On Bank Runs and Catastrophe

Financial collapse can happen suddenly and catastrophically.

We do not think about the possibility of bank runs very often anymore. This is not because they cannot happen, but because the modern financial system has created many ways in which to paper over (literally and figuratively) crises. But bank runs remain a very real possibility of the fractional reserve system when something compels bank clients to simultaneously withdraw large sums of their cash. A crisis at one “bad bank” can sow fear and panic, leading to multiple runs on banks, culminating in a full blown banking crisis. This issue has appeared at the fore in recent days on news that Canadian banks may act in an unprecedented manner by order of emergency.

Documented historical incidence of bank runs go back hundreds of years. Early events could arise as a result of a bad harvest year resulting in debt default. Bank runs were a consistent feature of the Great Depression. The 2007-08 financial crisis has been characterized as one long bank run which was later used to justify policy interventions such as nationalizations, deposit backstops and increased capital requirements (all of which arguably have made the financial system less stable, but that’s for another day). 

Since 2019, it has been apparent that banks are again under enormous pressure. The overnight lending markets were signalling impending trouble by the fall, as John Chappell writes. But given the recent rhetoric around federal government seizure of private assets, what would it take for a banking crisis to actually occur?

Catastrophe theory emerged some 60 years ago to model dynamic systems where humanistic components played a significant role in outcomes. Though early models centred on physics and biology, to explain natural phenomena such as tidal waves, and fear and aggression responses (respectively), the implications for finance were soon obvious. Business cycles have been observed to follow asymmetries observed in nature, and irreversibility is a common feature of financial time series. Since the advent of nonlinear dynamics in the 1960s, along with the technological ability to model these dynamics, there is ample evidence of volatile systems in the natural world.

The path to insolvency is often assumed to be smooth and continuous, but the reality is that complex systems often fall apart very suddenly and very completely – recall the old Wall Street saying: everything takes longer than you expect, and then happens faster than you expect.

Catastrophe theory has some underlying requirements in order for these events to occur. Namely, that small changes in variables and behaviour can lead to big changes in outcome, and that discontinuous or surprise events may occur. Both of these elements are present in the modern financial system. Were a government (or other entity) able to sow enough panic, and hesitate to authorize some counterbalance to quell that panic, it is conceivable that our banks would run out of cash; this would sow further panic and create a volatile system that would take political force (and will) to stop. Canadians have little experience in this regard.

It is probably worth pointing out that none of this is necessarily bad for society. Instability is a good thing, for without it, nothing would ever change. To the extent that instability causes great discomfort in the short run, governments would be wise to avoid pushing systems out of equilibrium for little (or, cynically, political) gain. The only silver lining to our current predicament may be that periods of instability tend to produce large jumps in human living standards, and that political popularity itself is a volatile system subject to the same winds of change. The 20th century alone provides a very rich history lesson – on both banks and politics. The Canadian government would be wise to take note.