Stock Market Expections – A Lesson in Rationalism

There’s a saying among economics types: The key to success in forecasting is to either give ‘em a date or give ‘em a number, but never give ‘em both.
As a money manager, it is my job to form expectations about things. Visitors to my office and readers of my columns often ask me for my forecast for the markets. I respond that I don’t really have a forecast. This response seems at odds with what I just said about expectations, so I’ll try to explain.
There is an important difference between a forecast and an expectation. Each has to do with stating an estimate of some future event or condition. The difference is subtle, and has to do with the time frame involved, and with the precision of the estimate.
The weather is a good example. If you were planning a California wedding for August of next summer, you have an expectation about the weather. Based on historical records and personal experience, you would quite rationally expect the day to be nice. You can have a high enough confidence about your expectation to go ahead and book an outdoor location, without bothering to rent a tent. This is the nature of an expectation: while it is far from a sure thing, you have great confidence in the general pattern of things. The averages are going to work in your favor.
As the wedding day grows closer, it is time to watch the weather forecast more closely. If the wedding is on a Saturday, and Thursday night’s weather forecast is for an unusually high chance of rain, you might decide to rent that tent at the last minute.
When making any investment decision, we need to have expectations. If you don’t expect a particular investment to pay some given rate of return, then how can you possibly make the investment? Investment expectations should be rooted in the time period over which the investment will be held. A buyer of stocks should ponder long time horizons where the final result is a bit uncertain. That does not relieve the stock investor of the duty to form an expectation.
I develop clearly stated expectations for the broad stock market over the next five to seven years. My expectations are rooted in my study of long-range pressures and patterns in the economy and in the stock market itself. I expect that the stock market will be as volatile as ever. I expect that the price-earnings multiple will fall over the next five to seven years. I expect another recession sometime in the next decade [note: This piece was written in mid-2003]. I do not, however, have a forecast for what is going to happen tomorrow or over the next year. Nor do I have a forecast for exactly when any of my expectations will come about. Give ‘em a number or give ‘em a date, but don’t give ‘em both.
There is more than one right way to go about forming expectations for the market. I like my way, but it’s certainly not the only way. But I will say that one particular method is almost certainly the wrong way to form expectations. Unlike the weather, where past patterns are proven to occur with regularity, the stock market and the economy behave quite differently. We can make increasingly accurate forecasts about the weather as the time period shrinks. The opposite is true for the stock market.
The wrong way to form future expectations about investments is to look solely at what they have paid in the past. I can give you one shining example of the folly of expecting the past to repeat itself. Over the past couple of decades, rates on one-year certificates of deposit have averaged about 5%. Go down to your bank and ask them for one of those 5% one-year CDs. Please write to me and tell me whether they laughed or merely looked at you as if you were crazy. Certificates of deposit only pay 0.25 to 0.50% today. What they paid in the past is of no concern whatsoever to your future expectations.
You might think that, while my example applies to bank deposits, it surely doesn’t apply to the stock market. Well, it most certainly does. Those double-digit average returns we received since 1981 have no bearing whatsoever on what we should expect from the future.
You do not need to have a forecast for the stock market. In the short run, anything can happen. But over your holding period of seven or ten or thirty years, you most surely should form an expectation. In contrast to the wedding planning process, simple historical averages are the wrong way to do it.
Intermediate-term high-grade bonds have paid an annualized return of about 5.5% over the past 10 years*. Should you expect to earn 5.5% over the next 10 years? Certainly not – that same bond portfolio is only paying about 2.9% right now, so that’s what you should expect to earn. If you expect to earn more than today’s 2.9% on bonds, then you have formed an irrational expectation.
It’s easy to form rational expectations for bonds. You can look up the yields in the newspaper every day.
While it is easy to develop an expectation for bonds and other forms of interest-bearing investments, stocks are a bit trickier. Stocks don’t have a stated interest rate printed on them. Your returns are largely at the mercy of the markets and of the companies you own.
There is one simple little calculation that should serve as the starting point to form your stock expectations. Over time, the U.S. economy grows at about 2-3%, plus inflation. We call the 2-3% growth the real rate of growth, since it is what’s left after we take out the corrosive effect of inflation. Realize for a moment that the 3000 or so public companies that make up the bulk of the stock market can only grow their earnings as fast as the whole economy grows. This is an elementary concept, since corporate America cannot grow to become larger than the American economy itself.
It never fails to humor and amaze me when I read some Wall Street prognosticator predict an infinitely-long period of corporate earnings growth that exceeds the growth rate of the entire economy. As Warren Buffet said, “Trees don’t grow to the sky.”
So, if the economy grows at 2-3% plus inflation, how fast does the value of corporate America grow? Right! The total value of the stock market can only grow at a rate of 2-3% plus inflation. In reality, it grows closer to 1.5% than 2-3% because a significant share of our total economic growth is generated by faster-growing privately held businesses, not public companies.
In addition, corporate managers find a way to continually dilute the shares held by the public, so we really don’t get the full 2+% on a per share basis. (For example: Since it went public in 1980, Apple, Inc. has nearly doubled its number of split-adjusted shares outstanding – diluting the original shareholders at a 3.3% annual clip.)
At any given moment in time, we should only rationally expect our stock values to grow by about 2% plus inflation. If inflation is 3%, then our stocks should grow at a 5% annual rate. Fortunately, this is not the end of the story. We also get another source of return from stocks – dividends. At present, the net dividend yield is around 2%, giving us a total expected return on stocks of 7%, assuming inflation is 3%.
Any investor, stockbroker or money manager that does not accept this fundamental truth about stocks should be forced to march up and down Wall Street with a sign around their neck that says, “I can’t add.”
Now, compare that 7% number to what the historic charts tell you to expect from stocks. Stock returns in the past few decades have been much higher than 7%, and it makes sense to ask why. First, dividends were far higher in the past, adding as much as 7% per year to our starting point of 2% before inflation. In addition, inflation was higher and added to gross returns. But the single biggest reason that stocks paid double-digit returns in the recent past is the fact that the price-earnings multiple rose dramatically, from less than 10 in 1981 to over 22 today.
The extreme expansion in the price-earnings multiple was a singular accident of history that will almost certainly not be repeated. Realize that, to earn more than 7% on stocks indefinitely, the price-earnings multiple must grow without limit. Such growth is, quite literally, impossible. Trees don’t grow to the sky.
Achieving the same returns over the next 30 years that we got over the past 30 years will require the price-earnings multiple to expand to over 40. While that might happen, it is quite unlikely and should not be part of our expectations. Amazingly, most public pension funds assume that stocks will pay 10% or more indefinitely. Why? Because they paid that much over the past three decades, that’s why. It’s like walking into a bank and demanding a 5% CD – and expecting them to actually give you one!
Starting from a rational expectation of only 7% today, we can rightly ask whether there are pressures pushing stocks above or below that figure. In 1981, the pressure was upwards since the price-earnings multiple was very low. Today, the price-earnings multiple is 50% above its long-term average, and the pressure on our 7% starting point is decidedly downward.
It is not a simple business to form a reliable expectation for stock returns. In the face of uncertainty, we need to go back to first principles and add up the numbers. We can adjust those numbers a little if we think there are compelling reasons to do so. The naïve and simplistic way to form an expectation is to merely look up historic returns and assume that they will happen again. Market history can teach us some lessons, but it does not always repeat itself.
*Vanguard Total Bond Market Fund 3/27/02 to 3/27/12