My favorite curmudgeon, the Scrooge of Stocks, is a former National Geographic photographer named Charles Allmon. He lives in Potomac, manages about $100 million for clients, and is the founder and editor of Growth Stock Outlook, now in its 38th year.

When I first met him a couple of decades ago, Allmon was a bull. He actually owned a lot a growth stocks, hence the name of his newsletter. But as the 1980s wore on, he became more and more gloomy. His great coup was predicting the 1987 crash two months before it happened, but, as Business Week put it, "The trouble is, Allmon has remained bearish ever since." Since October 1987, the Dow Jones industrial average has gone from 1700 to over 11,000 and lately to 8700, but Allmon thinks it's too expensive. Last week, 78 percent of his assets were in cash.

We're poles apart in our view of the future of stock prices, but I love his newsletter, which always has a glistening nugget or two. The latest issue offers a little sidebar with this amazing fact: Since June 27, 1973, the date Allmon started publishing a model portfolio, the Wilshire 5000 index has returned 2,322 percent while the Standard & Poor's 500-stock index has returned 2,247 percent.

In other words, over 281/2 years, these two portfolios -- one currently comprising more than 6,500 stocks; the other just 500 -- have produced almost precisely the same gains. Start with $10,000 in 1973, and you would have $242,220 with the Wilshire and $232,470 with the S&P -- a difference of just 4 percent. And, by the way, during this period, Allmon's own portfolio rose to $240,000.

These figures offer two important lessons for investors.

First, over a generation that included rampant inflation, high interest rates, several severe recessions, the worst one-day crash in history, two major wars and a calamitous terrorist attack, stocks nonetheless generated gigantic gains.

Second, over time, as the title of a Flannery O'Connor book puts it, "Everything that rises must converge." Put together an intelligent diversified portfolio, and, chances are, it will come close to performing the way that the market averages do. As a result, the argument for putting your money into a low-cost index fund is a powerful one. After fees, about two-thirds of managed U.S.-stock mutual funds failed to beat the S&P over the past 10 years. Part of the reason, as I pointed out last week, is that many fund managers trade too much and make poor choices; the other part is the fees themselves.

I haven't completely given up on managed -- that is, human-run, as opposed to computer-run, mutual funds. I still think a few gifted managers can beat the market with consistency. But how much time and effort should a small investor expend in trying to ferret out such geniuses? And, since past performance is no guarantee of future success, is there even a rational way to find them?

As usual, the best wisdom on the subject comes from Warren Buffett, who wrote in the 1993 annual report of the company he chairs, Berkshire Hathaway, that by "investing in an index fund, the know-nothing investor can actually outperform most investment professionals." He added: "Paradoxically, when 'dumb' money acknowledges its limitations, it ceases to be dumb."

Buffett was not using "dumb" in a pejorative sense. He was simply saying that many people aren't inclined toward learning about businesses and markets. For them, computer-managed index funds, a gift of this technological age, are just fine.

The reason to invest in an index fund is that the U.S. stock market is a proxy for the U.S. economy, which, in all its robust diversity, has been the best economic bet on the planet for the past century or so.

The problem is that there's no single method of defining the U.S. stock market. The Wilshire 5000 total market index, maintained by Wilshire Associates, a Santa Monica, Calif., investment advisory firm, tries to capture every common stock listed on the three major exchanges -- at last count, about 7,000 of them. Wilshire's Web site properly claims that the index is "the best measure of the entire U.S. stock market."

The Wilshire, like the S&P 500, is "capitalization-weighted," which means that the influence of an individual stock on the overall index is determined by its market cap, or value according to investors. (To calculate a stock's market cap, multiply its price by the number of shares outstanding.) As of Wednesday, the five largest U.S. companies by market cap were, in order, Microsoft (MSFT), General Electric (GE), Wal-Mart Stores (WMT), Exxon Mobil (XOM) and Pfizer (PFE). Together, their capitalization was about $1.2 trillion, compared with a total capitalization of $9 trillion for all the Wilshire stocks.

