What's the one main driver for stock market returns? It's got to be economic growth, right? Growth is paramount to the increase in stock prices. After all, that's what we're told every single day by the financial media.

Actually, according to a new book by Crestmont Research's Ed Easterling, that's basically wrong. The single biggest determinant of stock market gains is the trend in inflation. His new book, Unexpected Returns: Understanding Secular Stock Market Cycles, is destined to be an investing classic.

Earlier this year, Easterling approached me to review his book before its release. I thought I was an odd choice for the job, because I'm generally not a macroeconomic, top-down kind of guy in my security analysis. Ask me what the market's done, and I may not know the answer. But the more I thought about it, the more I realized that I do make all sorts of top-down assumptions, even when identifying individual securities. I don't want anything to do with the homebuilders Toll Brothers (NYSE:TOL) or NVR (AMEX:NVR), for instance. I have as much absolute confidence in the fact that we're at the top of the cycle as I do in the fact that I cannot predict "when" things will change.

Business cycles certainly do matter. So do market cycles. And historically, the latter are led by the rise and fall of inflation. During the 16-year period of 1965 to 1980, the stock market essentially ended at the same price where it started. Meanwhile, earnings per share during that period rose at essentially the same rate as they did during the boom years of 1980 to 1999. In fact, the time period that included the biggest bull market in modern history -- when Intel (NASDAQ:INTC), Cisco (NASDAQ:CSCO), Yahoo!, and a handful of other simply amazing wealth-creating machines entered the mainstream -- showed lower nominal economic growth than the earlier period. This piece of research is counterintuitive and, well, baffling. The key element is that the rate of inflation during the former period was substantially higher than that of the latter. Also, the valuation at the beginning point in 1965, with the S&P 500 average price-to-earnings (P/E) ratio of 23, was dramatically higher than it was in 1980, when the S&P 500 average P/E was 7.

Or put more simply, in a way my fundamental brain can understand: Your expected returns on a market portfolio are determined by the price at which you invest in the first place. Buy when things are expensive and you may expect below-average returns over the long term. Buy cheap and you may expect the opposite.

Unexpected Returns is a follow-on offering to John Mauldin's Bull's Eye Investing, which was published in 2003 to much fanfare and acclaim. Easterling provided some data for that book, and Mauldin thought so much of the data that he had Easterling co-author two chapters of the book. Unexpected Returns focuses on two components. First, the stock market matrix, which shows how well the stock market -- as measured by the Standard & Poor's 500 -- has performed for any length of time over the past century. This matrix shows one clear piece of evidence: People who buy in expensive markets have a much higher risk of both disappointing results and bone-crushing declines.

The second element is a proprietary model that Easterling calls "Financial Physics." Very simply: Over the long term (the past century), the U.S. real gross domestic product has grown at a fairly stable rate of about 3%. So if stock prices were only a function of growth, P/Es would also stay somewhat constant. But they haven't. When we've been in a rising inflation environment, P/Es have tended to fall; the opposite, and they've tended to rise. Further, Crestmont's research shows that during secular bear markets, stocks tend to be substantially more volatile -- with enormous rallies and declines -- than during bull markets.

So, where are we now?
Mind you, Easterling is working with historical data, and it is macroeconomic in nature. While physics tends to operate based on fundamentally simple laws, financial physics does no such thing. But based upon historical relationships of inflation, P/Es, dividend yields, and long-term interest rates, Easterling suggests that we have a nearly classic set-up for a long-term secular bear market. To wit, at the moment the S&P 500 trades at a P/E of 20, according to Barra, which is quite high. Inflation, despite the government's "hedonic" adjustments, is rising, dividend yields are low, and so are interest rates.

Easterling's not alone here. Warren Buffett at Berkshire Hathaway (NYSE:BRKa) told investors in 1999 that they shouldn't count on the market performing at even average levels during this decade. Buffett doubted that the market would return better than 4% annually over the next 17 years, and he was looking at very much the same inputs. He spelled out his thinking in a December 2001 interview in Fortune with Carol Loomis.

Unexpected Returns isn't a Chicken Little book, meant to scare people away from stocks. As Buffett's own performance in the gloomy late 1970s suggests, individual stocks can perform extremely well even when the market does not. We've seen the same thing in the past five years with the explosive performance of such tech-boom-era castoffs as Church & Dwight (NYSE:CHD) and 2003 Motley Fool Hidden Gems recommendation Valero (NYSE:VLO). Despite overall market losses, those stocks have outperformed. Easterling's book shows in great detail what can happen, though, when the market as a whole starts expecting results that just don't have much basis in reality.

Bill Mann wonders when the NCAA will demand that states whose names are derived from Native American names and words change them. Or are we going to see games between "the state school of the small state just east of New York but not directly south of Vermont vs. another state school that is just north of Texas." Come on! Bill writes for the Motley Fool Hidden Gems newsletter, which you can sample for 30 days for free. He holds shares in Berkshire Hathaway. The Fool has a disclosure policy.