Is It Time to Buy Stocks Again?

Over the past few weeks, I've gotten several versions of the following question from readers:

Stocks finally look like they're headed up now. Should I switch out of bonds and into stocks, or do nothing?

To me, the answer is clear: Neither option is attractive unless inflation declines further and stays low for the next several decades, earnings grow at least as fast as they did in the 1990s, and geopolitical issues remain benign. The odds of all three things happening are comparable to drawing 21 three times in a row in a game of blackjack.

Inflation
In this interest rate environment, it's easy to make an argument that stocks are cheap relative to bonds. You can do that by converting nominal (pre-inflation) yields into real (after-inflation) yields on the different asset classes. (For this exercise, I'll use the 2003 year-to-date inflation rate of 1.2%, as reported on InflationData.com.) By doing that, you can clearly see that stocks represent the best relative value among the big three liquid investments (stocks, bonds, and cash.)

For example, long-term government bonds yield 3% to 4% today, in nominal terms. This works out to a 2% to 3% real (post-inflation) return. So if you invest in government bonds, you can expect your purchasing power to grow by about 2% to 3% if you hold your bonds until maturity. However, in a classic case of government policy diverging from economic reality, you will be taxed on your nominal income as though inflation didn't exist. If you factor this drawback into the bond yield equation, the real return you can expect on government bonds is close to 0%. Thus, even if you have the patience to hold a long-term bond until maturity, the best you can hope for is to maintain, not increase, your future purchasing power.

Of course, you could just hold cash. But at today's rates, cash (i.e., a money market fund) yields somewhere in the vicinity of 0.6% in nominal terms, or negative 0.6% in real terms. Take out the management fees and taxes you'll pay on your income, and the real return gets even worse. In fact, if you're a taxable investor, an argument could be made that $1 in a money market fund is worth about as much as $1 stuffed under your mattress or held in a no-fee checking account. After all, you don't pay taxes or management fees on those.

And that brings us to stocks. At today's market P/E of 32, the earnings yield on stocks is about 3.1% (100% divided by 32 = 3.1%). There's no need to convert this 3.1% into real terms because it's already in real terms. That's because, over the long term, earnings growth should (theoretically) be correlated with inflation; if inflation rises, corporate asset values and earnings will increase in tandem.

Further, with stocks you have the opportunity to defer taxes until a time of your own choosing. You can sell a stock today and pay taxes next year, or you can hold the stock for 10 years and defer your taxes until then. This tax-deferral choice is valuable.

Conclusion: With a 3.1% real earnings yield and the opportunity to defer taxes indefinitely, stocks seem like a better bet than bonds do these days.

Earnings Growth
You should also keep in mind that government bonds don't increase their payouts year after year, while stocks do. When you buy a bond, you get a fixed payout each year until maturity, at which point you get your original principal back. When you buy stocks, however, the yield on your original investment will grow over time.

According to the book Stocks for the Long Run by Jeremy Siegel, U.S. companies' earnings have grown at an inflation-adjusted rate of about 1.25% annually over the past 130 years. Since World War II, the rate has been higher at just over 2%. If we assume that real earnings growth will continue at 2% over the next few decades, this means that, in addition to the 3.1% earnings yield on stocks today, investors can (again, theoretically) expect an additional 2% total return from stocks annually, after inflation. In other words, if P/E multiples stay where they are, stocks should allow investors to increase their pretax purchasing power by roughly 5.1% annually over the next few decades, starting from today.

Conclusion: Relative to bonds, stocks get even more attractive when you factor in future earnings growth.

P/E Expansion
Okay, so we've established that stocks probably represent a better long-term value than bonds, but we've ignored something. Just because stocks are a better relative value doesn't mean they're a good absolute value. After all, you can't send your kids to college or retire early with good relative returns--only good absolute returns will allow you to do that.

This brings me to a discussion of the equity risk premium (ERP). (For a primer on this, you can read "The Best Time to Buy Stocks".) In a nutshell, the ERP is the excess return investors demand from holding stocks rather than treasury bills. The lower the ERP, the lower the earnings yield investors require to be compensated for holding stocks. When the ERP is very low, P/E ratios are very high because investors require a lower earnings yield to justify purchasing stocks.

Consider today's high P/E ratios as evidence that the average stock carries a very low equity risk premium. Investors have seemingly forgotten about their historical aversion to the volatility of stocks and are now willing to pay up for them. There aren't any disastrous geopolitical events that seem imminent, so people are piling into stocks.

The problem: The 5.1% real return on stocks I mentioned will come to pass only if P/E multiples remain as high in the future as they are today. This is a big bet indeed: Since World War II, the average P/E ratio of the stock market has been around 16.

Sure, dividend and capital gains taxes and transaction costs are lower than they used to be. But today's P/E ratio is twice as high as the long-run average P/E ratio of 16. Are we certain that the positive developments I've listed are enough to make stocks worth twice as much as they have been in the past? Ponder this: If, in 1946, Congress had passed a law lowering taxes to today's levels and mandating lower transaction costs, would stock prices have doubled instantly? Seems pretty unlikely.

By my calculations, if the market P/E ratio declines gradually from today's 32 down to 24 over the next 10 years, and my assumption about 2% real earnings growth holds true, the real return on stocks will be just 4.0%. This would offset much of the benefit of future earnings growth. And if the P/E falls to 18, still above the historical norm, the real return on stocks will be just 3.1%, completely negating the benefit of earnings growth. And remember, this calculation ignores capital gains taxes. Factoring those in would likely bring my estimate of the real return on stocks below 3%.

Conclusion: At today's prices, stocks probably represent a better long-term value than bonds do. That said, no matter where you invest, don't expect much. If you're planning for retirement, it's dangerous to assume you'll be able to get double-digit returns on stocks because there's no margin of safety in case something goes wrong.

Next week, I'll discuss a possible solution to this low-yield dilemma.