It's widely accepted by now that financial and economic models did little to prevent the Great Recession, or at least see it coming. But there was very little said in terms of what the Bank of Canada uses. Noticing this lack of information, the C.D. Howe Institute took an in depth look and found that the Bank still doesn't have to resources necessary to react appropriately during a crisis.
The Bank's models have "yet to be expanded in a number of directions if they are to capture relevant characteristics of finance in the real world," authors Philippe Bergevin, Pierre Duguay and Paul Jenkins reported. These men know what they're talking about. One is a former deputy governor at the Bank of Canada and another was a senior deputy governor at the Bank.
Chiefly, the Bank's most commonly used model, known as ToTEM, is flawed because it assumes that financial interdependencies do not matter for macroeconomic stability. In simpler words, it assumes that the banks' reliance on each other does not have a connection to macroeconomic fundamentals.
Although most financial sector activity was baked into the model before the financial crisis, the "glaring exception is the role of liquidity," the report stated. The last crisis proved just how badly that hurt the economy. With no liquidity between the banks, the world went into shock because suddenly everyone stopped borrowing and buying.
The Bank's model is also a forward-looking one, rather than one that uses historical information to find correlations -- which in most cases is a good thing. But to look into the future, it has to assume that people are rational consumers and borrowers. In a crisis, they are anything but rational.
Still, there are two important questions. First, does this matter for the Bank of Canada? There main role is price stability for the Canadian economy. Should it also be financial stability?
If only it were so simple. The bank certainly has a mandate, but in times of crises, it must be able to assess the situation so that it can determine what to do with interest rates, etc.
Second, how much can we really rely on models? Anyone who has taken introductory statistics knows that models always break down in certain conditions. Maybe we need to better understand the limitations.
In fact, the Bank's models have been a work in progress since the first one was designed 45 years ago. (Which raises a separate question: what did they do before then?) Because the first model was quite simple, a second more complex iteration was developed. Yet that one also broke down because it couldn't deal with the outbreak of inflation in the 1970s.
Learning from experience, there was then a string of new successive models, all of which were only worthwhile in certain circumstances. And now ToTEM is being re-worked, to account for things like inter-bank lending and the importance of bank capital. (It should be noted that this isn't the only model used. Others like BoC-GEM exist. But ToTEM is the most useful top-down macro model.)
Until the Bank gets it right, C.D. Howe reminds the public: "Models are and will always be, by nature and by necessity, simplified representations of the economy, and that no one model can answer all questions"
Tim Kiladze