a basic method for valuing stocks. To grossly oversimplify, what you pay for a share in a company is based (in large part) on a formula called “earnings multiple.” How much you should pay today is a function of how much the company is likely to earn next year. The Wall Street analysts who create those earnings forecasts put a multiple on it, lets say 15X and — voila! — we get a fair estimate for what prices stocks should be next year.
What could possibly go wrong with that? Plenty. Three major things: estimates, multiples and forecasting error.
Estimates: Not only can they be wrong, they often are very wrong — and in a consistent and predictable way. According to a study by McKinsey & Co., Wall Street’s analysts are nearly always too bullish. The authors noted that analysts have been “persistently over-optimistic for the past 25 years.” Their earnings estimates ranged from 10 to 12 percent a year. Actual earnings growth was half of that, about 6 percent annually. “On average, analysts’ forecasts have been almost 100 percent too high.” The only exceptions are during recessions, when they are (surprise!) too bearish.
Multiples: We could use an earnings multiple of 15x — about the average over the past century. However, at times this multiple swings to extremes, sometimes wildly so. Sometimes, investors think markets are worth more than usual — 20x, 25x, even 30x. Other times, they are willing to pay much less for a dollar of earnings — 12x, 10x even 7x earnings. History teaches us that these multiples eventually revert to the mean.
That is for the overall market. Different sectors command different multiples, and even companies within the same sector can trade at different earnings multiples. That multiple can be a function of popularity, media coverage, a celebrity CEO or a hot new product. But buzz and other trendy factors are not the best basis for determining the intrinsic value of equities.
So we see the first two legs of our “simple formula” are not so simple after all. Earnings estimates are not very reliable, while multiples vary widely. What else could go wrong? One big thing, and it could be the source of your queasiness.
Forecasting error: As it turns out, whenever the economy slips into a recession, these earnings estimates have their greatest error factor. In other words, those estimates that were not so great in the first place turn out to be simply terrible as the economy turns. Wall Street gets it most wrong at precisely the worst possible time.
Consider a study from James Bianco of Bianco Research. It looks at two groups of earnings estimates, one from Wall Street strategists who look at the macro picture, and, second, the Street’s analysts who study individual firms. According to most recent recession earnings forecasts:
●1990-91 recession: Strategists (top-down forecasters) were too optimistic by 10 percent. Analysts (bottom–up forecasters) who added up the 500 company forecasts, were too optimistic by 25 percent.
●2001 recession: Strategists and analysts too optimistic by 23 and 25 percent respectively.
●2007-09 “Great Recession”: Strategists and analysts too optimistic by 40 percent.
As Bianco notes, both strategists and analysts are getting significantly worse at predicting future earnings.
Doug Kass, who manages the Seabreeze hedge fund, thinks other factors are at work. Kass, who has been critical of Wall Street analysts, points to distinctions between cycles. “If you go back over the last eleven recessions — the first nine all had 5 to 15 percent profit drops. Perhaps the recession of 2001 was an outlier, caused by the bursting of the dotcom bubble’s sky-high P/E multiples. The 2008 recession was even more unusual — with unprecedented leverage in real estate, banking and in corporations.”
Kass’s points are well taken, but I suspect the error is about something else. Structural changes at Wall Street firms are just as likely to be the cause. Research analysts used to work for trading and asset management divisions of big Wall Street banks. Since the 1990s, they have mostly migrated to underwriting. That’s where all the money is made.
This change has changed the job of the analyst. They do far less critical analysis and far more “cheerleading.” Robert Powell, editor of Retirement Weekly, confirms it: Regarding the stocks that make up the S&P1500, Powell noted that not a single one has a Wall Street consensus “sell” rating on it. This is pretty damning proof that forecasting errors may be because of inherent structural bias.
Given all that, what might the volatilitymean for markets?
I think about day-to-day market action as a conversation between investors holding different expectations about the future. Their range of opinions affects their outlooks about earnings and prices. Some believe the recent deceleration is only temporary and the economy is poised to take off again. Some investors believe we will avoid a recession, but only just barely. More bearish traders believe a regular cyclical recession is inevitable. The most negative outlook is for a nasty recession, with rising unemployment, more foreclosures and another leg down in housing.
On any single day, market action is just noise. Over time, as more and more data accumulate, this “conversation” becomes more intense. The markets eventually gather enough information to determine who is correct, and prices adjust accordingly.
The Street’s earnings forecast for the S&P 500-stock index is about $95 for 2011, and about $112 for 2012. If the bulls are right and the economy rebounds, stocks could be undervalued by 30 to 40 percent. If the bears are right, earnings estimates might too high by as much as 30 to 40 percent, and a recession-driven earnings contraction means lower equity prices.
My views are in between. At the least, I expect more slowing in the economy, reflected in weaker earnings and lower prices. However, I do not have enough information to see around the corner to how severe or mild the slowdown will be.
As we have suggested before, you should take a proactive approach that recognizes these various possibilities and plan accordingly. That means holding a decent amount of cash and bonds (I am at 50 percent) so you can ride out the downturn and have some buying power when things get cheap.
Ritholtz is chief executive of FusionIQ, a quantitative research firm. He is the author of “Bailout Nation” and runs a finance blog, The Big Picture.