“For example, stocks are riskier than bonds, and over time always have greater returns.”
In a LW vein, it’s worth noting that selection and survivorship biases (as well as more general anthropic biases) means that the very existence of the equity risk premium is unclear even assuming that it ever existed.
(I note this because most people seem to take the premium for granted, but for long-term LW purposes, assuming the premium is dangerous. Cryonics’ financial support is easier given the premium, for example, but if there is no premium and cryonics organizations invest as if there was and try to exploit it, that in itself becomes a not insignificant threat.)
The survivorship bias described by wikipedia is complete nonsense. Events that wipe out stock markets also wipe out bond markets and often wipe out banks. Usually when people talk about survivorship bias in this context, they mean that the people compiling the data are complete incompetents who only look at currently existing stocks.
If your interest is in the absolute return and not in the premium, then survivorship is a bias.
ETA: I think I was too harsh on the people that look at the wrong stocks. But too soft on wikipedia.
In a LW vein, it’s worth noting that selection and survivorship biases (as well as more general anthropic biases) means that the very existence of the equity risk premium is unclear even assuming that it ever existed.
(I note this because most people seem to take the premium for granted, but for long-term LW purposes, assuming the premium is dangerous. Cryonics’ financial support is easier given the premium, for example, but if there is no premium and cryonics organizations invest as if there was and try to exploit it, that in itself becomes a not insignificant threat.)
The survivorship bias described by wikipedia is complete nonsense. Events that wipe out stock markets also wipe out bond markets and often wipe out banks. Usually when people talk about survivorship bias in this context, they mean that the people compiling the data are complete incompetents who only look at currently existing stocks.
If your interest is in the absolute return and not in the premium, then survivorship is a bias.
ETA: I think I was too harsh on the people that look at the wrong stocks. But too soft on wikipedia.