"Aggressive investors should hold stocks, conservative investors should hold bonds."
"Long-term investors can afford to take more stock market risk than short-term investors."
Now, John Y. Campbell, Otto Eckstein Professor of Applied Economics at Harvard University, and Luis M. Viceira, assistant professor at the Harvard Business School, say in a new book these two rules of thumb are off the mark, if not downright wrong in some instances.
The problem, they assert, is that the main tool for generating optimal asset allocations, the mean-variance optimizer, has two fundamental flaws: it looks at the results for one time period only, and it assumes the financial world is static.
Where the optimizer first misses the boat is that it assumes investors care how wealth is distributed only after the end of a predetermined time period. But what investors really care about is how they can cover their expenses over the long haul, the authors argue in "Strategic Asset Allocation: Portfolio Choice for Long-Term Investors," just published by Oxford University Press.
This theory was pioneered 30 years ago by Robert Merton - co-winner of the 1997 Nobel prize in economics - but was "then left on the shelf" until the past five years, when enough computer power became available to do the complex computations involved, Mr. Campbell said in an interview.
Thirty years ago, economists also thought the financial world was a pretty safe place. Interest rates were fairly stable, as was the equity risk premium. Since then, short-term interest rates soared in the early 1980s, fell in the early 1990s, rose once again in the late '90s, and today are very low by historical standards.
One consequence of these roller-coaster rates is that the long-term bond affords little protection to pension funds, Messrs. Campbell and Viceira write, arguing pension funds would be better off investing in Treasury inflation protection securities, or TIPS.
"The point of my book is that you have to judge how much inflation risk is out there, and you have to form a view," said Mr. Campbell, who also is a partner in Arrowstreet Capital LP, Cambridge, Mass. "A pension fund board needs to form a view on the long-term outlook on U.S. monetary policy. Is inflation risk actually tamed? Can we count on the Fed to get it right?"
Similarly, Mr. Campbell disputes Jeremy Siegel's contention in his book, "Stocks for the Long Run" (McGraw Hill, 1998), that stocks are safer than Treasury bills and bonds for long-term investors. Many pension funds bought Mr. Siegel's analysis during the late 1990s, boosting equity exposure in the belief stocks were a safe bet over time.
"There's a logical inconsistency in this argument," Mr. Campbell said. The only way the risk of holding stocks shrinks over the long term is because equity returns revert to the mean - that is, stocks go through bull and bear cycles.
"If there's mean reversion, that means stocks must be more right at some periods than others," he said. Instead, pension funds are better off doing long-term strategic market-timing - increasing stock holdings as the market drops, and selling holdings as the market goes up, but in a much more gradual way than in tactical asset allocation.
Mr. Campbell has a point. While academic studies show that investors do poorly when trying to time the market in the short run, he wants investors to exercise their judgment on 20-year market trends - surely what a prudent expert should try to do.
