No one knows how long stocks will keep going up. No one knows whether their inevitable fall will be a "correction" or a catastrophe. To put it plainly, no one knows anything -- yet investment decisions still have to be made.

So I'm offering a framework for making them. It's one used by professional investors to walk the high wire without losing sight of the safety net.

To explain how it works, let me start with an example used by Marshall Blume, professor of finance at the Wharton School in Philadelphia and a principal in the money-management firm ofPrudent Management Associates:

A euphoric investor put $10 into stocks in August 1929, just before the Great Crash. Not until 1945 did he recoup his losses.

A Blume-taught investor, on the other hand, put $5 into stocks and $5 into bonds and maintained the 50-50 split. When the bond market rose, he cashed in some of his profits and bought stocks, bringing his portfolio back to an even division between stocks and bonds. When stocks rose, he took some profits and bought bonds.

He rebalanced his investments every month to keep on the 50-50 track. He recouped his losses by October 1935.

Blume's investor was practicing careful "asset allocation," which -- simply put -- means deliberately dividing your money among stocks, bonds, cash and other types of investments. According to a study by the First National Bank of Chicago, asset allocation is the long-run determinant of your investment success or failure, not how smart (or dumb) you are in picking particular stocks or mutual funds.

But the moral is more than "diversify to protect your rear." The key to this strategy is continually returning to the original division among stocks, bonds and cash to limit your risk.

Most investors keep more and more of their money in stocks as the market rises, which makes their positions ever more dangerous. When the bear strikes, they have more to lose.

Conversely, they are afraid to buy when times are bad, so they lose their chance to get stocks cheap. That's another beauty of keeping your stock and bond investments at a fixed ratio: It combats those bad-timing impulses that cost you money.

A high-risk portfolio is 100 percent in stocks. A moderate risk might be 50 percent stocks, 50 percent bonds. Low risk might be 20 percent stocks, 30 percent bonds and 50 percent cash -- for example, money-market mutual funds.

On a high-risk portfolio, the upside potential is roughly 30 percent a year over the next five years, by Blume's calculations. The downside is a compounded loss of 6 percent annually, if the market suffers one of its periodic drops. In general, only young people can afford to shoot for these top returns, because only they have enough time to make up for the big losses that can occur along the way.

On a moderate-risk portfolio, your potential gains are 22 percent annually over five years and your potential losses, 1 percent compounded. So by angling for moderately lower gains, you vastly decrease your exposure to risk -- an appropriate stance for the middle-aged.

On a low-risk portfolio, your upside is 14 percent annually over five years. Your downside is a comforting 3 percent gain compounded over five years. This suggests that the newly retired, who cling to their insured CDs, can afford to keep a little money in stocks and bond mutual funds. Such investments can raise your returns without seriously affecting long-term security.

Rebalance your portfolio periodically to keep it in line with the level of risk you want to maintain. Blume says that you'll probably do okay if you rebalance once a year. Richard Brignoli of Brignoli Models, who allocates assets for institutional investors, suggests rebalancing whenever a portfolio gets 10 to 15 percent out of line.

As a practical matter, you can't use this system with individual stocks and bonds because you can't diversify enough. But it's a cinch to rebalance with no-load (no sales charge) mutual funds.

In general, you might want to divide stock participation between aggressive funds and conservative ones. On the bond side, you'd split between low-risk, short-term funds (maturities of around one year) and the riskier longer terms (maybe five to 10 years).

For portfolios over $200,000, Brignoli would add investment real estate and, after that, gold. But staying in balance is the key. Since you cannot predict the future, you build your fortress and stay put.