What a Bunch of Balderdash

July 28, 2008

I apologize ahead of time if this post is a bit intemperate.

Buried around a truism, the New York Times has produced a misleading and rather silly piece on the value of “predictions.”

The thrust of the piece is that predicting the markets and the economy, especially in the short term, is fraught with peril. True enough. That is why whenever giving our view of the future it is in terms of the risks one takes relative to the returns one might get. Hence the name of this site.

However, some things can be said, and often with a great deal of confidence. Timing is an issue, the magnitude can cover quite a bit of ground, but things can still be said. The problem for investors is that in a world filled with conflicts of interest and myopia how do you to tell which views on those tradeoffs to which one should pay attention? It is not that “predictions” (which is a term they use loosely) are to be ignored.

Anyway, that is the stated context for the article. The subtext however is the denigration of “outlandish” predictions, by which they mean those emanating from us negative nellies. Thus, the somewhat cautious claim of Bill Gross of a trillion dollars in losses related to the credit crisis is trivialized:

But despite this decidedly mixed track record, forecasters still enjoy a rapt audience, particularly at a moment when so much in the markets depends on the uncertain course of the housing market and the broader economy.

At investment and commercial banks, losses tied to bad mortgage investments, which now exceed $450 billion, are certain to rise further if home prices continue to decline and more
people default on their mortgages. Last week, Bill Gross, a prominent bond fund manager, offered another forecast for the final bill: $1trillion. Some market watchers say the figure could be even higher.

Obviously we will not know for sure until some time down the road, but it is interesting that Gross, as well as a number of others who have been hooted down for their alarmism, have consistently been proven to be too optimistic. The 450 billion dollar figure which everyone recognizes will go dramatically higher (that is math, not opinion) is already dramatically higher than we have been repeatedly told would be the most the losses would be.

Our firm, as well as a number of others, are not band wagoning here. We “predicted” that we would be fortunate if the losses didn’t reach a trillion or more. No precision, no claim to knowledge we didn’t have, just a little simple math which showed it would be “bigger than a breadbox.

Despite that obvious track record which shows that some people saw at least the general outline, if not the exact shape of things, we get this kind of piffle:

Ideally, predictions on companies and stocks would be “thoughtful, nontheatrical forecasts that take a look at long-term fundamentals,” said Abby Joseph Cohen, a longtime strategist at Goldman Sachs.

But, Ms. Cohen added, less dramatic forecasts rarely make headlines. “If what is being provided to viewers and readers are these theatrical forecasts, that is what many people will pay attention to because that’s what they have available.”

Abby may be a wonderful person to have over to dinner, and a dear friend to have, but she is also someone who has led more investors over a cliff than any major strategist of the last ten years. Abby has been the outlandish one, as anyone who was unfortunate enough to listen to her in 1999 and 2000. I am sure she views the thoughts of those who have consistently stressed the risk levels and overvaluation of stocks over the last 10-12 years as being outlandish in their claims. However, look at the returns of the market over the last 10+ years (10 year return for the S&P500: 2.55%.) Pitiful.

“We have gone from an abnormally calm period, and we’ve blown right through normal volatility,” Ms. Cohen said. “We are in an exceptionally volatile period.”

Does anyone fact check at the Times? That is an outlandish claim. It is most decidedly not true. For those who wish to get a more accurate description of volatility today versus history, the essential Ed Easterling of Crestmont Research will bring you up to speed (pdf.) Listen to Abby at your peril.

This however tales the cake. First, the truth:

Another source of investment guidance used to come from research analysts, who try to predict quarterly earnings at companies. But there is a great deal of guesswork involved here, too. Analysts correctly predict earnings only a fifth of the time. Nearly two-thirds of quarterly earnings beat estimates, and the rest come in too low, according to data from Thomson Reuters. Many companies, of course, try to defuse overly optimistic forecasts to manage investors’ expectations and deliver “better-than-expected” results.

