When I read Robert Schiller’s “The Life-Cycle Personal Accounts Proposal for Social Security: An Evaluation”, the third chapter of Mark P. Kritzman’s Puzzles of Finance, which is entitled “Time Diversification”, went to the top of my read pile. Let me suggest that Donald Luskin start with Mark’s primer at the end (Brad DeLong and Duncan Black are suggesting Don learn to read first). A couple of things that Luskin said left me shaking my head:
What the “paper” really said was that, in computer simulations of past investment performance, an investment strategy that Shiller invented out of whole cloth underperformed the returns of the existing Social Security system in 32 percent of the trials.
In a nutshell, a life-cycle account automatically moves an investor from higher risk holdings such as stocks into lower risk holdings such as bonds. A professional investment counselor might advise you do just that as you get older. It just so happens that I know quite a bit about life-cycle accounts and how they work, because I invented them.
Luskin was certainly not the first to think time diversification reduced the risk associated with putting one’s retirement money into stocks. And if he bothers to read Kritzman’s excellent discussion of Paul Samuelson’s Scientia 1963 paper, he might finally understand that it is the variability of terminal wealth and not the variability of average returns that matter.
But the first Luskin statement (above) is even worse because to understand the possible outcomes of an investment strategy where there is risk requires one makes some sort of assumptions as to expected returns and volatility of those returns. Samuelson assumed variations in returns are indepedently and identically distributed but even if they are not, investing in stocks long-term does not guaranteed that actual returns will average 7% as is so often assumed by the Cato Institute and Club for Growth discussions that Luskin trumps. To call Shiller’s exercise that simulated a particular strategy against a 3% fixed cost of borrowing (note Luskin did not even capture this aspect of Shiller’s paper properly) “invented out of whole cloth” is unfair. But those Cato Institute and Club for Growth assumptions that one can earn a guaranteed real return equal to 7% strikes me as weaved out of whole cloth. But if Mr. Luskin knows where I can get a riskless real return equal to 7% per year for 30 years, I’ll pay his patent filing fees.