Early Warning Indicator Model of Financial Developments Using an Ordered Logit – Introduction

Early Warning Indicator Model of Financial Developments Using an Ordered Logit - IntroductionThe recent financial crisis starting 2007 has shown that boom-bust-cycles can have devastating effects on the real economy. At least since the Great Depression, economists and policy-makers have become aware of the potentially damaging effects of large fluctuations in asset prices, such as equity and property prices. The recent experiences in the 1980s-1990s in Japan, in 2000s in Iceland as well as Ireland and other countries have confirmed that, in some circumstances, boom-bust-cycles in asset prices can be very damaging as they may lead to financial and ultimately to macroeconomic instability.
Against this background, it is important to have indicators to assess the possible implications of large asset price movements and the building up of financial imbalances in the economy. In this respect, several recent studies have shown that the analysis of monetary and credit developments may be very useful (see for example Borio et al. 1994, Adalid, Detken 2007, Gerdesmeier, Reimers, Roffia 2010, 2011). There are, in fact, several reasons why monetary and asset price developments tend to be positively correlated. One reason is that both sets of variables may react in the same direction to monetary policy or cyclical shocks to the economy. For example, strong money and credit growth may be indicative of a too lax monetary policy which leads to the creation of excessive liquidity in the economy and fuels excessive price changes in the asset markets. Moreover, there can be self-reinforcing mechanisms at work. For example, during asset price booms the balance sheet positions of the financial and non-financial sectors improve and the value of collateral increases, permitting a further extension of the banking credit for investment which may reinforce the increase in asset prices. The opposite mechanism can sometimes be observed during asset price downward adjustments.
Most of the studies have in common that the development of the financial indicator is mapped into a bivariate variable. It gets a unity for boom or bust periods and a zero elsewhere. This paper contributes to the literature on the early warning system that it maps the financial indicator development into three phases: booms, normal period and busts. It extends the works of Gerdesmeier, Reimers and Roffia (2010) analyzing busts and Gerdesmeier, Reimers and Roffia (2011) investigating booms. To forecast the movement of the variable an ordered logit-type approach is applied. This approach uses the higher information content of the variable than a multinomial logit model suggested by Bussiere and Fratzscher (2006) which explain the pre- and post-crisis as well as the normal periods. Explanatory variables are, for example, lending boom, interest rate spread and house price growth. Moreover, it characterizes the comprehensive development of the indicator from boom over normal to busts periods. The description of the whole boom-bust-cycle is more general than the approach of Singh (2010) who uses an ordered probit approach, however, defines two phases of downturn.
The paper is structured as follows. Section 2 briefly summarizes the available evidence on the indicator properties of money and credit for detecting asset price imbalances, with a focus on the most recent contributions. Section 3 briefly describes the data used for the empirical analysis and describes the criterion to define an asset price phases. It also presents some results based on an ordered logit-type approach, using the pooled estimation procedure. In Section 4 we present some robustness checks of the model, and Section 5 draws some conclusions.