The role of credit as a predictor of the economic cycle

The current crisis has clearly shown the need to deepen our understanding of the linkages between macroeconomic and financial variables. Indeed, the science of economics has been accused of being incapable of predicting the crisis, even though apparently unsustainable imbalances were building up during the preceding expansionary phase. In particular, there have been many references to rapid credit growth and how it has caused upward pressure on assets, particularly real estate. As a result of this experience, mechanisms to predict the economic cycle are currently being proposed that are based on using credit as a “leading variable” for the cycle, in other words, as an indicator containing information predicting the likelihood of a recession. It is also increasingly common for macroeconomic supervision by international institutions to use credit growth as a signal of macroeconomic imbalances or as a warning sign of future recessions. However, the statistical basis for this type of inference draws on studies carried out since the start of the current crisis. This work by both academics and analysts examines the behaviour of the economy drawing on the information available at each point in time, which is used to attempt to describe the past. It is a different question whether the information available at each point is time can predict the future, or, other words, what variables in real-time economic prediction models contain predictive information allowing turning points in the economic cycle to be detected. In this paper we present a summary of recent work addressing this issue. Specifically, our study aimed to determine whether, with the information available at a given moment in the expansionary phase prior to the crisis, it would have been possible to make a clear diagnosis of the economy’s situation at that time. And, above all, it asks whether changes in any of the financial variables (credit, in particular) would have been a useful predictor of the onset of recession. This paper starts by describing the combined evolution of credit and the economic cycle in several countries over the last few decades, and highlighting a number of empirical studies that have sought to base economic predictions on this information. It then discusses the origin of the difficulties involved in using credit as a leading indicator of the cycle, and puts forward an alternative methodological approach for the use of financial variables in models predicting the economic cycle.