Abstract This article brings the research on survivorship together with research on predicting the stock market. A key argument of the article is that studies of long-term prediction can and should take into account the fact that a market has survived. Survival looks like a rebound, ex post. If stock market crises were always followed by rebounds, then markets would have the wonderful feature of reversing themselves after bad times, and of rising after large increases in dividend yields. Alas, history is not so kind. As a result, researchers have to account for the “bounce” when they look at countries such as the United States and the United Kingdom and also have to try to figure out what to do with the gaps and interruptions of other stock markets from other countries, should they decide to use them.