State variables, macroeconomic activity, and the cross section of individual stocks | Researcher: Martijn Boons
There is a long-standing tradition in both academia and practice to use state variables, such as the dividend yield, the default spread (between corporate and government bonds), the term spread (between long- and short-term government bonds), and so on, as indicators of future consumption-investment opportunities.
Finance theory, going as far back as Nobel laureate Robert Merton’s intertemporal capital asset pricing model, suggests that if these state variables do really predict something meaningful about the future, that is, asset returns or economic growth, investors should care about the exposure of their stock portfolio to these state variables. Indeed, suppose a high default spread is bad news for the future. Then, a stock that pays off when the default spread increases is attractive as a hedge and investors will pay a premium to invest in this stock.
Tying the role of state variables to economic fundamentals
This intuition is the focus of the research paper by Martijn Boons (Nova SBE) forthcoming in the Journal of Financial Economics. The author sets out to shed light on a recent discussion of what exactly defines bad times or worsening consumption-investment opportunities. This is important, as the set of potential state variables has grown considerably, leading Fama (another Nobel laureate) to suggest that the intertemporal capital asset pricing model is a fishing expedition, where state variables are routinely used in models without tying their role back to economic fundamentals.
The author proposes a simple test to resolve this problem:
- The first part of the tests consists of asking which state variables are able to predict macroeconomic activity in the time series.
- The second part of the test consists of asking whether the state variables that do predict in the time series capture a risk premium in the cross section of stocks (and, whether the variables that do not predict, do not capture a risk premium).
For a large set of candidate state variables, the author finds such time-series and cross-sectional consistency. For instance, the default spread and term spread are found to predict macroeconomic activity (measured using industrial production growth in the US) with a negative and positive sign, respectively. Consistent with the hypothesis outlined above, investors are found to pay a premium for stocks that are exposed to the default spread, which allow investors to hedge, whereas they pay lower prices for stocks that are exposed to the term spread, which are not attractive to hedge.
This article sheds light on how prominent state variables used widely in the models of academics and practitioners are priced in the stock market. Measuring exposure to state variables is useful in practical applications, for instance, when investors wish to tilt their equity portfolio away from macroeconomic risk to hedge or toward macroeconomic risk to capture a premium. For academics, the paper represents a late, but certainly welcome validation of fundamental economic theory.
This article is based on the working paper accessible here
Know more about Martijn Boons, Assistant Professor at Nova SBE.