CHICAGO - DiMeo Schneider & Associates LLC has developed a portfolio management system that rebalances a portfolio when there's a shift in its risk/return characteristics, not just a shift in specific asset classes.
The proprietary model, called Portfolio Engineer, has been in the works for about a year, said William Schneider, managing director and co-founder of the Chicago-based investment consulting firm.
Mr. Schneider came up with the idea a couple of years ago as a way to take advantage of the wild swings within styles in the market. The model allows the winners to run a little bit, said Mr. Schneider, but not at the overall risk of the portfolio. Mr. Schneider, with the help of quantitative analyst Matt Rice, set out to identify an optimal percentage of risk and return for the entire portfolio, as opposed to setting percentage bands around specific asset classes, a policy most pension funds use.
The Portfolio Engineer model uses a theory of optimal percentage, or optimal R, which is a range the portfolio can drift from the target asset allocation before rebalancing is necessary. The optimal R is based on the portfolio's overall asset allocation, factoring in its return and standard deviation characteristics. For example, a particular asset class, say small-cap value, may be allowed to drift 12 percentage points from its target as long as the overall portfolio has not gone out the optimal range established by the consultant.
Mr. Schneider would not disclose the details of how the model determines a pension fund's optimal R. "What it does, is take into account the fact that the basic assumptions that we use to do an efficient frontier are flawed," said Mr. Schneider.
Higher returns, lower cost
The firm tested the method against a monthly rebalancing scenario and found returns were slightly higher, excess risk was slightly lower and rebalancing was needed less frequently. Rebalancing was only required 0.71 times annually, or once every year and a half, which results in lower transaction costs.
"It shows that you don't need to rebalance if there is a little bit of drift," said Mr. Schneider. "The question is, when does drift cause a real change in risk and return to the whole portfolio?"
The firm just rolled out the new process,. which Mr. Schneider said is unlike any other in the industry.
Most consultants look at asset allocation ranges for rebalancing, agreed Richard Michaud, founder of New Frontier Advisors LLC, Boston. However, that is slowly changing.
New Frontier, an investment adviser and consulting firm, developed its own rebalancing software that it markets to institutional investors. The New Frontier model, based on Mr. Michaud's patented resampled efficient frontier process, rebalances when a portfolio is determined to be non-optimal for a given level of risk using resampled statistical data.
A recent study by Greenwich Associates, Greenwich, Conn., found 60% of plan sponsors rebalance when asset classes drift from fixed trigger points; 23% have a tactically flexible policy that does not require rebalancing when asset classes drift out of target ranges; and 9% use the calendar method and rebalance at given time periods.
Ed Asam, director of capital markets research at CRA RogersCasey, Darien, Conn., said his firm uses a volatilty-weighted rebalancing system that has broader ranges for more volatile asset classes and narrower ranges for less volatile asset classes. The idea behind the strategy is that winners have a chance to run without increasing the overall risk of the portfolio, he said.