The endowment model is in need of a new asset allocation framework, and two former Hewlett-Packard CIOs know how to fix it.
According to Ken Frier and Gretchen Tai, the Yale model’s shortcomings could be overcome with more dynamic asset allocation, diversification beyond illiquid assets, separation of alpha and beta, and a strong governance structure.
“Asset allocation decisions have more impact on future investment results than any other single choice that an investor can make,” the authors argued in a paper.
Instead of sticking to a steady—and often inflexible—policy portfolio, Frier and Tai suggested investors form a custom allocation model that can adapt to changing markets.
For example, using a combination of value and momentum factors could be a powerful way to predict asset class price and returns.
“Use value metrics to decide what is best to own or not own, and then wait for a confirmation from momentum before making a change to portfolio composition,” the paper said.
Value alone could help investors understand the expected return/risk relationship of an asset class, Frier and Tai wrote, especially during extreme market conditions. Momentum, on the other hand, could help investors lose less and even spot signs of market trouble, they continued.
“Avoiding just one really bad year can have an enormously positive risk/return impact, even in the context of an 18-year track record,” the authors wrote. “One doesn’t necessarily have to pay a hefty fee to hedge fund managers or deploy tail-risk hedging strategies to protect against downside losses.”
The combination of these factors—as well as others such as carry, defensive, and liquidity—could also help investors identify “superior managers” and gather substantial cross-asset information,” Frier and Tai said.
With this framework in place, investors should then diversify and add assets with low correlation to public equity. The former CIOs suggested going long markets with good value and momentum, and short ones with bad value and momentum.
Furthermore, investors can separate alpha and beta to take advantage of both dynamic asset allocation and manager alpha, Frier and Tai said.
It is possible, they continued, to “maintain a physical allocation with a well-performing manager,” (alpha) while adjusting the overall market exposure (beta) with derivatives overlay programs.
“External managers can be screened and selected primarily on the basis of the expected alpha to be generated instead of conducting manager search by asset class,” Frier and Tai concluded.
Related: How to Fix the Endowment Model & HP CIO Gretchen Tai Steps Down