Asset Allocation Post-Crisis
Coverage of the 2011 Global Investment Conference.
BY Sebastien Page | May 10, 2011
For the last sixty years portfolio theory has received both praise and criticism. To balance return against risk at the portfolio level is a worthy objective, but the models we use to achieve this objective can be misleading. The recent financial crisis reinforced the importance of a number of key developments in the field of asset allocation. As they seek robust post-crisis asset allocation, pension plans, foundations, endowments, and individuals should pay close attention to each of the following:
Shifting from Data-driven Models, Deploying Risk Factor Diversification
Too often, data-driven models ignore the current state of the world. For example, most data samples include periods of declining rates and moderate inflation. In contrast, over the next few years, rates may rise, and inflation might creep up. Similarly, while developed markets sovereign bonds historically have been viewed as safe havens, downgrade risk may cause them to have a much higher-than-historical correlation with risk assets going forward. In general, in the wake of the financial crisis, investors have recognized the importance of combining quantitative information with a qualitative, macro view. Historical data can be useful (after all, we don’t have future data), but only to the extent that it helps formulate a relevant view about the future.
The recent financial crisis has reinforced the notion that asset class returns are driven by common risk factors. Hence, risk factors – as opposed to asset classes – should be the building blocks for portfolio construction. They include, for example, interest rates, the slope of the yield curve, corporate bond spreads, equity returns, investment style returns (momentum, value, and size), changes in volatility, commodity returns and changes in liquidity. In his recent article titled “Beyond Risk Parity” (Journal of Investing, Spring 2011), Vineer Bhansali demonstrated that equity risk contributes 97% of the volatility in a typical 60/40 portfolio (modeled as 60% S&P 500 and 40% in Barclays Capital US Aggregate). Asset class diversification can often mask true risk factor diversification.
A More Dynamic Process
The traditional approach to asset allocation typically focuses on a three-to-five year horizon. The asset mix is optimized (usually in the context of an asset-liability study), tolerance bands are set and eyes are closed until the end of the horizon. Unfortunately, significant events may occur along the way. The number of times the earth completes a revolution around the sun has nothing to do with how abruptly expected returns and risks may change. Expectations about the future, and therefore optimal asset allocations, must take into account both the cyclical and the secular horizons. Investors with a rigid governance process and no clear mandate to follow a dynamic process, can outsource the process to external, multi-asset mandates, such as a global tactical asset allocation strategy.
Unforeseen market crises are often referred to as “tail risk events” because of the way they appear on the “normal”, bell-shaped curves often used to illustrate market outcomes. When investors allocate to the liquidity premium through real estate, private equity, and hedge fund investments, they implicitly sell an out-of-the-money put option on liquidity risk, which increases tail risk, because when the “put option” comes into the money in a severe market scenario, they are basically a forced buyer of liquidity when it has become extremely expensive.
Indeed portfolios that were heavily allocated to illiquid assets fared poorly during the recent crisis. When the option becomes in-the-money, the strategy must be able to survive the storm and pay up. In such markets, a strategic allocation to a tail risk hedging strategy may provide liquidity when it is needed the most.
In summary, the financial crisis has sparked a renewed skepticism of portfolio theory and financial engineering in general. Asset allocators should be aware of the pitfalls of a naïve approach to portfolio engineering that relies on the normal distribution and that fails to incorporate forward-looking views. By relying on these tenets, an approach to analyzing portfolios should help address these concerns and provide opportunities to build better-diversified portfolios.
Sebastien Page, CFA is executive vice president and head of Client Analytics with PIMCO.