The Ann Kaplan professor of business, finance and economics at Columbia University, Andrew Ang, who also consults to the Norwegian sovereign wealth fund, describes the shortcomings of research on asset allocation and illiquid assets, and how to overcome behavioural biases.
Andew Ang’s paper, “Liquidating Harvard”, details the workings of the world’s largest endowment, its own catastrophe during the recent financial crisis, and importantly the choices it faced to meet the return shortfalls.
“I looked at Harvard, that’s a big name, but many have the same problems. The endowment model didn’t work because it didn’t have a floor for returns,” Ang says. “Harvard has one-third of its operating costs from the endowment so its choices were: it could liquidate the endowment, raise taxes, cut costs, or a fund-raise. It goes back to what’s the endowment for: is it a buffer, like a sovereign wealth fund. Harvard University chose to issue debt – $2 billion of it – and in doing so doubled the leverage ratio.”
The experience of Harvard highlights the implications of its various available actions on the portfolio management of the institutional investment pool, and the lack of research in the area of how to conduct asset allocation with a long-time horizon and illiquid assets.
“There is little research on how to do that, my paper is the first coherent framework,” he says.
“There is a general equilibrium question, do you invest in illiquid assets (or diversification); then given that how do you structure the portfolio to liquidate when you can’t.”
One of the key areas Ang highlights is a mean variance model that does not take into account that liquidity will give you an allocation to illiquid assets that is too high.
Investors in illiquid assets also lack control, he says, because the vehicles are complex, opaque, and investors have no right to take their money out.
“If you invest in equities you do have control, and you should you own that capital,” he says. “It would be good with illiquid assets to separate the monitoring, money paid to the manager etc, from the investments themselves but we are a long way from that. Where’s the balance of power? In liquid assets it’s with asset owners, with alternatives it’s with the talent.”
The other difficult area is in rebalancing, with a right skew inherent because of the difficulty in rebalancing.
Ang is a big believer in factor analysis and says the ‘ideal portfolio for a large investor should be factor oriented’.
“Investors do a disservice by putting asset labels on allocations. Instead they should conduct factor analysis, including looking at macro factors like inflation and growth, but also investment factors like active managers use such as value, growth, momentum, volatility, credit,” he says. “This decision should be made with the asset owner and there are some cheap ways to do this that can be in-house and automated.”
In fact Ang says allocating according to factors, and diligently rebalancing are two keys to successful portfolio management for institutional investors.
“It is naturally counter-cyclical to rebalance, that when an asset is doing well you need to sell. But the biggest mistake investors make is pro-cyclical, it’s when discretion gets in the way, when behavioural aspects get in the way,” he says.
“But you have to be brave and to buy what someone’s selling and vice versa, it’s very hard to do.”
He says the most simple counter-cyclical mechanism is to make sure investors buy low and sell high, and if processes, with well-defined events to override those are in place, behavioural tendencies can be overcome.
“It’s like with the Mississippi flooding, the army has criteria about whether to breach the level, which would mean flooding the area, people would die and towns would be ruined. The Army Corp has training and processes to deal with this, because the guy at the front may have a friend in that town and can’t make that decision,” he says.
Similarly he says, analysing factors and allocating assets according to that, as well as diligent rebalancing should be the base line for investors.
“If you can do better then do it, but you must have the base line in place. With short-term asset allocation deviations, unless they are done properly behavioural biases always creep in.”