Can Private Asset Allocations Lead to Privation?

Strategic asset allocation is not static asset allocation; how are investors managing embedded beta risks in private asset portfolios?

Arun Muralidhar (Photo by Princeton Headshots)

Memory is short in asset management, and often that can be a good thing, but more often than not, it is not. As Mark Twain noted, “A clear conscience is the sure sign of a bad memory.” The rising allocations to private assets in institutional portfolios globally would not normally be a reason for concern, except that investors seem to have forgotten the damage that these allocations did to Ivy League (or “Yale model”) portfolios in 2008 and seem somewhat oblivious to the embedded beta risks in these investments.

To jog the memory of those of us who experienced 2008, and to share the history with those who did not, high allocations to private assets, with little to no risk management associated with the allocations, led to major cash flow issues as equity markets tanked in 2008. The global financial crisis led many asset owners into either fire sales, having to borrow in capital markets or doing incredibly silly things—like selling cash equities, putting on futures to free up margin cash without realizing that they were levered and long (this actually happened at a major endowment). It led to privation—professors were terminated, student aid was reduced, department budgets cut and capital investments curtailed. In the pension world, this led to a reduction in funded status and to increased contributions at the worst possible time.

Why have institutional investors flocked to private assets? The generous argument is that they get higher returns and diversification benefits. The more rigorous argument, made most emphatically by Oxford University’s Ludovic Phalippou and also by Nobel laureate Myron Scholes, is that on a risk- and fee-adjusted basis, investors could get better results and liquidity from investing in the Russell 2000 index.

A correlation argument is fallacious, as private assets are marked-to-market at the discretion of the asset manager (ignoring the governance issues that arise when the manager is marking your assets), and so investors cannot calculate correlations with this return stream and public assets without intentionally deluding themselves.

The more reasonable argument is that the infrequent marks help protect the portfolio valuation from excessive volatility—plain and simple! Interestingly,  Cristina Chen-Oster, founder and CEO of valuation provider M2M Capital, is offering clients an impressive method to get frequent marks for these assets (full disclosure—I am an adviser to her enterprise), but I fear she will not be successful, as no one from the CIOs to the regulator probably wants this transparency and frequent marking of assets. The rumors I hear offline about secondary deals on good private assets going at 30% below the marked value do not bode well for getting a true market valuation on 30% of one’s portfolio, especially if the portfolio includes some dud assets (like the dead unicorns)!

If CIOs believe we are not on the brink of another crisis, then they are choosing to ignore the data: the rising tri-polar geo-political risk, slowing economies with reasonably persistent inflation, bloated central bank balance sheets, an imploding commercial real estate market and large illiquid allocations limiting liquidity provision in a crisis. A recent lunch with four extremely smart CIOs confirmed this concern.

But “risk measurement is not risk management,” as I wrote back in June 2012 for CIO. The biggest risk in these private assets is still beta—probably 90%, given the leverage in these assets. You can be sure that the private asset managers are not managing this risk, as they are not paid to focus on the risk management of the beta of the asset—they are in the business of finding non-listed companies and turning them around and selling them for a profit—an alpha-oriented trade. So the responsibility to manage private asset beta is on the pension funds and endowments that invest in them, and to not do so is, in my humble opinion, terrible governance. Having talked to representatives of pension funds and endowments in Canada, the Netherlands, Switzerland and the U.S., I am genuinely concerned about the systemic risk of massive shares of portfolios—amounting to billions of dollars—being left un(risk)managed.

What should an innovative fund do? Not increase leverage, as we have seen some funds implement. Before even initiating these investments, funds should have a process in place to map their private asset exposures to liquid futures contracts. Current software offers risk measurement parameters like value-at-risk or exposures to weird factors like growth or inflation; since risk management is about tilting portfolios dynamically, investors want to emphasize the beta exposures/futures that are likely to rise and sell/hedge these exposures when they are likely to decline. Strategic asset allocation is not static asset allocation. Markets are dynamic, so doing nothing with the portfolio is a tactical view too; namely that the beta exposure implied by the decisions of the external managers are appropriate.

The most efficient way to do this includes using liquid futures—where the data is freely available—and removing investors’ dependence on any third-party risk measurement providers. San Diego County implemented this approach in 2008 and had a net cash inflow of close to $50 million, while the Ivies were struggling because they shorted the Russell 2000, thereby liquefying illiquid assets by extracting beta when it was bad. That’s as simple as it is. CIOs can do the same with private credit and other illiquid assets.

The naïve investor would argue that this is market-timing and tactical and is an impossible process to implement. All investing is market-timing and tactical. To assume that private assets will return 9% per year over 10 years is a tactical call, and doing nothing over the next 10 years to manage the beta is market timing. Moreover, since most assets lose money at least 45 to 47% of the time, to do well, especially in bear markets, requires investors to be right about 55% to 58% of the time, hardly a difficult bogey for well-compensated and trained CIOs to overcome.

CIOs like Don Pierce at San Bernardino County, who has added more than $1 billion by rebalancing intelligently with a robust internal model (and implemented with futures), demonstrates that with the right effort, intention and governance, these results can improve returns and risk management.

To not do something along these lines is to succumb to Mark Twain’s other warning: “There are two times in a man’s life when he should not speculate: when he can’t afford it and when he can.”

Dr. Arun Muralidhar is co-founder of Mcube Investment Technologies LLC (www.mcubeit.com) and of AlphaEngine Global Investment Solutions LLC (AEGIS).

This feature is to provide general information only, does not constitute legal or tax advice, and cannot be used or substituted for legal or tax advice. Any opinions of the author do not necessarily reflect the stance of ISS Stoxx or its affiliates.

 

Tags: , , , , ,

«