Unstable Banking
What to do about the banks?
Those of you who have read my comments on health care know that I believe any proper regulation of a system -heath care or banking – requires a set of incentives designed to induce the desired behavior. This is the way in which we can align behavior with socially acceptable outcomes. I have discussed that with respect to health care ad nauseam, but have so far not talked about it in terms of banking.
So here goes.
Let’s start with the basics:
- Banks areinherently risky. There is no way to have a bank system that is not riven through with risk. The whole purpose of a bank is to ‘mediate’ or translate short term deposits and liabilities into longer term assets like loans. The risks there are obvious: there is a risk of not being repaid, in which case the bank still owes its depositors but has no assets to turn into cash so it can repay them; and there is time dependent risk, this is when interest rates on deposits, because they tend to change frequently – they are short term don’t forget – end up being higher than the interest being earned on the assets, in which case the bank loses money on its assets even tough they end up being repaid. These are the two major risks banks undertake in the normal course of business.
- Banks mitigate these risks by trying to understand the chances of default: they check credit scores, they model likely outcomes, and they go over the books of the businesses they lend to.
- They also try to match the cash flows of their assets and liabilities on their balance sheets to reduce interest rate risk.
- And they keep a pot of capital they can dip into in case of emergency so that they can write off bad assets rather than take a loss on their annual income statements.
- In order to have this emergency pot of capital banks regularly set aside, from their annual income, a ‘reserve’ for potential loan losses. This reserve, plus the accumulated profits and shareholder capital, are the buffer that prevents a bank from losing money even when its assets go bad.
So that’s the essence of banking.
The problems get deeper when we realize that the ‘spread’ between the assets and liabilities, that is the difference in interest rates a bank pays its depositors and what it charges its borrowers, is normally very thin. Banking can be very expensive: all those branches are costly to run, the infrastructure to account accurately for all that money is very expensive because it has to be close to flawless, banking is people intensive, credit analysis and data gathering cost money, and so on.
To pay for all this overhead banks don’t simply lend out dollar for dollar what capital they have – that would never cover the bills. Instead they borrow even more and lend that money out too. In other words their liabilities are not simply capital owed to shareholders, but also a variety of creditors: depositors like you and me, other banks, the Federal Reserve Board, and bondholders etc.. This piling on of other liabilities in order to expand the pool of cash that can be lent out for profit is called ‘leverage’ in the US and ‘gearing’ in the UK. You ‘lever up’ the profit opportunity.
This is where things get dodgy. And this is where incentives come into play.
Before I go into the details let me recommend the academic paper that provoked my thinking on this. It is byLucian Bebchuk and Holger Spamann of the Harvard Law School. For those of you involved in bank regulation it is worth the time. Plus it’s free.
Back to incentives:
Shareholders sit at the bottom of the liability heap. If the bank simply uses its capital pool to generate assets it will make very little money. So the shareholders have every incentive to pile on as much leverage as possible. The more leverage the more profit and therefore the higher the ‘return to equity’ [‘ROE’] generated. Shareholders love high ROE’s because that’s what flows through to their pockets. And an easy way to pump up ROE is to have as little ‘E’ as possible per dollar of assets.
Let’s illustrate that: an income of $10 on $100o of assets provides a 10% ROE if there is equity of $100 supporting those assets, but 20% if there is only $50 of equity. In the first case there is a ten to one relationship between equity and assets in the second case twenty to one. Further: in the case of 20:1 leverage a bank with $100 of equity can have $2000 of assets whereas at 10:1 it can have only $1000.
So shareholders love leverage: it can make them rich.
They love it so much that before this last crisis many of our leading investment banks had leverage ratios in excess of 30:1. Their ROE’s were fabulous.
Bankers love leverage also.
They get paid for all the cash that the leverage produces. The income potential obviously goes through the roof if you are highly leveraged. that means the bonus polls go up too. As I said bankers love leverage just as much as shareholders.
Neither shareholders nor bankers have much downside in this leverage game. The shareholders have at risk only a small portion of the total liability base of the bank: their downside is limited at their equity stake. The creditors have far more at stake. In a bank with leverage of 30:1 for each dollar the shareholders stand to lose, the creditors stand to lose twenty-nine. The bankers themselves stand to lose their bonuses and, you would have thought, their jobs. While that may sound drastic it pales in comparison to the plight of the creditors.
So the basic incentive structure of banking is not just to take risks, but to take enormous risks. The payoffs are gargantuan while the losses are limited by comparison. Bankers and shareholders are playing with other people’s money. Inevitably this will lead some bankers to cross the line from judicious risk taking into outright gambling.
That’s fine when only a few rogues do it. It is disastrous when the entire system does it.
Unfortunately our entire system went on a gambling binge and lost.
Some of them lost their jobs, although not many – which I find outrageous, and the shareholders of the likes of Citibank and Bank of America lost their stock value. You and I lost billions.
Why?
Because the incentives are all screwed up.
