For banks, how big is too big to fail?

One issue at the heart of the crisis response is resolving the moral hazard of institutions which are ‘too big to fail’: where the estimated cost to the economy and broader society of a collapse is greater than the cost of a taxpayer-funded bail-out, governments have tended to act to save such institutions.

"Leaving aside the all-too-predictable US moves, supported neither by clearly articulated policy nor evidence, the issue of too big to fail remains complex.”

This was particularly the case in the United States and United Kingdom where taxpayers were on the hook for billions. In Australia, major banks explicitly paid for an insurance policy so they could maintain their funding base.

Leaving aside the all-too-predictable US moves, supported neither by clearly articulated policy nor evidence, the issue of too big to fail remains complex.

The broad argument – that no institution should be so large it poses a systemic threat – is logical but actually in some financial markets there is actually an economic argument for dominant institutions.

As Wharton finance professor Chaojun Wang explains there are reasons some markets are particularly concentrated and it is not all about market power.

“Sometimes it is efficient,” he writes. “Sometimes it is inefficient in the sense that there may be too much concentration or not enough concentration.”

Wang uses the example of trade intermediation and the huge role played by institutions like JP Morgan and Citi.

“On the one hand, there is some efficiency gained if everybody trades with JPMorgan because JPMorgan will be very efficient in balancing its inventory very quickly, handling all of the trading orders from all of the customers,” he says.

“On the other hand, you do want to have more than one dealer because there is a desire from the customers, like insurance companies and mutual funds, to have more competition among the dealers.”

“These two forces, the inventory efficiency and the inventory balancing, versus the competition among the dealers, will determine the structure of financial systems in such a way that most of the trades will be concentrated among very few large dealers.”

Specialised markets

Wang’s research focussed on particular, often specialised markets, and not retail banking. However, these forces are at play more broadly.

An underlying tension in the strategic response by many institutions in recent times has been the balance between economies of scale and diseconomies – essentially the efficiency of the big firm versus the cost of complexity.

This has driven an industry-wide focus on simplification – at the product level and the business model level. Notably, in Australia, there has been a move by banks to sell off businesses like wealth management and insurance.

The rationale here is clear: while universal banking models, which offered customers everything from basic banking to insurance, wealth management, stock broking and financial advice, made sense in theory, they were very complex to operate.

The regulatory response to complex banking models has also been more intense. Globally, there has been both this trend to simplification for efficiency sake but also in response to the too big to fail measures.

The over-arching strategy of global regulators, under the auspices of the Bank for International Settlements and its agency the Financial Stability Board, has been to make financial institutions more robust by increasing capital requirements and less reliant on the taxpayer in cases where they still get into trouble.

This applies particularly to what the BIS calls Globally Systemically Important Banks or G-SIBS – those banks which globally are too big to fail. A second tier, D-SIBS, applies to domestically important banks.

Importantly, this regulatory agenda is also driving simplification – from a regulatory perspective, there is a much more expensive impost on large, complex institutions. So there’s a carrot and stick to simplification.

In theory, this all makes a lot of sense. Taxpayers shouldn’t be on the hook for the failure of banks to self-insure. But there is, as always, unappreciated consequences.

The research of Wharton’s Wang finds it is economically rational for some players to be dominant players.

“The original intention of all of these financial regulations was to increase financial stability by requiring the banks to have a larger capital buffer in case of a crisis,” he says.

But in some cases, notably liquid-but-small asset classes, inventory management is actually more desirable than large numbers of market participants.

For large, liquid markets you do want many traders.

 “What my theory predicts is that you want to have lower capital requirements because you want to encourage dealer participation for the highly liquid assets such as treasuries,” Wang writes. “You don’t want to have a very large, monopolist bank that dominates a huge market like the treasuries market.”

So Wang believes there needs to be more tailoring of capital requirements to reflect efficient markets.

The other implication of simplification is the question of convenience. In an ideal world, consumers want less complexity and that can mean dealing with fewer institutions – the driver of the universal banking model.

Somebody like you

Now that banks have found the model too big to manage it doesn’t mean it will disappear. What it may mean is a giant non-bank will take on the role of ‘bank holding company’. Someone like Amazon.

As CB Insights explained in its latest report on Amazon’s financial services strategy, from payments to “lending to insurance to checking accounts, Amazon is attacking financial services from every angle without applying to be a conventional bank.”

{CF_IMAGE}

In essence, Amazon is starting with a blank sheet of paper and building its model on pretty well what every senior bank executive is saying: it all comes down to the customer experience. Customers do care about price and security but they particularly care about service and a convenient experience.

As Bain & Co says Amazon “can afford to go after this previously unprofitable segment (basic transaction accounts) in part because it will be able to transform the economics of banking; Amazon does not have the burden of an expensive branch and contact centre network, which we estimate comprises roughly 40 per cent of a North American retail bank’s costs on average”.

“Instead, Amazon could steer new customers to ‘just ask Alexa’, its voice assistant on the Echo device,” Bain speculates.

“The company can also avoid a lot of the customer acquisition costs borne by most direct banks because it already has digital relationships with so many Americans. Given these two advantages, Amazon’s incremental costs will be almost nil.”

“Amazon will not legally become a bank. Rather, the bank it partners with would probably hold deposits, while Amazon would design and manage the customer experience and distribution.”

“The arrangement allows Amazon to avoid dealing with bank regulatory compliance and managing the balance sheet.”

That is, Amazon would become a vertically integrated, universal bank – exactly what many incumbent banks are moving away from. It would provide a full array of financial services but, critically, would not directly own all of them – particularly those that are heavily regulated.

Amazon is not the first non-bank to try this. Department stores and supermarkets in the UK had a shot, similarly based on perceived strong customer relationships. Microsoft tried it in the 90s.

Essentially those attempts came back to the same central challenge: how to manage the complexity of a large, diverse business model with different customer demands, different regulators and different cultures.

Andrew Cornell is Managing Editor at bluenotes