You Don't Need a Weatherman

John P. Hussman, Ph.D.
All rights reserved and actively enforced.

"You don't need a weatherman to know which way the wind blows"

- Bob Dylan

Near-term market forecasts are useless. The next rally or decline in the market can't be predicted. Nobody can pick the bottom. Nobody can pick the top. It is impossible to tell where the market will be six months from now. I wholeheartedly believe all of these. The best way to draw a useless answer in an interview is to ask me where the market is headed.

One of the ironies of the financial markets is that the people who believe that the market can't be predicted are usually the ones most eager to make predictions. Watch about two hours of CNBC on any given day to see this principle in action. In some cases, the forecasts are as impressive as a gumball machine horoscope for their exquisite detail, and undoubtedly just as accurate.

It's not necessary to forecast the markets in order to invest well. In fact, except for the projections that we can make about very, very long-term returns based on valuation levels, forecasts are anathema to successful investing. As soon as your investment position is based on a forecast, every market movement thereafter is either a confirmation or a refutation, and you're open to a roller coaster of pride, worry, indecision, and second-guessing. One of our investing principles is that you can sail a boat anywhere you wish without forecasting the wind. What is required is the willingness to align yourself with the prevailing wind, to measure general conditions often, and to change the tack of the sails when you observe a change.

Probability Distributions versus Point Forecasts

But isn't all that talk about aligning ourselves with the prevailing wind just the poetic equivalent of forecasting?

No. The distinction involves the difference between a "probability distribution" and a "point forecast." Fill a few hats full of red and green balls, some with slightly more red balls than green ones, and some with slightly more green than red. There. You've got a set of probability distributions. Try to forecast whether the next ball drawn out of a particular hat will be red or green, and you're making a point forecast. Market timers try to make point forecasts. We just try to identify the probability distribution.

Think of it this way. If you always carry an umbrella in April, regardless of whether it's sunny or rainy that morning, are you predicting the weather? If you generally expect to lose money in Vegas, does that mean that you think dice and roulette wheels are predictable? If you refuse to buy overpriced garbage stocks with bad balance sheets and heavy price/volume distribution, are you forecasting a near-term price decline?

In each case, the issue isn't whether you can predict a particular day's weather, or a specific game's outcome, or a single stock's direction. The issue is whether you can distinguish a favorable probability distribution from an unfavorable one. From the standpoint of rain, April has a different probability distribution than most other months. That certainly doesn't mean you won't have brilliant, lovely days in April. But if you don't like getting soaked, and you're willing to tote an umbrella despite some marvelous days, you'll be well served to carry an umbrella in April. Expecting to lose money in Vegas doesn't mean that the dice won't periodically come up 7 or 11. It's just that you can't predict those particular outcomes, and on average, you'll lose money. Ditto for stock selections. The fact that a stock has low quality in terms of valuation, financial stability and market action does not ensure that the stock won't go straight up. It's just that on average, positive returns load on other, more favorable factors.

At present, stocks are in the most hostile Market Climate we define: unfavorable valuations and unfavorable market action. But that is not a forecast of a near-term decline in stock prices, or any decline, for that matter. Rather, it's a statement that if you were to group together all the historical periods in which this Climate has been in effect, the average result is not good in terms of return and risk. There are certainly some periods in which stocks performed swimmingly. But there are others in which stocks were hammered. Once we've identified a general Climate, we have no additional ability to forecast what the next specific outcome will be. We simply operate on the basis of the market's average return/risk profile in that Climate.

But if the average is "not good," isn't that still a forecast? Well, it might be if we were certain that any particular Climate would persist for a long time. But for all we know, market action might improve enough to shift us to a more favorable Climate a week from now. In the meantime, the possible range of outcomes (roughly +/- 2% on a weekly basis) dwarfs the average outcome (roughly -0.5% on a weekly basis), so we don't even have a statistically significant forecast for the coming week, much less further into the future.

Suffice it to say that we align our investment positions with the Market Climate that we observe at any given time, and shift our position when the evidence indicates that the Climate has shifted. We align our sails with the prevailing wind. We spend a great deal of effort developing proper ways of measuring valuations and the quality of market action in order to identify that wind correctly, but none of those require forecasts. The key is identification.

So while our investment discipline doesn't give us assurance about the near-term direction of the market (which we don't believe is possible), it does give us clear guidance as to the investment position we want to accept at any point in time. You don't need a weatherman to know which way the wind blows.

Market Climate

The Market Climate for stocks remains hostile, with both valuations and market action unfavorable. It's worth repeating that this does not imply any meaningful forecast of upcoming market direction. The reason is, very simply, that the Market Climate could change a week or two from now, perhaps on a market advance rather than a decline, at which point the average return/risk profile of stocks might be favorable once again. We simply have no attachment at all to any forecast of upcoming market direction. When the Climate shifts, so will our position. For now, however, the Strategic Growth Fund is fully hedged against the impact of market fluctuations based on the Market Climate we currently observe. On average, investors have not been rewarded for taking market risk in the current Climate.

In bonds, the Market Climate continues to be characterized by neutral valuations and unfavorable market action. The Strategic Total Return Fund holds a portfolio duration of about 3.25 years, primarily in Treasury Inflation Protected Securities. The Climate for precious metals shares also continues to be favorable here.

Given the persistently low trading volume in the market, there is not much development in market action that conveys strong information. Despite the full awareness of investors that the Fed will move rates by 25 basis points, and a near zero chance that it will move 50, the market seems almost paralyzed in anticipation of the obvious, much like a deer in the headlights. There's no real point in second-guessing the market's reaction, at least from the perspective of our investment discipline. Whatever the outcome of the upcoming Fed meeting, we'll alter our investment position only when we observe a Climate shift, and not otherwise.