Equity, credit and the business cycle

Abstract

Both domestic economies and financial markets are affected by cycles that are often represented through multi-state models such as Markov Switching (MS) models. This article discusses the performances associated to the government bond, the equity and the credit cases along the business cycle, using both an European and a US dataset over the 1987 to 2010 period. Periods of noninflationary growth have been strongly supportive to the credit universe, whereas inflationary growth has led to a strong performance of the equity asset class. On the contrary, recession periods are characterized by strong performances from government and investment grade bonds. These statements hold both in the US and in the European cases.

Keywords:

• asset allocation
• Markov switching
• density test
• equity
• credit

JEL Classification::

• G14
• G11

Notes

¹ Rabault (1993), Krolzig and Toro (1999), Filardo (1994), Durland and McCurdy (1994) or Lam (1997) perform such an analysis in the US case. In the European case, an MS approach is applied to the European business cycle in Artis et al. (2004) and Krolzig and Toro (2004).

² Here, I am interested in an accurate description of the cycle's regimes and not into forecasting these regimes, which is why I rely on this in-sample analysis and not on an out-of-sample one.

³ Each series has been tested for stationarity through an Augmented Dickey–Fuller (ADF) unit root test, that confirmed the stationarity of the data. Tests results are not reproduced here for the sake of space saving but are available upon request.

⁴ As the aim of this article is to relate the business cycle of each zone to the performances of the assets related to its business cycle, asset classes such as commodities or foreign exchange rates are excluded here. The commodities asset class can hardly be related to the US or the eurozone and the FX performances should be related to differences in economic activity and inflation – at least – between the domestic country and the rest of the world, which is beyond the scope of this study.

⁵ To this purpose, I simply estimate a two-regime MS model using the previous economic data. One regime is found to be an expansion regime and another a recession one, as found in the existing literature. Corresponding tables and figures describing these regimes are not reproduced here, for the sake of saving the space. They remain, however, available upon request.