The authors make the case for the use of a macroeconomic indicator based in the real economy, the ratio of world capital to output, as an accurate way to anticipate both equity and fixed-income returns across multiple developed markets. Their findings provide further evidence of market convergence, with implications for traditional strategies of diversification.
The authors use the ratio of world capital to output (i.e., capital stock, physical stock of capital to GDP = k/y), a production-based measure, as a diagnostic proxy for the global macroeconomic business cycle and find this variable has significant predictive power for stock market returns across a group of major industrial economies. Unlike a number of other financial market–based predictors, such as the world dividend price ratio and the risk-free rate, k/y exhibits strong out-of-sample forecasting strength in almost all countries studied. It also shows utility for projecting key dimensions of fixed-income markets, such as bond returns, interest rate changes, and credit spreads.
How Is This Research Useful to Practitioners?
As portfolio managers have increasingly incorporated holdings from outside their home markets, in concert with growing economic globalization, international measures of business conditions have become more relevant for investment decisions. The authors argue that world k/y is a useful tool under such evolving circumstances, offering a fairly robust 2+% excess return forecasting advantage relative to the customary alternative estimate of historical averages. Its countercyclical nature parallels the time-dependent behavior of the equity risk premium. In addition, it is smoothly integrated into established conditional asset pricing models, improving, for example, the cross-sectional performance of the international capital asset pricing model.
World k/y is versatile; it is also applicable in debt investing. There is evidence regarding the predictability of time-varying risk premiums in fixed-income markets that illustrates a degree of international macroeconomic integration. Furthermore, world k/y is an output-based measure better suited for markets lacking the detailed consumption-based data available in the United States. Although some US analysts have derived useful models of returns that rely on investment-to-capital ratios, the authors caution against depending on aggregate investment data. Such data are susceptible to stock mispricing and/or long lags, whereas world k/y lacks these deficiencies.
How Did the Authors Conduct This Research?
The authors first calculate k/y by using Organisation for Economic Co-Operation and Development series data (lagged one quarter) of the natural logs of the quarterly real capital stock of the business sector divided by real GDP, both denominated in US dollars. The sample period is 1970–2010 and thus covers multiple business cycles. The analysis is focused on the United States, the United Kingdom, Japan, Italy, France, Canada, and Switzerland because of their continuous availability of data on the physical stock of capital.
Of course, given the lengthy period under study, many currently significant and growing economies (e.g., Brazil, India, and South Korea) are not included. So, the applicability of the findings to these emerging market economies as they fully mature into developed status over time remains to be seen. A series of regressions is run to determine how effectively stock returns in these countries can be predicted using k/y (applying Morgan Stanley Capital International stock price and dividend data) and to determine out-of-sample forecasting power. The latter also serves as a check against the possibility of a small sample bias (i.e., too few countries in the sample) coloring the results.
For both the equity-oriented and fixed-income applications of k/y, the authors identify some proportion of country-level risk premium variability linked to international business conditions. This variability reflects tangible economic integration across both geography and asset class, which can have consequences for the expected benefits of diversification.
Identifying a forecasting tool with both accurate and stable predictive power that is not prone to trend following (and is also soundly rooted in economic theory) would be perceived by many within the community of CFA charterholders as having real value. The authors present a strong case that k/y can track variations in expected returns of financial securities across developed economies and business conditions; they thus give it the status of an indicator worth watching regularly.