Do We Need Big Banks?

Humans can’t handle more than 150 relationships. It’s a throwback to our caveman past – when tribes exceeded 150 people, bonds of trust started to weaken and no one really knew what was up.

Big banks in New Zealand have thousands of employees and hundreds of thousands of customers. Big banks in the US have tens of thousands of employees and millions of customers. No one really knows what is going on because there are so many people involved in the banking sector.

Back in the olden days, before I was born, you used to have a relationship with a bank manager. When you went to apply for a loan, he knew your banking history, probably your parents and your grandparents too. If you were applying for a farm loan he’d have a decent knowledge of your farming ability.

This meant that the distance between savers and borrowers was small. Relationship banking meant knowing your customers, knowing the local community and knowing that if someone had good character they were likely to pay the bank back eventually.

Back in the olden days, there was also a lower incidence of bad behaviour. Defrauding banks still happened, but it was a lot harder than it is today. Because communities were smaller, the social shame effect could be used to lower the likelihood of default.

This is all in stark contrast to how the banking sector functions today. Credit approval is granted based on points and complicated models that evaluate a potential borrowers prospects.

Putting aside the role that fancy models played in the collapse of Long Term Capital Management, the NASDAQ technology collapse, the housing boom and rise of sub-prime, the derivatives on sovereign debt and even domestic housing developments, the big banks today have enormous distance between the savers and the borrowers.

In any system, when you centralise decision making, you lose knowledge at the coal face that matters. In his essay The Use of Knowledge In Society, F A Hayek argued that the reason central planning failed was because a committee couldn’t make price decisions for every single actor in an economy.

Central planning in the banking system is represented by centralised credit approval systems. They use fancy models that use crap models like “Value at Risk” because the regulators, who know very little about the real world, have mandated that those models are the true representation of the risk a bank is exposed to.

Dispersed knowledge applies just as much as the private sector as it does in the public sector. Dozens of local bank managers making subjective decisions is likely to be more stable than a centralised credit approval system.

The conceit, of course, is that a model is a simplification of reality. It is made up of assumptions, can disregard potentially relevant factors as “not useful” and applied to situations it really shouldn’t be.

Big banks in New Zealand today are walking blind through a minefield. Their loan default models are based on past default rates. Their “worst case scenario” stress testing models forget that the “worst case scenario” is always worse than the last worst case scenario before that!

We don’t really need big banks because the arguments they use to justify their existence, like they provide payment networks and the like, are redundant in the 21st century. They are a utility, the flash offices I see in Wellington belonging to big banks offends me deeply.

They should be as boring as water companies. Not some sexy, profitable, lavish executive pay with no clawbacks for imposing systemic risks on everyone else nirvana. The big banks have to go if we want any return to a realistic distance between savers and borrowers.