It would be really nice if we could tell the difference between 1929 and 1990 ahead of time, but the fact is that we have a pretty good historical record telling us that people are not so good at performing that task.  Think back to the early 90's - Japan in a crash, the U.S. set to go to war with Iraq, possibly causing massive oil disruptions, the Soviet Union breaking up with uncertain consequences, coming off the 1987 market crash and S&L crisis, the feeling that the greedy 80s were about to be punished with a bear 90s, recession imminent, housing prices crashing in the Northeast, etc.  Selling all your stocks in 1990 and staying out during the 90s for valuation reasons would have cost an investor just as badly as lump-summing one's entire lifetime investment in 1929.   

1929-1950 was indeed a pretty bad run, but of course any reasonable person who intended to hold investments until 1950 would have sought out further capital and investing values somewhere in between (look at Graham and Fisher).  Also he or she would have made great returns post 1950.  If we were actually sure that the next 20 years were going to be as bad as 1929-1950, stocks would already have crashed and there'd be nothing to talk about.  Don't let pessimism blind you - missing positives is at least as detrimental to long-term investing profitability as taking negatives.  There are always things to worry about, but as long as you take some profits in high times, find ways to buy value after crashes, and maintain some sort of asset allocation that suits your long-term goals and short-term cash needs, you're doing what you can.