Then this nonsense:

This year, Wall Street’s crystal balls have performed even worse than in the past. As earnings season for the second quarter winds down, 67 percent of companies reported earnings higher than what analysts had predicted, and 22 percent reported earnings that were worse. Only 10 percent of companies matched analysts’ expectations.

We go from a decent point about how wrong analysts are to this misleading………

Okay, I will attempt not to get steamed. The problem with analysts is they are consistently too bullish, and by huge amounts. Exactly the opposite of what that statement implies. Yes, companies consistently beat the estimates when they are finally reported, but only after analysts slash those estimates almost down to the day the reports are made. This year companies have consistently failed to make estimates that were made just weeks earlier. Compared to what was “predicted” 3, 6 or twelve months ago the earnings have been dramatically lower. There is a problem with analyst predictions, and investors should be warned. However, this just muddies the issue of what those problems are. Conflicts of interest, behavioral biases, being too optimistic about the particular industries they cover (even when not conflicted) etc. It is not that they are too cautious about earnings prospects.

Then, to add insult to injury, we get this:

Poor predictions are nothing new in the financial world: in 1999, a pair of prognosticators — James K. Glassman and Kevin A. Hassett —published a book titled “Dow 36,000”; the blue-chip index closed last week at 11,370.

But investors seeking light in a dark period may just have to stick with no one’s predictions but their own.

That helps investors. We get evidence that some of the silliest predictions of the past were wrong; a false implication that analysts are often wrong, but too bearish; a plea from one of the most disastrously bullish strategists of our era not to listen to outlandish claims from people who have been proven correct; and people who make valid points (such as Taleb) are placed in a context with which I am sure they are uncomfortable; all in the service of a message that no one’s advice is worth taking. Never mind that most of the greatest investors of the past half century warned repeatedly that returns were likely to be unsatisfactory, regardless of this present crisis. Never mind that many, if not most, of those same voices warned repeatedly that the housing market was set for a fall, that a credit crisis was likely imminent, and on and on.

No, we shouldn’t listen to anybody.

Since the Times doesn’t want to tell you this, or examine it, I’ll jot down a few things to keep in mind.

  • Analysts are almost always too bullish.
  • As one of her sources, Taleb, would likely agree, economic models based on assumptions of equilibrium (essentially all of them) unsurprisingly almost always (until the horse is out of the barn) predict some sort of equilibrium, or at least a tendency to move back towards it. Thus economic activity is often well outside of their predictions. More importantly, “fat tail,” or extreme, events occur at rates far more frequent than their models assume. There are literally dozens of market and economic events which have occurred over the last century, when statistically (according to the assumptions of modern finance and economics) it was unlikely even one would happen.
  • The real returns from stocks are much lower than most market commenters and professionals realize.
  • You can with a fair degree of confidence predict most asset class returns over 5-10 years of time. I said with a fair degree, not certainty. Unfortunately most attempts to do so use unrealistic assumptions or historic returns. The first problem is obvious, but requires knowledge of what is reasonable. Most people use the latter, historic returns, as their guide. Big mistake.
  • Also, somewhat off topic, though important nonetheless, stocks are not cheap yet, but they may still go up.
  • I think it is unlikely they will not go lower from here.
  • For Clients: If stocks do go up smartly and we don’t do as well as we could have, we have had a pretty darn good 12 months (or five years for that matter.) Trailing a bit now would hardly be a crisis, and as last year proved, we may not trail even then.

Thanks for visiting Risk and Return. Please feel free to contact us with any questions and/or comments. Please note our disclaimer.

Tags: , , , , , , , ,

Leave a Reply

Your email address will not be published. Required fields are marked *


Thompson Creek
"Your Partners for a Life Lived Well"

Private Client Fee Only Advice

Click here


Email Newsletter icon, E-mail Newsletter icon, Email List icon, E-mail List icon Sign up for our Email Newsletter- The View From The Bluff
For Email Marketing you can trust
View Lance Paddock's profile on LinkedIn