Australian National University

A Simple Framework for Analysing Bull and Bear Markets

Email address: adrian.pagan@anu.edu.au

Abstract:
The behaviour of asset prices has always been at the center of academic, media and business attention. A study of the DGP's of asset prices is important. Economic models of these DGP’s are beginning to emerge e.g. the study of Gordon and St-Armour (1998) in which the evolution of the S&P 500 is accounted for by allowing for stochastic risk aversion. and Campbell and Cochrane (1999a). Both models allow the intertemporal marginal rate of substitution to vary stochastically but in different ways. Other models focus upon the nature of the dividend process e.g. Donaldson and Kamstra (1996) and Timmermann (1996). One of the features that these models are designed to capture is ``bull and bear markets'' and this paper enquires into whether the implied DGP’s are capable of producing historical bull and bear (B&B) markets. Since B&B markets are the analogues of expansions and contractions in the business cycle literature we use methods from the latter to date the turning points that demarcate the markets. This exercise is performed on monthly data for the equivalent of the S&P500 over the years 1835/1-1997/5. We formally show that the nature of B&B markets will depend upon the type of DGP which generates capital gains in the market. This leads us to consider the ability of various statistical models to produce series with the right characteristics---random walk, EGARCH, duration dependent etc. We then focus in some detail upon economic models such those of Gordon and St-Armour, Campbell and Cochrane (1999a) and Donaldson and Kamstra (1996). The latter is designed to explain the B&B markets of the 1920s and we study the likelihood of such extreme events by simulation methods.

PDF file of paper: pagan.pdf

Session: Econometric Methods in Finance

Time: Saturday, 7 July, 8am - 9:30am

Room: A