I recall having lunch with an associate on 8/25/87, the day the market reached a new high, and in fact, the high before the crash. He asked what I thought of the stock market. No, I didn't call the top. I told him that it had made some great gains, but that people who dollar cost average weren't too concerned with short term tops, or bottoms. Now, one thing I'll mention. How much did stocks fall for all of 1987? 20%? 10%?
What if I told you the S&P was up 5.74%? The index itself was up 2.03% plus 3.71% in dividends. You can find people who are ready to twist the facts, and of course if you bought on August 25 at 337.89 and sold at the October bottom of 216.46, you'd have lost 36% and perhaps been scared from the market for good.
Now, to my point. Lets go back and say you are in this for the longer term, still you bought, lump sum, at the high, 337.89. The market did well into the 90's, reaching a peak of 1530 in september 2000. Instead of discussing the return from peak to peak, let's look at the next low, March 12, 2003. The S&P bottomed at 788.90. The index return was 5.61%, but the index doesn't include dividends. With included dividends, the return over that 15-1/2 year period was 7.99%.
Buying on the very top before a crash, selling at the bottom, after the tech bubble, and you still enjoyed a return of 7.99%. Dollar cost averaging would both reduce the price going in, and increase it going out for higher returns, but I think the point is clear. I'm happy to hear any other views on this.