JPMorgan Asset Management officials are rethinking asset allocation, tossing out the traditional mean-variance optimizer in favor of a new approach that assumes returns don't fit neatly on a bell-shaped curve. It also incorporates downside risk.
Experts have been pointing out flaws in the optimizer for years, but JPMorgan executives are at the forefront in coming up with an alternative.
JPMorgan's approach rejects the conventional wisdom that returns are normally distributed on a bell-shaped curve and that returns are independent of each other. By assuming that “fat-tail” events — such as the 1998 Russian debt default crisis, the bursting of the tech-stock bubble in 2000 and the current subprime mortgage disaster — happen more frequently than expected, using standard deviation as a measure of risk gets thrown out the window.
Instead, JPMorgan's research assumes a non-normal distribution of returns and incorporates a measure of downside risk called conditional variance at risk, defined as the average real loss (or gain) in the worst 5% of cases in a simulated return distribution.
JPMorgan is not alone in seeking a better solution to the asset allocation puzzle. Officials at the $175 billion California Public Employees' Retirement System, Sacramento, are considering using downside risk in the fund's upcoming asset allocation review (Pensions & Investments, March 23). Separately, some managers — including Pacific Investment Management Co. and Acadian Asset Management Inc. — incorporate tail risk in some of their strategies.
“The Morgan work is very, very thoughtful and it's provocative. It goes along with other work that is being developed to start the discussion of how modern portfolio theory needs to be revisited in light of (changes to) both our regulatory and investment environment,” said Cynthia F. Steer, managing director, chief research strategist and head of beta research, Rogerscasey Inc., Darien, Conn.
“Traditional finance theory is taking a beating here,” said Kathleen DeRose, senior managing partner and head of portfolio management and research at Hagin Investment Management, New York. “If you're using a risk model that comes from the old theory, and the old theory has been turned upside down, chances are you will have to remake your risk model,” she added.