The risks I mentioned at the beginning – of default and interest rate mismatching – pale in significance when compared with the risk introduced by excessive leverage. The problem is that while banks may feel obligated to manage the first two risks because they can’t earn profits unless they do, they have much less incentive to manage leverage risk. On the contrary they are impelled by the profit motive to let it rip. They are behaving entirely rationally when they have 30:1 ratios.
This leads us to a vital conclusion: financial markets are unstable. AsHyman Minsky taught us, or rather taught those who paid attention to him, financial markets are prone to regular disasters. It is an inherent property of finance.
Why is this?
Minsky’s hypothesis is that as bankers observe stable economic conditions, say during a boom, they will naturally feel able to take on more leverage: they are like gamblers on a roll at the casino, they keep upping the ante. The problem is that bankers, when they ‘up the ante’ are creating liquidity: they add to the economy’s store of money by making loans. Remember they are using the same level of capital to make more loans [or other assets like mortgage backed securities etc], so the economy itself becomes less secure: there is less capital backing a larger volume of business. Eventually, according to Minsky, this gambling will implode. Something will cause the edifice to topple over and all that leverage causes cascades of losses through the economy.
This is why Minsky talked of stability leading to instability. He was right.
Wait it gets worse:
Because banks lend to each other one bank’s problems with defaults becomes another banks problems. If bank ‘A’ – let’s call it Bear Stearns – has a problem and has borrowed tons of money form bank ‘B’ – let’s call that one Bank of America – then Bear Stearns inability to pay back its creditors now becomes Bank of America’s default risk. And since Bank of America borrowed from bank ‘C’ – let’s call that one The Royal Bank of Scotland – the ripple effects trickle through the entire system. Actually they don’t trickle they get amplified: banks behave in herds, they tend to do the same things at the same time. This is entirely rational because they are all reacting to the same information. An impending Bear Stearns implosion causes everyone else in the game to retrench. This simultaneous retrenchment causes them all to call on Bear Stearns for repayment, which creates a ‘run on the bank’ for Bear Stearns who then goes bust since it has nowhere to get funding.
Why can’t Bear Stearns go to the Federal Resere Board for funding?
Because it is not regulated as a bank. Deregulation, absence of regulation, and poorly enforced regulation have left vast financial organizations with trillions of creditors and assets sitting outside the banking systems: Goldman Sachs, Lehmann Brothers, Bear Stearns, GMAC, Countrywide, Morgan Stanley. A veritable rogues gallery of massive risk takers and gamblers, none within the system. None officially able to get backing from the Fed in times of crisis, and so none immune to a ‘run on the bank’ situation.
No problem you might say: let them go bust. The taxpayers are not on the hook because they are outside the New Deal era safety net.
Umm. Not exactly.
Those creditors. Some are within the safety net. If Lehmann is allowed to go belly up, heck the free market should be set free, then who cares?
Well apparently everyone. The entire system crashed. Welcome to another risk: systemic risk.
The entire world came close to an end.
Most of our big banks within the system lost their shirts. We had to step up and bail them out. Not just this, but we ended up bailing out organizations not in the system. Goldman Sachs got a free ride. We bailed it our even though it never had to conform to any rules. Lucky dogs. Welcome to yet another risk: the public guarantee risk which induces even more instability since we provide a nice net for the gamblers to fall into so they don’t hurt us too much when their gambling fails.
Worse yet: Goldman Sachs is on the brink of its most profitable year ever. Courtesy of taxpayer money that stabilized the financial system, and benefitting from the continued guarantee that no ‘too-big-to fail’ bank will be allowed to go bust – memories of Lehmann haunt us – it continues to play with other people’s money and will pay record bonuses this year. And many of its ‘top’ employees will moan and complain over their fancy drinks about the nerve of the administration bailing out GM.
The lesson to learn from all this is that incentives matter. They override all else.
If Minsky is correct, and I believe recent history proves he is, then stability does indeed lead to instability.
And since you and I end up underwriting the cost of the instability part – think ‘bail outs’, but don’t profit from the stability part – think bonuses, we have an incentive to limit the damage the bankers can do to us.
If, as I have tried to explain, the natural incentive structure of finance is to encourage wild eyed gambling on the part of bank shareholders and bankers, and if we feel the need to continue to maintain our guarantee of the gambling – which I feel we should to limit the damage these yahoos create – then we need to have rules that they should play by. That way we can take the toys away when we need to. We intervene and tilt the incentive structure towards socially acceptable behavior. You know: of the kind where these cowboys don’t scupper the entire economy in their quest for the next quarter’s big bonus payment.
So when we debate the upcoming reform of banking, we should all pay attention to the way in which its provisions compensate for the skewed and anti-social behavior that the incentive structure of the market produces.
That means addressing banker pay and bonuses, and the size of banks.
Unfortunately, even though the proposals acknowledge the role of both in creating the crisis, it leaves very vague what we should do about them.
That’s why I am not a fan of the administration’s efforts so far. They talk a good game, but are very tentative in action.
Meanwhile those incentives look set to cause us trouble again sometime soon.
Addendum:
A special nod to John Fleming for correcting the math in my little example. Perhaps silly math is why I liked banking?