Decision-making under varied economic regimes

It is a well-recognized empirical observation that different asset classes respond differently to different economic drivers. For example, fixed-income assets tend to respond to anticipated movements in interest rates, among other factors; bond prices fall when interest rates rise. Commodities respond to, and sometimes drive, inflation expectations; commodity prices can rise fast when inflation expectations are rising, and they can fall quickly once inflation appears to have peaked.

It is also well-recognized that asset class behavior can vary significantly over shifting economic scenarios. For example, business cycles tend to affect cyclical and non-cyclical companies in markedly different ways, primarily because of sensitivities of consumers and producers to economic growth.

Yet, while asset class performance certainly varies with changing conditions, traditional asset allocation approaches make no effort to adapt to such shifts. Instead, traditional approaches seek to develop static “all-season” portfolios that optimize efficiency across a range of economic scenarios.

In this piece, we consider a compelling alternative. We define economic trends in their totality as a single complex system that changes over time to produce what we call economic “regimes.” And we investigate whether regime-based asset allocation can effectively respond to economic regimes at the portfolio level to provide better long-term results when compared against static benchmark-based approaches.

Regime-based investing is distinct from tactical asset allocation. While the latter is shorter-term, higher frequency (daily or monthly) and driven primarily by valuation considerations, regime-based investing targets a longer time horizon (a year or more), and is driven by changing economic fundamentals. It straddles a middle ground between strategic and tactical. A regime-based approach is designed to give investors the flexibility to adapt to changing economic conditions within a benchmark-based investment policy.

As a preface to our complete work on this topic, we offer the following preview of how regime analysis may be implemented to improve portfolio results. We also provide insights into the key benefits and challenges associated with implementing such an approach.

Step 1: Develop the relationship between economic factors and financial markets

In developing these relationships, we find that economic regimes can be broadly defined in terms of four key factors, which tend to dominate financial market performance. These factors are economic growth, inflation, monetary policy and labor market slack. Our research indicates that developing insight into near-term changes in these four factors — rather than their absolute levels — can provide an effective framework for executing a regime-based asset allocation policy.

We also find that regime-based investing requires understanding the state-dependent relationship between financial markets and the broader economy and devising a method of modeling the non-linear nature of such relationships. Modeling such relationships is quite complex for both conceptual and practical reasons. For example, a simple scatter plot of S&P 500 returns and U.S. real gross domestic product growth may not at first reveal much of a relationship, at least not a clearly linear one (Exhibit 1). But advanced non-linear statistical techniques can help in identifying and defining this relationship. While prospects for economic growth improve, equity prices tend to rally. Beyond a certain threshold, however, the relationship starts to break down.

Of course, asset class performance can be driven by more than just economic fundamentals. Changes to valuation (e.g. price-earnings ratios for equities) are also often a significant driver of performance. It is possible for financial markets to reflect extreme optimism or pessimism (from a valuation perspective) for long periods of time — rather than pure economic fundamentals. In such cases, regime-based frameworks may prove inadequate for the purposes of developing robust and resilient portfolios.

Step 2: Model the behavior of different asset classes in different economic regimes

Once we have developed relationships between our economic factors and asset classes using our non-linear framework, we can model the regime-dependent returns of various asset classes. Exhibit 2 outlines six possible economic regimes defined in terms of our four key factors and ranks five asset classes for their relative performance potential, from best to worst, within those regimes.

(Click here for Exhibit 2)

It is clear from this analysis that shifts in asset class leadership are so broad and varied that no static portfolio weighting could be optimal across all regimes. Just on an intuitive basis, regime-based asset allocation appears to be the most logical response to shifting economic regimes.

Step 3: Assess the impact of different economic regimes at the total portfolio level and optimize portfolio allocations depending on economic insight and the investor's risk constraints

Our finding is that successfully developing and executing a regime-based asset allocation strategy does not require perfect economic forecasting skills. Yet even imperfect economic foresight — which we define as forecasting only directionality rather than exact magnitude of economic changes — is not necessarily easy to achieve given the confluence of factors affecting the economy. These influences include — but are not limited to — changing patterns in consumption, savings, investments, taxation and fiscal policy on the broader economy. In our complete research, we analyze the effect of various degrees of economic foresight using the concept of average information coefficient.

However, even with perfect economic foresight, asset class response can be extremely difficult to capture. This is particularly true when the economy and financial markets experience new paradigms, relative to history. In such circumstances, the relationship between economic factors and financial markets can change quickly, leading to underperformance of a regime-based investing approach developed on historical data. So, it is important not to underestimate the complexity of the challenges of foresight, or overstate its power to capture returns.

For a hypothetical diversified portfolio, regime-based asset allocation has the potential to significantly increase portfolio efficiency. Exhibit 3 compares the results of a static benchmark-based policy with those of a regime-based policy, under the regimes outlined above.

The hypothetical portfolio is most vulnerable to “tail” scenarios of a severe recession or an overheated recovery with inflationary pressures. Purely measured by portfolio returns, the regime-based portfolio produces superior results.

It is important to note here we do not advocate abandoning benchmark-based investing. Institutional investors set their portfolios' strategic benchmarks based on a desire to meet liabilities and other important strategic goals. Within these broad objectives, however, we argue that investors may be handicapping their portfolios by being regime agnostic — which is what a strategic benchmark is.

Instead, we believe that investors would benefit from being “regime aware” and allowing themselves the freedom to adjust allocations around the strategic benchmark in response to shifts in economic regimes.

We believe, based on our analysis, that regime-based investing can offer a compelling alternative to the static “all season” approach. n