The use of a four-state probability model can more accurately predict stock market phases and improve investment and risk management decisions, said Tom McCurdy, professor of finance and Bonham chair in international finance at the University of Toronto’s Rotman School of Management, during the Canadian Investment Review’s 2022 Risk Management Conference.
The proposed model incorporates trends and sub-trends labelled bear markets and bear rallies, bull markets and bull corrections, he noted. “I think that individual models or thought processes have a limited domain. . . . What’s really important is, when are we going to have a transition from [a] low interest rate to a high interest rate or from recessions or expansions . . . or bear rallies to bull markets?”
These sub-models can provide a lot of information about the aggregate distribution, including a fuller picture of risks. The four states are also linked by a transition matrix that determines the movement from bear to bull market, he added. “I’m always trying to model the full distribution if I can, rather than just the first moment, and sorting these phases based upon both the risk and the return provides very useful information.”
To be fully effective, this approach requires as much data as possible, said McCurdy, adding while the model looks at how typical market corrections and rallies have played out and makes forecasts based on these past events, it’s still difficult to predict the timing of market crashes.
“We’ll find the sequence of transition from a bull correction back to the bull [market] can be different in different bull markets and the realized structure of the bull markets can be different over time, but we’re still utilizing information about what a typical bull correction looks like to improve our forecasts for the bull market phase.”
McCurdy applied the model to the market heading into the coronavirus pandemic in early 2020, which started in a bull market before quickly sliding into a bear market by the end of February. While some economists were quick to declare a return to a bull market as the market slowly improved in April 2020, McCurdy’s model pointed to a bear rally, amid high levels of volatility.
“The model still thinks there was quite a bit of risk [at that time] and, although we were thinking about it as a positive trend, we weren’t confident enough that we were in this long-run bull market.”
Based on the model, McCurdy noted it’s unlikely the bull probability state will increase to more than 50 per cent until the current market volatility settles down. “Our model would say, ‘Be careful, we’re still in a bear rally.’ Once we get back to normal levels of volatility for the bull state, then it will give us a higher probability [of a bull market].”