The impossible trinity and financial stability. The incidence of trilemma regimes on the (in)satability of stock markets and credit aggregates (1922-2013)

Abstract

This dissertation discusses the accumulation and unwinding of financial imbalances in stock markets and credit aggregates in advanced economies since the interwar years until the present. The project is framed within the literature that arose with the Great Financial Crisis (2007-09) and that, against the mainstream consensus at the time, posits the existence of a financial cycle best characterised by the joint evolution of asset prices and credit. As a theoretical framework, we build on the open macroeconomy trilemma which indicates that policymakers can choose two out of three desirable and competing goals: fixed exchange rates, free capital flows, and autonomous monetary policy. We construct exchange rate and capital control regimes, for twelve advanced economies, from several sources to identify whether the behaviour of the boom-bust cycle in asset prices and credit growth and the drivers of financial instability are contingent on changing institutional conditions. Additionally, we test for the existence of a global financial cycle, proxied by the United States, which may be driving the evolution of both dependent variables. In the first part of the dissertation which covers the literature review (chapter 1) and the data and methodology (chapter 2), we tackle the issue of measuring financial imbalances. To shift from the usual framework of financial crises towards the study of financial stability, we develop a new indicator for expansions and contractions in financial time series: the Local Bull Bear Indicator (LBBI). This technique, which constitutes the main methodological contribution of the dissertation, identifies above or below trend risk-adjusted returns, expressed as standard deviations. It allows for the identification of expansions and contractions to different time horizons, according to the persistence of shocks. As a data contribution, we offer a dating and characterisation of boom and bust phases in stock markets and credit aggregates, to different time horizons, for all the countries in the database. Regarding the empirical findings, the second part of the dissertation presents panel data evidence using a panel corrected standard errors model (PCSE) for stocks (chapter 3) and credit (chapter 4). First, regarding the literature on global financial integration, we find evidence of a global financial cycle proxied by the US stock market and credit aggregates. Secondly, we find that the relationship between the dependent variables and a selection of control variables and trilemma-associated covariates is contingent on the way in which countries resolve the macroeconomic trilemma at different times. Finally, regarding financial stability, we find that beyond price stability, the variables that drive LBBIs are contingent on the way countries resolve the macroeconomic trilemma at different moments in history. The third part of the dissertation presents time-series evidence for two case studies on the United Kingdom. Chapter 5 discusses whether the evolution of the stock market is contingent on domestic policy or trilemma regimes and finds that the usual story of global financial integration following a U-shape with peaks during the gold standard and the present-day inflation targeting regime requires nuance. The long-run co-movement between the stock market in the UK and the global cycle seems to peak in periods of heightened systemic risk such as the Second World War and the two oil shocks (1973, 1978). Finally, Chapter 6 finds that there is a financial cycle in the UK and that the causal relationship between stocks and credit for the UK is both time-varying and contingent on the regulatory framework in place. As the financial system becomes unregulated, credit growth seems to drive the behaviour in asset prices, allowing us to conclude, preliminarily, that while price stability is important, policymakers should include the evolution of credit aggregates in their policymaking function.