The 1998 Canadian Bank Merger Decision and the 2008 Financial Crisis: Factual and Counterfactual Investigations

The observed stability of large Canadian banks during the 2008 financial crisis, in contrast with several American and European counterparts, poses a puzzle. This puzzle is explored through three essays, each pursuing a question with a different method. Yet all three are linked to the historic interlude marked by the 1998 government rejection of proposed mergers of four large Canadian banks and by the 2008 crisis when the banks remained stable. Essay One is a historical essay that asks if the behavior of the Canadian banks themselves, above and beyond that required by regulation, played an important role in avoiding instability. A political economy framework is used to trace out the evolution of the Canadian banking order and describe the structural context in which Canadian banks behaved leading up to and during the crisis. An expectation-based relationship between the banks and the prudential regulator is found at the centre of this banking order, making it difficult to disentangle autonomous bank behavior from the regulatory relationship. This renders the essay's question indeterminate. Essay Two uses econometrics to analyze the financial characteristics of large banks and evaluate whether mergers of equals (the sort proposed by the Canadian banks in 1998) contribute to riskier behavior by the consolidated banks. While the evidence is ambiguous in its support of this hypothesis, it does show that both a bank's funding behavior and risk appetite are significant predictors of bank performance during the crisis. Essay Three tests a counterfactual hypothesis: if these four Canadian banks had merged in 1998, the resulting merged banks would have faced a greater risk of failure in 2008. The counterfactual framework applies synthetic data of two fictive merged banks in a stress test model simulating the 2008 crisis. The simulations describe conditions where the imagined merged banks would face greater insolvency risk. But when considering the conclusions of Essay One, these conditions would have been precluded by the regulatory relationship, suggesting the merged banks would not have faced a greater risk of failure in 2008.