People have often danced around the issue of how inflation affects the investment return on common stocks. But virtually no one has offered clarity on the issue.

At times, over the years, one would hear that inflation was actually good for common stocks since it tended to boost earnings growth. The investment business has often taken this conventional wisdom, even publishing it as thoughtful analysis, when factual analysis of historical data and events might offer different conclusions. One key point sometimes made is that price/earnings ratios (P/Es) are primarily controlled by interest rates.

When you buy any stock and hold it over any period of time your total return is made up of three, and only three, components:

Return = Earnings Growth Rate (%) + Dividend Yield (%)
+/- Change in Price/Earnings (P/E) Ratio (%)

Actually, the earnings growth and P/E terms are multiplied by each other and not added but this conceptual equation will do for purposes of discussion.

The reason earnings growth is broken out is that stock prices tend to ultimately go where the earnings go. Look at Figure 1, Wealth Indices - 10 Year Moving Average. Here, we have three indices: stock principal return (i.e., price only), earnings growth rate and consumer price inflation. The Standard & Poor's 500 index is what is referred to as the stock. A 10-year moving average is used on all three indices (which is indexed to 1.0 in 1880) since one wants to focus on trends and not year-to-year noise. Note that over time, the stock price ultimately goes where earnings go. Now, note the really interesting part.

Look at how stock price over time trends above-then below-earnings. One will see that price tends to cut upward through earnings when the inflation index is rolling over or slowing down. This is a long period of P/E expansion. Conversely, price tends to move downward through earnings when the inflation index is accelerating upward or increasing its rate. This is a long period of P/E contraction.

The second component in the equation is the dividend yield (the annual dividend in dollars divided by stock price and shown as a percent). The third is the percent change in the P/E ratio. Anyone who was invested in stocks in the 1970s knows very well the negative effects from P/E changes. The opposite was true in the late 1990s when expanding P/Es caused many stocks to go far above fair values.

What About Interest Rates?
Conventional wisdom would seemingly be true. That is, since ambient interest rates present competition to stocks, it should be interest rates that have a significant effect on stock P/E ratios. In other words, as interest rates get higher, it would make more sense to purchase safe, interest-bearing vehicles in lieu of taking on the volatility and risk of company shares. But look at Figure 2, 10-Year Average Annual Interest Rates, Inflation, P/E.

The top panel of this chart shows both long-term interest rates (20-year, high-grade corporate bonds) and short-term interest rates (90-day Treasury bills). Both show 10-year averages in their rates so we can focus on trends and not year-to-year noise.

The center panel shows a running 10-year average of U.S. consumer price inflation. Take note of the past 100 years and the long periods of rising inflation followed by long periods of declining inflation. The bottom chart panel shows the 10-year running average P/E ratio for the S&P 500 index. Note the long periods of declining P/Es followed by long periods of rising P/Es.

Now, compare the P/E chart to either interest rates or inflation. From roughly 1950 to 1980, interest rates were in an uptrend in the U.S. Over this period, P/E did a complete cycle from low to high and back to low. The relationship between interest rates and P/E appears doubtful when viewed over the long term. Maybe one could argue that it is over the short term that interest rates influence P/Es, but this is also dubious, and anyway, making good investment decisions in the short term is extremely difficult.

The relationship between P/E and inflation, on the other hand, is absolutely dramatic. P/E peaks and valleys tend to lag inflation valleys and peaks by a few years. This makes sense since investors always tend to fight their last battle for a few years. For example, when the wrenchingly high inflation of the 1970s suddenly stopped in the early 1980s, investors who had lived through it could not believe that it was actually gone. It wasn't until about the middle of the new decade that they really began to accept that the demon was dead.

How could inflation be more important in P/E determination than interest rates? Because when you buy a stock, you are actually paying the discounted present value of all future earnings. Conventional wisdom would use the ambient long-term interest rate as the rate to discount future earnings back to present value. However, it appears that in most instances it is the rate of inflation that is used to discount future earnings to present value.

This applies in most instances, except than when interest rates are so extremely high that they outshine potential stock returns.

A Fair Value Channel

Figure 3 shows the "fair value channel," a mathematically derived path through time marked by the movements of the Standard & Poor's 500 stock index. The channel is defined by a computer model that never changes its mathematics, and has only two inputs: earnings growth and consumer price inflation. In other words, the model hinges on the belief that these are the two primary forces driving the direction of the stock market. For past results, the model uses actual earnings growth and inflation, while the future results rely on predicted growth and inflation.

The fair value channel ties everything together, starting with the idea that stock prices ultimately follow earnings growth, but then showing how periods of low- or high- trending inflation will detour stock prices in a way dictated by the historical relationship between inflation and P/E ratios.
This relationship can be generally quantified based on 100 years of history. The width of the channel is always the same on semilog paper. That is, the top of the channel is always the same percent above the center line of the channel, while the bottom is always the same percent below the center line. The sophistication of this model lies in its simplicity: The stock market is not trying to defy and confuse investors. It simply does what it does in the long term because of a couple of rather simple variables.

