Should we be imposing some more taxes on banks?
Fri 5 Apr 2013 by Matt Nolan.

Following the pain of the Global Financial Crisis, banks have come in for a lot of flak.  Additional regulations and taxes have been suggested for banks.  Although a justification for a certain type of bank tax can be found, many of the measures suggested to date have been inappropriate and damaging.  In order to figure out what a reasonable type of tax is, we first need to discuss what is going on in the banking sector.

In the first three Infometrics articles on Interest.co.nz (one, two, and three) we made a point of discussing the Global Financial Crisis – the catalyst, why it persisted, and where it came from.  We concluded that the crisis was set off by central banks forgetting that they were a lender of last resort, but in the final article we also noted that this lender of last resort function for a central bank does lead to problems of moral hazard in the financial sector.  These are the problems we will be discussing today.

Banks link borrowers and lenders in a more efficient and timely manner than the individuals involved can lend to each other.  By pooling risk over a large set of loans, banks can also reduce the risk to lenders, in turn making it attractive to use them.  Furthermore, banks have a greater ability to borrow short and lend long than other types of potential lenders – thereby funding long-term investments that take a long time to realise value, while still allowing depositors reasonable access to their funds.

When banks go bad

However, sometimes things can go wrong.  The fact that banks aren’t just holding people’s money (which is a service they are assumed to provide), but are lending it out for long-term illiquid investments, makes them vulnerable to bank runs – a situation where a bank may be solvent in normal circumstances, but is made insolvent due to a lack of liquidity when a lot of people want their funds at once.  If depositors had knowledge of the true solvency of the bank this wouldn’t be an issue – but given that an insolvent bank may “fake it” and pretend to be solvent, there is an information problem.

Given this, governments around the world have turned around and offered to protect banks.  As a result, deposit guarantees and bailouts are expected by financial institutions.  Contrary to many people’s expectations, shareholders are often wiped out when these sorts of things happen, but the banks management and the depositors in the bank (often) get away scot free.  However, there is no such thing as a free lunch, and so this support has to come from taxpayers.

The reasoning given is that allowing the bank to fail will create broader economic duress, which in turn will cost the taxpayer more!  The failure of Lehman Brother’s illustrated that there was some truth to this at a point in time, and has led to deposit guarantees being viewed as something that society will provide.

We can’t look at banks without looking at time

But we can’t look at these things at a point in time.  Remember, the entire point of lending and borrowing is moving funds, investment, and consumption through time.  The fact that governments are expected to provide a backstop for depositors, and are expected to provide a backstop to the bank’s management, in turn reduces the risk for these people.

If a bank, and the depositors to a bank, expect someone else to pick up the tab when things go wrong they will behave very differently than in the case when their own money is on the line.  This is a situation of moral hazard.  We saw this behaviour in New Zealand with finance companies – people were chasing an extra 0.5%pa with extremely risky non-bank financial institutions because they didn’t believe they could lose their money.  When the government did let them fail, there was a massive public out roar.

Our Reserve Bank recognises that this happened and wants such bankruptcies and risks to be transparent, which is one of the reasons for the Open Bank Resolution (OBR) – a program that determines how financial institutions will be managed for bankruptcy, and includes a loss for depositors.

Now we may not be able to convince depositors that they could lose out – as we’ve pointed out before, they may really believe that the government will always come to save the day.  Given that people make actions on what they expect, this is very relevant, and we will want to make sure we have the right policies in place given these expectations.

In this specific case, we need to make sure that people do “price the risk” they are placing on society through the assumed bailout.  We can do this by setting a deposit levy.

By setting this levy at a rate that accounts for the estimated implicit “subsidy” to the banking sector, the government raises the funds to support depositors during bad times from the depositors during good times.  With deposits now taxed, depositors will be less willing to put their money in a bank unless they are given a rate of return that accounts for this.  As a result, this can be seen as a form of compulsory deposit insurance for people who decide to put their money with a bank – and its existence will increase the cost of bank funding to a level closer to the true “social cost”.

I would note that this isn’t perfect.  If deposits with a series of banks are all viewed as riskless, depositors will chase the one with the highest rate of return – which is likely to be the riskiest.  In this environment, the moral hazard issue still exists – it has just been partially muted by the levy.  This is why the RBNZ understandably prefers the OBR.

Not a Tobin Tax

Now this is not a Tobin Tax – this is a levy on deposits.  A Tobin Tax is a tax on all financial transaction.  It can take a number of forms, but in its most general form it would be like a GST without rebates along the value chain – so it would tax more complicated forms of production more than less complicated ones!  The often suggested versions of a Tobin Tax doesn’t target the underlying issue and what is driving the perceived market failure – which is why such a tax is inappropriate here.

What about the bankers!

But what about our bankers?  Aren’t bankers also taking on too much risk to boost their own bonuses at our expense?  This issue is a bit different.

In the case of the United States, we may very much see this occurring – and this has to do with the generally poor institutional structure of these banks.  However, in many ways this is linked to the fact that “debt is cheap” for banks.

For a bank, a loan is an investment that provides a rate of return.  They fund this through a mixture of debt financing (the funds depositors give them, including funds from unregulated money market vehicles) and equity financing.  When there is a deposit guarantee, debt financing is subsidised for banks relative to equity financing.  As a result, banks tend to raise “too much debt”, keeping equity/capital levels low and increasing their leverage – in turn making them fragile.  Furthermore, the low amount of equity reduces oversight, and encourages an institutional structure that takes on risk.

If the subsidy on debt financing is removed by introducing a deposit levy, then the incentive to use debt instead of equity to fund lending is diminished.  A larger equity holding implies that shareholders have more skin in the game, and so will have an incentive to deal with perverse risk taking and change managerial incentives within banks.

Appropriate monitoring of the financial sector, an acceptance that there are systemic risks that are not taken into account by individual banks, and an acknowledgement of people’s expectations around future government policy are all important elements of any future regulation.  In many ways the Reserve Bank of New Zealand has been, and still is, ahead of the curve on these issues.

However, if we accept that bailouts will occur in certain situations, and that the public will assume deposit insurance will occur (viewing the event as akin to a natural disaster), then a levy on deposits may save the taxpayer money while still giving depositors the insurance they already assume exists.

Related Articles