When I mentioned the topic of this article to some of my colleagues, they were instantly curious about what book I call my favorite. Huckleberry Finn? The Bible? John Bogle’s Little Book of Common Sense?

Nope. All of those are excellent books. But in this case my favorite book is one you can’t read yourself because you can’t buy it or find it at a public library.

(And unless you are totally fascinated by large tables of numbers, you wouldn’t find this to be good reading at the beach or anywhere else.)

This 64-page book, a reference work for investment advisors and economists, has a simple title: “MATRIX BOOK 2009,” published by Dimensional Fund Advisors. Most of its content is annualized rates of return for dozens of kinds of investments with data as far back as 1926 and as recent as December 2008.

I especially like this book because it gives me “just the facts and nothing but the facts,” letting me figure out for myself what might be important without anybody else’s biases or emotions getting in the way.

Hidden in these pages, for people who know what to look for, are some fascinating lessons about what has worked in the past. Even though the future won’t be the same as the past, I expect many long-term patterns to be similar.

(In the examples below, unless otherwise noted the returns cited are annual compound rates.)

Now I’d like to share 10 very interesting things this book teaches me:

Lesson 1: The long term can be much longer than we might think. Here’s what I mean. Most people would agree that over the long term, stocks have had higher returns than bonds. And most people understand that for short periods, the reverse can be true. And most people probably think 20 years is plenty long enough to validate this “long-term” expectation.

But it’s not. Here and in some of the other lessons, let’s compare the most recent 20 years with the most recent 50 years.

From 1989 through 2008, long-term bonds (both corporate and government) had higher compounded annual returns (8.5 percent and 10.1 percent respectively) than the Standard & Poor's 500 Index (8.4 percent). Going back 30 more years, to 1959, we find the stock index returned 9.2 percent, beating long-term government and corporate bonds (7.5 percent and 7.2 percent respectively).

Lesson 2: Similarly, it turns out that the most recent 20 years wasn’t enough for U.S. large value stocks to demonstrate their very-long-term advantage over U.S. large stocks that blend growth and value (in other words the S&P 500 Index). From 1959 through 2008, large value stocks beat the index by 1.2 percentage points (10.4 percent vs. 9.2 percent). But from 1989 through last year, the index (8.4 percent) outshined value stocks (5.5 percent).

Lesson 3: Sometimes 20 years is more than enough for an asset class to live up to our expectations. For example, we expect smaller stocks to outperform larger ones. From 1989 through 2008, U.S. small-cap stocks beat the S&P 500 Index by 0.4 percentage points; U.S. small-cap value stocks outperformed large-cap value stocks by 4.3 percentage points. (Over the past 50 years, the annual advantage was 2.1 percentage points for small-cap stocks and 3.2 percentage points for small-cap value stocks.)

Lesson 4: We are often asked why we don’t recommend growth stock indexes. My favorite book contains the answer. Growth stocks, of course, are very popular. But in the 20-year and 50-year periods we looked at, both large and small growth indexes underperformed their respective value counterparts by 10 to 25 percentage points. By recommending blend funds instead of growth funds, we weight the portfolio toward value but we still include plenty of growth stocks.

Lesson 5: This is one of the biggest surprises in this book, hidden on pages 36 through 39, where U.S. stock returns are compared with those of the United Kingdom. If you thought about it, you’d probably expect U.S. returns to be higher than those from the U.K. But for the 50 calendar years ending in 2008, large U.K stocks outperformed those in the U.S. by 1.2 percentage points (10.4 percent to 9.2 percent). Small U.K. stocks compounded at 13.4 percent, vs. 10.9 percent for small U.S. stocks. (These U.K. returns are computed in U.S. dollars.)

Lesson 6: Japanese stocks have been on a roller-coaster ride in recent decades. From 1970 through 1989, Japanese large and small stock indexes rose at 22.4 percent and 29.7 percent, respectively. It was perfectly obvious to many people in the late 1980s that Japan was on the road to overtaking the United States as the world’s economic leader.

Then Japan’s good fortune came to a screeching halt. For the 19 years from 1990 through 2008, small-cap Japanese stocks lost 30.5 percent cumulatively; their large-cap counterparts lost 40.5 percent cumulatively.

Many Japanese investors, undoubtedly being “loyal” to their country’s companies and sure a turnaround was always just around the corner, learned a very painful lesson in the high risks of investing only in domestic stocks.

Lesson 7: Not all international stocks are equal. This sounds obvious, but too many investors skip over the details and thus take the risk of acting on the wrong conclusions. There’s a widespread sense that international stocks have done better than U.S. stocks. True, and also untrue. Here’s the true part: From 1970 through 2008, small-cap international stocks compounded at 14.4 percent, vs. 9.5 percent for U.S. small-cap issues. But in the same 39 years, large-cap international stocks compounded at 9 percent, vs. 9.5 percent for the S&P 500 Index.

Lesson 8: For most people, the biggest long-term losses don’t come from picking the wrong mix of assets. They come from the insidious, almost invisible force known as inflation. And an investor’s best weapon against inflation is owning stocks. From 1926 through 2008, ultra-safe one-month Treasury bills compounded at 3.7 percent without ever experiencing a losing year. That’s before inflation. After inflation, the return was reduced to 0.7 percent. (Taxes, computed on the 3.7 percent nominal gains, likely took the real return into negative territory.) In that same period, the S&P 500 Index had an inflation-adjusted return of 6.4 percent.

Lesson 9: If you didn’t care about risk (measured by volatility), you could have put your money into some interesting assets that subject yourself to much more risk than necessary. From 1988 through 2008, the highest asset class return shown in my favorite book was emerging markets, with an 11.4 percent compound return. In second place were U.S. small-cap value stocks, at 10.7 percent. Both those asset classes were quite volatile. If you had the benefit of hindsight, you might have been happy to choose the third-place finisher, long-term U.S. government bonds. They returned 10.1 percent.

Lesson 10: Perhaps the most important lesson of all is how hard it is to accurately predict long-term investment returns. The past has no risk, because we know just how things turned out. And we know what we “should have” done. As always, the future is unknown.

I place my greatest faith in the returns from the very-long-term past. Based on everything that I know, I believe stocks will likely give me more future return than bonds – in the long run. But because life exists in the short run, I include plenty of bonds in my portfolio.

And even though I have the benefit of this wonderful book filled with numbers from the past, I must admit that I know absolutely nothing about which specific asset classes will turn in the best returns over the rest of my lifetime. For that reason, I am most comfortable having my equity investments diversified very broadly over many asset classes.

My best advice to you is to do the same.

Paul Merriman is founder of Merriman.