看了下面这段明白了,
Suppose we invest $250 in each of four stocks. The first stock earns 50 percent, the second stock earns 20 percent, the third stock earns 10 percent, and the fourth stock goes bankrupt mid-year. We end the year with $950 ($375 + $300 + $275 + $0), or a loss of 5 percent on our $1,000 investment. Survivorship bias occurs when one calculates the returns on this group of investments using only the survivors at the end of the year, which in this case are the first three stocks. The survivors returned over 26 percent, much more impressive than a five percent loss. Survivorship bias is a virus that can infect indexes and the composition of mutual fund peer groups, but not just that.
也就是说,从历史算是指根据历史的return倒推,而不是从历史的required return倒推equity risk premium。
而用历史数据往往会把那些倒闭了的公司(往往return是负的)给排除在外,所以return综合起来就偏大了,equity risk premium也偏大了。 |