Bernstein amends policy portfolio mantra

In a case of right church but wrong pew, Peter Bernstein says his doctrine wasn’t fully developed when he said earlier this year that institutional investors should throw out the policy portfolio and return to market timing.

The problem was that Mr. Bernstein was on the right track in his hugely controversial speeches and papers, but he now realizes he didn’t ask the right question.

Investors must ask “which `policy’ a policy portfolio is designed to fulfill,” he writes in his Aug. 15 client newsletter. The answer, says the noted economic historian and financial consultant, is simple: The goal of any investor, institutional or individual, is to fund a stream of liabilities.

Thus Mr. Bernstein is joining a growing camp of pension experts who argue that pension assets should be invested in a way that aligns with pension liabilities, using financial engineering techniques such as portable alpha.

Robert Litterman, managing di- rector in charge of the quantitative-resources group at Goldman Sachs Asset Management in New York, says, “Where he’s come out now is much closer to where we are.”

Instead of creating a policy asset allocation based on return forecasts, the investor must figure out a mix “with the highest probability of being able to pay for the groceries when the time comes,” Mr. Bernstein states in his newsletter.

Not recanting

He isn’t, however, recanting his earlier view that the conventional policy asset mix should be tossed out the window. “When we never know what the future holds, how can we kid ourselves into believing we can estimate long-run rates of return and co-variances for a broad menu of financial assets?” Mr. Bernstein writes.

He credits Kevin Kneafsey, an investment strategist at Barclays Global Investors in San Francisco, with helping him come to his recent epiphany. “Thanks to the work of others … we now understand what we have been talking about,” Mr. Bernstein explains.

Like many others, Mr. Kneafsey takes issue with Mr. Bernstein’s earlier contention that the policy portfolio is obsolete. “I read his piece and just thought it didn’t jibe with what we were thinking,” Mr. Kneafsey explains. So he started writing down his thoughts and passed them on to Patricia C. Dunn, BGI’s vice chairman, who is a friend of Mr. Bernstein’s.

To Mr. Kneafsey’s surprise, Mr. Bernstein not only acknowledged his paper but said it was “must reading.”

In “Solving the Investor’s Problem,” Mr. Kneafsey agrees with Mr. Bernstein that “the time is ripe for change, and it is time for apocalyptic thinking about the policy portfolio.” Mr. Kneafsey argues, however, that the policy portfolio shouldn’t be abandoned but rather constructed and used differently.

Now investors pursue a “two-hats” policy: The policy portfolio attempts to both hedge the liability and shoot for growth above that liability, he says.

First, investors calculate liabilities. Then they construct a policy portfolio designed to maximize returns while minimizing contributions at tolerable risk levels. Finally, they find investment strategies to fill the portfolio.

However, the traditional policy portfolio hangs on the long-term financial forecasts that Mr. Bernstein finds unacceptable.

But Mr. Kneafsey concludes that market timing is the wrong solution because it involves too few choices: stocks versus bonds, domestic versus international, and traditional assets versus alternatives. The fundamental law of active management says the information ratio – the ratio of alpha per unit of active risk – increases as skill increases and the opportunity to apply that skill rises, he notes. With market timing, there aren’t enough ways to add value consistently.

Instead, Mr. Kneafsey writes, investors should adopt a “one-hat” approach. While they should first determine the liabilities, the policy portfolio should become “a pure liability hedge,” not geared toward producing growth.

Carrying over alpha

Next, investors should “find in vestment products expected to deliver alpha irrespective of the policy portfolio,” Mr. Kneafsey states.

Using derivatives or structured products, investors can neutralize the market exposure from sources of alpha, effectively carrying the alpha on to desired asset classes.

Mr. Kneafsey writes that the one-hat approach offers better beta because it is focused on the sole task of hedging liabilities.

What’s more, he continues, it produces better alpha – the excess above the benchmark return – because the alpha is no longer tied to the underlying beta, or market exposure.