Retirement begins in your 20s or 30s—or, better yet, at birth. That's when you should start investing to build a big enough nest egg so that you can live off income and capital gains from stocks and bonds, rather than off the sweat of your brow and your brain.

After 30 years of studying finance, I have found few eternal verities, but the most important—the Golden Rule of Accumulation—is this: Start early!

Time is the most powerful weapon in an investor's arsenal. In your 50s and 60s, or even your 40s, you probably won't have enough time left to build a substantial retirement portfolio—unless you own a business, work for a company that doles out generous pensions, or have made some profitable gambles in real estate.

Let's do the numbers. The average annual return for U.S. large-company stocks (as represented by the Standard & Poor's 500 Index) for the past 80 years has been about 10 percent after transaction expenses but not taxes. Say your goal is to build a nest egg of $1 million by the time you are 55. If you start at age 24 and invest $5,000 a year at an average return of 10 percent annually—through, for example, an index mutual fund or exchange-traded fund (ETF)—then you'll reach your goal. But if you wait until you are 34 to start, you'll accumulate only $357,000 by age 55. If you start at age 44, you'll have just $107,000.

Here's a second example, even more dramatic: One investor, call him Ishmael, begins at age 25 to put $2,000 a year in a low-expense mutual fund with an average return of 10 percent annually. At age 35, he's put $20,000 into the fund. Then he stops investing entirely. But of course, the value of Ishmael's holdings keeps rising, and by the time he is 65, he has a portfolio worth $556,000. [We are assuming this is a tax-deferred account like an IRA or 401(k).]

Now consider a second investor, Isabel. She, too, invests $2,000 a year at 10 percent, but she waits until age 35 to start. She keeps investing for a full 30 years—a total of $60,000 out of her pocket. At age 65, Isabel's got just $329,000.

In other words, those extra 10 years (between ages 25 and 35) produce 70 percent more money for Ishmael at retirement time, even though Isabel's out-of-pocket investment is three times greater.

How can this be true? The answer lies in compounding, the fact that interest increases the value of interest as well as the value of principal. If you earn 5 percent on $1,000, then after a year you'll have $1,050. After two years, you'll have not $1,100 but $1,102.50

As time passes, the power of compounding accelerates dramatically. Imagine that when your daughter is born, you give her a one-time-only gift of $10,000 worth of stock. Assume that the stock appreciates at 10 percent annually, on average. On her tenth birthday, your daughter's stock account will be worth $26,000 (I am rounding up to the nearest thousand in all cases); that is, it will grow in value in that first decade by $16,000. But over the second decade, her account grows by $41,000; over the next, by $107,000; over the next, by $278,000. If the account were earning simple interest of 10 percent on the principal, without compounding, then growth would be a mere $10,000 per decade.

One other important fact about compounding is that a small increase in the rate of return you achieve can have a huge impact over time. In the case of a gift to your newborn daughter, if her portfolio returns 10 percent, then $10,000 grows to $4.5 million by the time she is 65. But if her portfolio returns 8 percent, then it grows to only $1.4 million. If it returns 5 percent, it grows to a mere $227,000. In other words, half the rate of return produces an account that's less than one-20th the size.

But enough numbers! If you're a young person, all you need to know is that you must start early. If you are an older person who has young progeny or young friends, encourage them to start early. If you're particularly generous, set up a long-term trust or a Section 529 tax-advantaged college savings plan, or simply open a mutual fund account that the young person promises not to touch (or better yet, doesn't know anything about).

The next step is deciding what to put into the account. What is the best investment for the next 40 or 50 years? No one can be certain what the world will look like in 2050, but here are some modest assumptions:

1. For the U.S. economy and U.S. companies, the future will look pretty much like the past. We can expect stock-market values to rise at roughly the same rate as they have over two centuries, a little over 10 percent annually, on average.

2. The power of "creative destruction," as the economist Joseph Schumpeter described the capitalist process, will lead to the decline, merger or bankruptcy of many companies that are currently flourishing.

3. No one can possibly predict which business sectors will enjoy the best growth in market value in the decades ahead. Right after World War II, for example, financial guru Benjamin Graham predicted a boom in the use of commercial airlines, but he warned that copious passenger-miles would not necessarily translate into copious profits. He was right.

4. Globalization will increase opportunities for investment outside the United States.

5. The greatest risk to the buying power of a nest egg will be inflation.
Add three more guiding principles, and you have the basis for devising an ultra-long-term strategy for investing for retirement: keep expenses low, watch out for taxes, and diversify.

A simple but effective portfolio that would meet these criteria might look like this:

• 50 percent U.S. large-cap stocks, held in a managed mutual fund with minimal expenses or in an index fund or ETF.

• 20 percent U.S. small-cap stocks, held similarly.

• 30 percent international stocks, in managed or index funds or ETFs.

What, no bonds? Since 1926, according to the Ibbotson Associates data series, long-term U.S. government bonds have returned an annual average of about 5 percent; corporate bonds, 6 percent; stocks, 10 percent. After inflation, the differences are even more striking: bonds return between 2 and 3 percent; stocks, 7 percent. One dollar invested in large-cap stocks in 1926 grew to be worth $239 in buying power (that is, after inflation) by 2004. One dollar invested in Treasury bonds grew to just $6 in buying power.

The only reason to invest in bonds at all is to dampen risk. Bonds exhibit far less volatility in the short term than stocks, but in the long term, bonds are just as risky as stocks, mainly because bonds are more exposed to the ravages of inflation.

And when you construct a Start-Young Retirement Portfolio (which I'll christen the SYRP, like the stuff you pour on pancakes), you have vast stretches of time in front of you—30 years or more. Even for 20-year periods, stocks have been less risky than bonds. The Ibbotson figures show that since 1926, the worst 20-year period for large-cap stocks produced average annual returns of 1.3 percent after inflation, which is only about one percentage point below the average period for bonds. The worst 20-year period for bonds? An annual average loss of 3.2 percent. To put it briefly: for the SYRP investor, bonds stink.

Volatility, however, is worrisome only if you might have to liquidate your investment. With a SYRP, such short-term needs are not an issue, except perhaps as you approach retirement. When you get into your 50s, you can start moving assets that you hold in tax-deferred accounts from stocks to bonds. In taxable accounts, you simply devote your new contributions to bonds and leave the stocks alone, to build up capital gains.

Which brings up the final point: pay attention to tax consequences. With tax-deferred plans like the 401(k), 403(b) and IRA, capital gains and dividends build up tax-free (and should be automatically reinvested). But even a taxable SYRP doesn't have to generate much in the way of taxes if it's invested in a mutual fund like Dreyfus Appreciation (DGAGX) with an average annual turnover of just 8 percent, an indication that a typical stock is held for 12 years. The fund carries an expense ratio of a little less than 1 percentage point, and over the five years ending Nov. 30, 2005, it's ranked in the top 70 percent of all funds in its category.

Of course, you can control turnover and taxes best if you manage a portfolio of individual stocks yourself. Be my guest, but recognize the Schumpeterian truth: Some companies are going to vaporize over the decades, and you'll have to pay attention to your holdings and make adjustments to your SYRP. Just remember that the key is to pour it on. Now!