Many investors get all balled up in attempting to do the impossible. That is, they attempt to outguess the market in the short term, encouraged by some shouting heads on television or supposed gurus holding conferences in nice locations.

Note in 1974 how the market was well below fair value. Then see in 2000 how the market was well above fair value. It's true that attempting to time major shifts into and out of stocks is generally a loser's game. But it is still always good to feel you understand the dynamics of how the market works, and our model could be very helpful for an investor pondering when to raise cash needed for withdrawal or when to invest new cash added to an account.

If one could accurately predict earnings growth and inflation for the S&P 500 over the next decade, one would have narrowed a multitude of variables down to just two and one would have a good handle on where the market was headed. In addition, the value of these two variables could give a good indication about whether classic "growth" stocks or classic "value" stocks would be the better performers.

After defining what growth and value stocks are, you can use this inflation-earnings analysis in different economic cycles to determine when each investing style would likely perform the best. However, that is a topic for another time.

Inflation Momentum As A Predictor
Now that the point has been made that consumer price inflation is a key factor in common stock investing, it would be extremely useful to have a measure of "inflation momentum" through time, in other words, a measure of inflation that was not volatile from year to year or even over a number of years. In such a measure, when inflation momentum hits a top or a bottom, it would be a trend-shifting top or bottom and not just a fake. It would be interesting to see how the Standard & Poor's 500 index performed in the decade after just such a major top or bottom in inflation momentum.

The stock market has a long history of total return in the vicinity of 10% per year in compound average annual return. This comprises approximately 6.5% per year from earnings growth and 3.5% per year in dividend yield. As seen in Figure 2, stock P/Es seem to be inversely correlated with inflation. Sometimes the P/E change has a positive additive effect on returns (as it did in the late 1990s) while at other times it has a negative effect (in the 1970s). The point is that while the long-term average return is 10% the market never spends much actual time at 10%. It seems to spend about half the time below 10% and the other half above 10%.

In Figure 4, Inflation Momentum vs. Stock Returns, the top panel shows the calculation of inflation momentum, where you can eyeball the tops and bottoms in inflation momentum since 1915. Over 92 years, there are three tops and three bottoms. In the bottom panel is the average annual total return for the Standard & Poor's 500 stock index over 10-year periods. This includes price changes and dividend yield. Note that the chart has been moved back 10 years (to the left). The reason for this is that, given any date on the inflation momentum chart, one wants to quickly be able to see the stock market return for the following 10 years.

In general, when inflation momentum peaked, the following 10 years was generally good for stocks. That is, a period where the market was on the high side of its 10% per year average. Conversely, when inflation momentum bottomed out, the following 10 years were generally subpar for stocks. Historically, whenever inflation has gotten started, it has generally been difficult to reverse its course. Recently, the Federal Reserve has let inflation rise unchecked cutting interest rates continually in hopes of avoiding an awful recession brought on by the subprime loan fiasco and other things.

It is generally better not to be retiring just after inflation momentum hits a major bottom, as it did recently. This is because stock returns in general might tend to be subpar for a number of years. Consider a 30-year period where stocks return 10% per year. In one instance, assume stocks return 5% per year for the first 15 years and then 15% per year for the next 15 years. In the second instance, reverse it so the 15% per year comes in the first 15 years then the 5% per year. With annual withdrawals from the portfolio, one is much better off at the end of 30 years where the higher returns come first.

While the historical relationship between inflation and stock returns (to include P/Es) is a given, there is the possibility that this relationship will not hold up this time.  Nevertheless, history always has something to teach us and it can never hurt to understand the past.

The current economic malaise is worldwide, and the solutions may be difficult to implement. The weak U.S. economy is currently not in shape to easily tolerate the Federal Reserve's traditional inflation-fighting methods.  A free enterprise economy that goes to excess is typically followed by more prudent behavior by businesses and investors, but only after they have been properly chastised into a phase that can be disconcerting.

The Federal Reserve may quickly get inflation under control.  Also, the slowing economy could well kick in as a drag on high inflation. And there are different types of securities for different inflation phases, but that is a subject for another day.

R. Stewart Eads, CFA, is president of Eads & Heald Investment Counsel and co-founded the firm in 1987.  He has an MBA from Wharton Graduate School, and BEE and MSEE degrees from Georgia Tech.  He was a senior portfolio manager with Wellington Management Company for 15 years.  Prior to that, he was with the Investment Department at The Hartford Insurance Group.  He also has worked in the Office of the Director, U.S. Central Intelligence Agency, and as a NASA engineer.