A second approach to buying the market is to own the S&P 500, another market-cap-weighted index, composed (with a few exceptions) of the 500 biggest U.S. companies. As Standard & Poor's says on its Web site (www.standardandpoors.com): "The S&P 500 focuses on the large-cap sector of the market; however, since it includes a significant portion of the total value of the market, it also represents the market."

In other words, 500 stocks stand for 7,000 stocks. And they do. At last report, the total market cap of the S&P was $8.5 trillion -- or more than 90 percent of the market cap of the Wilshire.

Who gets into the S&P? Size counts, of course, but a stock also has to be heavily traded and have a "reasonable price" and four consecutive quarters of profits (after extraordinary items). The index now excludes all foreign-based companies -- such firms as Unilever (UN) and Nortel Networks (NT) were recently purged -- and the S&P selection committee, headed by economist David M. Blitzer, reserves the right to add and subtract stocks to achieve "sector balance."

That balance gives the S&P 500 a diversification similar to that of the economy as a whole. The index is divided into 10 sectors. Recently, the 81 financial stocks represented 20 percent of the S&P's total market cap while, at the other extreme, 37 utilities represented just 3 percent. Information technology is the second-largest sector, with 77 stocks and 16 percent of market cap.

So far this year, the index has deleted 21 stocks, including WorldCom, which was bumped on May 14, more than a month before revelations of phony profits. The 21 replacements include eBay (EBAY), the online auctioneer, and Anthem (ATH), a large health-benefits company that went public only a year ago.

In the end, the difference between the S&P and the Wilshire is so tiny that it's not worth worrying about. Yes, the S&P is more concentrated than the Wilshire: Its top five holdings recently represented 14.4 percent of total assets, compared with 11.8 percent for the Wilshire. And, of course, while the S&P includes no small-caps, such stocks represent about 10 percent of the Wilshire's total value. There may be times, especially if small-caps have been depressed for a long time, when the Wilshire may be a better choice than the S&P, but, in general, the choice of index is a coin flip.

Now, let's look at how to buy an index.

The most popular mutual fund that mimics the Wilshire is Vanguard Total Stock Market (VTSMX). It owns only about half (3,306 to be exact) of the index's stocks. For the 10 years that ended Nov. 30, the fund has returned 9.4 percent. Over the same period, Vanguard 500 Index (VFINX), the largest mutual fund in the world, has returned 10.1 percent by tracking the S&P 500. Over the past five years, the two funds have finished in almost a dead heat. Return for Total Stock Market: minus 0.003 percent. Total return for 500 Index: plus 0.8 percent.

The key to deciding which index fund to buy is the expense ratio: How much is the managing firm putting into its pocket? In the case of Vanguard's Wilshire fund, the answer is 0.2 percent -- that is, 20 basis points, or one-fifth of 1 percent. Vanguard's S&P 500 fund charges only 18 basis points, rock-bottom. By contrast, the T. Rowe Price Equity Index 500 fund (PREIX, which also tracks the S&P) charges 35 basis points. Wilshire Associates has its own total-market fund, Wilshire 5000 Index Portfolio (WFIVX), charging 60 basis points. If that sounds high, remember that the average managed fund charges 140.

Another index choice is Schwab 1000 (SNXFX, which simply owns the 1,000 largest U.S. stocks. As a result, 21 percent of its holdings are mid-caps, compared with 13 percent for Vanguard Index 500. Again, that's a tiny difference. If you have a Schwab account, buying the fund is a breeze, but it's not cheap. Expenses, according to Morningstar, have averaged 46 basis points annually for the past five years.

For all their advantages, index funds won't guarantee a smooth ride. If history is a guide, an index fund will lose money, on average, every three or four years, and, by definition, it will never, ever beat the market averages, always falling short by its expense ratio. Still, again according to history, the average index fund ought to whip the average managed fund. Sad for us humans, but true.

James K. Glassman's e-mail address is jglassman@aei.org. The paperback version of his latest book, "The Secret Code of the Superior Investor" (Three Rivers Press), is being published this month. He will discuss investment issues in a live online discussion tomorrow at 11 a.m. at www.washingtonpost.com/liveonline.