Rules-based adaptive approaches to multiasset-class investing can be designed that protect investors from potentially devastating losses while participating when market conditions are more favorable. Such an approach is intended to adapt to shifts in risk appetite, financial markets and the real economy. Based on history, all of us know that markets do not behave “efficiently” at all times. Sometimes investors benefit from owning risky assets (e.g., equities, real estate investment trusts, commodities, high-yield bonds and emerging market debt), while at other times (think 2008) holding these assets generates significant losses.
A first step toward developing this framework is to examine the history of annual returns, risks and correlations for a series of asset classes. These asset classes include equities, bonds, commodities, REITs and cash. A second step is to identify the macroeconomic and financial market factors that favor or work against each of these asset classes. In our opinion, these factors comprise the following data series: (1) measures of market sentiment, (2) interest rates, (3) household and business balance sheets, (4) real economic factors and (5) asset prices.
To illustrate how one might design this framework, in the chart below we construct a simplified regime framework that uses three data series: GDP, the VIX and the S&P 500. The chart below standardizes and plots quarterly data for these three indexes.
In this chart, we discern key relationships. Namely, when GDP growth is strongly positive and the "fear index" (the VIX) declines, the S&P 500 index generally increases in value. However, when the VIX increases and GDP growth slows, the S&P 500 tends to decline in value (the tech bubble in 2001-'02 and the financial crisis in 2008). A regime framework can be constructed that adapts to these changes in market conditions.
This data can be segmented using a statistical process that generates a framework consisting of five regimes. These are labeled: (1) safety, (2) add risk, (3) adjust risk, (4) fade risk and (5) neutral. The table below ranks the Sharpe ratios by regime for the 10 asset classes using historical returns since 1990.
The next step is to allocate the portfolio, which can be accomplished by imposing some rules. For example, the allocation can be determined by dividing the historic Sharpe ratio for each asset class by the sum of all Sharpe ratios for asset classes in that particular regime. If U.S. equities has a Sharpe ratio of 0.50 and the total for all Sharpe ratios in that regime is 2.50, the U.S. equity allocation would be 20%.
In addition, these rules should factor in valuations. For example, an adjustment factor can be applied to the Sharpe ratio for an asset class if it is over- or undervalued. And there may also be reason to hold a cash allocation in a specific regime (e.g., safety), given opportunity costs and other considerations. Implementation of the portfolio can be achieved using highly liquid, low cost exchange-traded funds.
Importantly, proper regime segmentation will drive success or failure. In the safety regime, only gold and U.S. Treasuries generated positive performance. Alternatively, in the risk-on and adjust-risk regimes, virtually all asset classes generated positive performance. In applying this process to history, it is important to see clear distinctions between the various regimes. If they all perform more or less similarly, this will defeat the intended performance objective, which is to capture periods of under and outperformance.
An adaptive, regime-based framework, if properly constructed, should provide protection in years such as 2008, while adding value when market conditions are more favorable. This framework should be designed to capture the evolving relationship between the real economy and the financial markets. Given the uncertainty that exists about future market performance, an adaptive approach offers a very sensible investment alternative to a broad swathe of investors, who seek to add incremental value while protecting capital.
John Balder is co-founder of Investment Cycle Engine Inc., Newton, Mass.