Scientific Beta: Estimating macroeconomic risk in equity portfolios

June 13, 2023

By Mikheil Esakia, Quantitative Research Analyst, and Felix Goltz, Research Director, both working at Scientific Beta

In research published earlier this year^[1], we presented a methodology to estimate stock-level exposures to various macroeconomic risks. Investors may be interested in harvesting the equity premium, while benefiting from rising or falling interest rates or from rising or falling inflation, for example.

While such targeted exposures are attractive for investors, a major challenge in designing such equity portfolios is the reliable measurement of exposures to macroeconomic risks.

Investment practice does not have a convincing answer to this challenge. Investment managers mainly rely on off-the-shelf building blocks, such as sectors or style factors to manage exposure to macroeconomic risks. However, such building blocks have never been designed to efficiently target macro risks.

Moreover, strategies that target macroeconomic risks are often discretionary. For example, there are actively managed funds targeting strong outperformance in periods of high inflation. However, the lack of transparency and replicability of discretionary approaches makes them hard to evaluate.

Forward-looking variables and macroeconomic surprises

Our methodoloy relies on innovations in forward-looking variables that quickly incorporate investors’ expectations about economic conditions, such as short-term interest rates, term spread, credit spread and expected inflation.

Moreover, our methodoloy relies on surprises in macroeconomic variables, instead of levels. The level of a macroeconomic variable that was fully anticipated by investors will not lead to different price reactions across stocks. Using surprises in such fast-moving variables is important in order to identify a reliable relationship between stock returns and macroeconomic variables.

Our robust measurement procedure

To capture exposure of stocks to these macro surprises, our methodoloy applies robust measurement tools from statistics and textual analysis. The resulting exposure measures allow the construction of equity strategies that come with statistically and economically significant exposures out of sample.

Our robust measurement procedure delivers a more robust and reliable classification of stocks than standard estimation approaches. When constructing equity portfolios that target high or low macroeconomic exposure, our procedure leads to 50% higher realised out-of-sample exposures and doubles the t-statistics associated with these exposures.

Capturing macro exposures

Our approach of building dedicated portfolios using stock-level exposures captures macro exposures more reliably than using off-the-shelf building blocks, such as sector or factor indices.

Our methodology shows that allocating across sectors or factors leads to weaker exposures than our dedicated macro portfolios, and sometimes even leads to exposures that are inconsistent with the objective. The reason behind this is that sectors or factors are not designed to discriminate between assets by macroeconomic risk exposures, while the stock-level approach enables the heterogeneity of macro exposures in the cross-section of equities to be fully exploited.

In addition, our proposed strategies are fully systematic and do not rely on common beliefs or the discretionary choices of an asset manager.

Finally, our methodology shows that our macro exposure strategies do not suffer poorer performance compared to a broad equity portfolio. The standalone returns of eight macro exposure strategies as well as their Sharpe ratios are not significantly different from the market portfolio in our sample. They also do not come with negative alpha in a multi-factor model that includes the usual style factors. 

Opportunities for investors

Our methodology opens several opportunities for investors. First, using our measurement techniques, investors can inform themselves about the macroeconomic risks they are exposed to. In fact, many equity investors may be substantially exposed to macroeconomic risks without intention and our method offers a way to detect such hidden exposures in their portfolios.

However, our methodology can be used beyond such ex-post analysis. Since it allows macro exposures to be measured with reliability out of sample, such exposure measures can be used to construct portfolios that target exposures in line with investor preferences.

Such dedicated macro portfolios allow investors to protect themselves from sudden changes in economic conditions, such as surprises in interest rates, credit risk, or expected inflation. These strategies come with market-like performance in the long term, but provide outperformance when surprises in respective macro variables are in the direction that they target.

Towards a systematic and transparent methodology

The need of investors to target well-defined macroeconomic exposures is readily apparent from the amount of attention that investors devote to topics such as rising or falling inflation, changes in interest rates, shifts in the term spread, or variations in the credit spread. Addressing such needs with strategies that are discretionary and thus hard to evaluate – or with standard building blocks that were never designed to achieve such objectives – does not appear satisfying. 

Our contribution is to propose a methodology for portfolios that target macro exposures by design and that do so in a systematic and transparent way.

^{[1] Esakia, M., and F. Goltz, Targeting Macroeconomic Exposures in Equity Portfolios: A Firm-Level Measurement Approach for Out-of-Sample Robustness, Financial Analysts Journal, Volume 79, Issue 1 (2023)}