He subsequently amended that advice, saying institutions' policy mixes were not tied to the right goal, which is to pay their liabilities. "I said, ‘Throw the policy portfolio away and play it by ear.' I think you can't do that because you need some kind of benchmark to know what your performance has been," Mr. Bernstein said in an interview.
"But to put (the policy asset mix) in cement, or at least to put it in warm cement for three years, is not realistic in the type of world we live in," he added.
Consultants at Watson Wyatt Worldwide, Reigate, England, think Mr. Bernstein has struck a nerve. "For our purposes, we agree with his concern that policy portfolios have become a substitute for thinking," consultants wrote in a paper last winter.
"Funds have had had a false optimism that it is possible to have a policy that is designed for all weathers. The real benchmark should be the return required by the structure and timing of the liabilities," they wrote.
Watson Wyatt officials wrote that capital markets are subject to changing regimes that might last five, 10 or 20 years. This view abandons efficient-market theory, which says securities prices move up or down in a "random walk," showing no apparent pattern.
Instead, Watson Wyatt officials have embraced a concept of "rational beliefs," developed in the mid-1990s by Mordecai Kurz, an economics professor at Stanford University, Palo Alto, Calif.
Under this school of thought, investors possess different information and views on how the markets will unfold. Thus, one investor might be optimistic about the stock market's future, while another might be negative. In addition, their views will vary over time.
This lack of certainty about market prices explains far better than traditional finance theory why there is so much volatility in securities prices. It also explains why there can be lengthy bull or bear markets that are not explained by fundamentals.
This theory has three key beliefs for investors, according to the Watson Wyatt paper:
cMarkets tend to trade for long periods of time at "wrong" prices — those not justified by fundamentals — but eventually revert to the "right" level at some time;
cIncorrect prices can be exploited, but only long-term investors can exploit infrequent market corrections; and
cEquities deliver a better payoff for long-term investors on a risk-adjusted basis.
To exploit this theory, investors need to adopt what Watson Wyatt calls "dynamic strategic asset allocation."
This approach has three elements to it: setting a risk budget, where target risk levels (expressed as value at risk) and asset allocations should be expressed in ranges; revising the risk budget regularly, perhaps annually, to take into account shifting market dynamics; and quarterly or monthly rebalancing, for both the asset mix and risk levels.
Strategic Economic Decisions' Mr. Brock, a disciple of Mr. Kurz, said talking about long-term regime shifts misses the point.
In his view, adopting a dynamic approach eliminates the distinction between tactical and strategic asset allocation. A "robot" makes shifts in the asset mix, based on historical market patterns. "The system tells you when to get in and to get out," he said.
Mr. Brock's view is that there is enough "predictability" in the data to indicate how markets will likely perform. Traditional methods don't assume there are patterns in the data, but assume the likelihood of a change in market conditions is random.
In addition, investors will have different portfolios, depending on their risk tolerances.
Officials at Watson Wyatt and Mr. Brock have been accused of promoting market timing, but they insist their approaches are more involved.
"Dynamic strategic asset allocation is more than glorified market timing," said Roger Urwin, Watson Wyatt's global head of investment consulting.
"Market timing," said Mr. Brock, "is the wrong word." He defines market timing as an "alpha-adding strategy," while his strategy depends on shifting market exposures.