Bank Capital: More Please

Our banks remain chronically undercapitalized. That is to say they run on too narrow a cushion of equity capital that can be extinguished by losses before they have to be bailed out or allowed to fail.There seem to be two causes for this.

First: it benefits management to have as little capital as possible. The lower the capital, the higher the return on that capital – as long as earnings stay constant. This is the classic leverage incentive that overwhelmed the banks in the crisis. Of course wafer thin capital at the bottom of the balance sheet means that there is precious little room for error. It doesn’t take much of a loss to plunge a poorly capitalized bank into crisis. As we learned during the crisis our banks are not very good at managing risk. In fact, were one to be dispassionate about it, their recent record is absolutely terrible. Were the market for risk managers truly responsive to actual performance rather than hyped up reputations we would be awash with unemployed risk managers and traders. They would be changing careers towards something they are better at. Like demolition derby driving. Unfortunately hype dominates and theses precious people still lurk about our banks ready to trash them at a moment’s notice.

But why does this benefit management? Because they are often, most often, evaluated on their ability to produce a return on equity. Shareholders want that ROE to be as high as possible, so they compensate managers who reach higher targets and penalize those who lag behind. I remember this pressure myself. It is always very easy for bank managers to convince themselves that they have risk under control when a major consequence of that decision is a fatter bonus.

Add in the very short term, and thus absentee, nature of shareholders in a modern economy – the old fashioned notion of shareholders actually “owning” a company is quaint to say the least and you compound this obsession with ROE. In these days of stock churning and institutional investment the idea that shareholders control management is a farce. Short term investors think they can bail out of a bank stock before the risks of producing a high ROE come home to roost. So a key risk control mechanism – that the shareholders don’t want to lose their investment – has been undermined.

So the combination of wafer thin capital and rotten risk managers is toxic for the financial system. That was where we were back in 2008. And it is pretty much where we are still.

The second reason that banks have too little equity is that it costs more than debt. This is because the cost of debt can be deducted for tax purposes whereas the cost of equity – the dividends paid on that equity – cannot. So even if the pre-tax cost of debt is higher, the after-tax cost isn’t. That makes debt very attractive as a funding source. Thus a bank’s total capital tends to include big chunks of long and medium term debt in all their varieties. The owners of this debt – the bondholders – have less at risk since they are contractually owed whatever they lent, plus a stream of cash payments to pay the interest on the debt. If the equity is wiped out by excessive losses the next layer of capital to absorb loss is that provided by bondholders. That’s why we call it capital and not a deposit. But one of the strange facts of our recent crisis was the extraordinary lengths everyone went to to prevent bondholders absorbing any loss at all.

That’s what all the talk of “haircuts” is about. A haircut in this parlance is the amount of a trimming given to the value of the bonds owned by bondholders. If they lent $100, and get back only $90. The haircut was $10.

Now here’s an interesting conundrum: wouldn’t you, as a bondholder, want the bank you lent to to have enormous equity? That would reduce the risk that you have to take a haircut on your bonds.

More to the point: wouldn’t you as an equity holder also want the bank to have plenty of equity capital to absorb losses. The bigger the cushion the less likelihood that the bank will fold and take your investment with it.

It would seem as if it is in everyone’s interests to have banks with as much capital as possible. The more the merrier. That would make the banks safer, sounder, and more reliable. It would lessen the risk in the real economy of damage from bank runs and collapses etc.

Apparently not.

A result of our having banks withy too little equity is that the bondholders have found themselves too exposed to losses. More so than they thought. Hence the great lengths and sacrifices that governments are going to the prevent bondholders taking haircuts. Indeed it is the zeal with which some countries and regulators have gone to in order to screen bondholders from losses that has caused the rush to austerity in the budgets of governments who have bailed out banks. In effect the taxpayers are protecting the bondholders. In effect we are protecting bondholders from having the responsibility to figure out how much risk they take when they invest. The implication being that being a bondholder in a bank, which is by its nature a risky thing to do, is now seen as being something that ought to be risk free. Perhaps a better way to look at it is that taxpayers have now been revealed as contingent equity holders who can be called upon to provide a layer of protection to bondholders. Whether taxpayers will provide this service free of charge for much longer is questionable.

Bondholders were clueless about the real riskiness of their investments – their ability to calculate risk is obviously very poor in view of the outcome and their cries for help – and they had evidently bought into the same delusions that managers had. They didn’t need all that equity sitting around, the bank risk managers were so good.

Whoops.

Even with this awful track record, there has been huge resistance to raising bank capital targets.

Bank executives drone on about how it would raise the cost of capital and thus make capital within the economy more expensive. This would, they argue, reduce investment and slow growth. Possibly. But the wild gyrations of an unstable financial system built upon the twin illusions of rotten risk management techniques and wafer thin capital ratios more than offsets the higher cost of capital from a safer and less risky system. What they are really worried about is their loss of ROE and their bonuses. The cost of capital in a more capitalized system is likely to trend down as investors require less coverage for their investment risk.

Meanwhile, the European banks, full as they are with bad sovereign debts, probably couldn’t raise new capital at the moment either. So their governments appear hell bent on scuppering or reducing the new and higher capital requirements. This has produced another line of argument for the American banks: they suggest that higher capital for them would put them at a competitive disadvantage, and would thus hinder the US recovery.

This is nonsense for the simple reason that we should want our banks to be well capitalized so that they no longer threaten our economy. Why should we emulate European death wishes? Our banks, by being better capitalized, would attract capital and business. They would be seen as more stable, have better credit ratings, and all the other accoutrements of soundness that we used to see as essential to a banking system.

A potential problem is during the transition phase. As banks are forced to raise capital they will face higher costs because the supply will be less than the demand. So they should be allowed to phase in towards the new requirements. But those new requirements should be very high. Frankly, the new standards being discussed are ludicrously low. The idea that a giant bank should be able to get away with only 7% of high quality capital is absurd given what we have just gone through, and the social costs that those same banks imposed on taxpayers around the globe.

I have always advocated that banking should be dull and not very profitable. Banks are intermediaries. They are the ultimate middlemen. The service they provide is essential, but is akin to a utility. Their skill should be in capital allocation and the judgement of risk. The compensation for these skills and the utility they provide should be low. After all bank profit is a transaction cost borne by the economy – it is the cost we absorb for having our capital moved around. So the more our banks profit, the more the burden to the rest of our business and economic activity. My view, is of course, hopelessly out of date. Banks nowadays simply see themselves as conduits through which capital flows. They prefer to charge fees rather than hold interest earning risk assets. The former boosts ROE without consuming much equity. The latter requires big chunks of capital to support it, and has all the headaches associated with managing loans. For instance, if you generate and hold onto loans, you need actual loan officers who are adept at managing them, especially when an economic downturn reduces the cash flows of your customers. Heck you might even have to endure a rash of loan renegotiations. They are time consuming and expensive. So banks gave up this old fashioned approach and simply sold off their loans. They reduced the amount of loans they held onto – and had, therefore, to manage – and started dumping them into the secondary market. This released capital for other things – like gambling in proprietary trading – and produced nice fee income.

The whole idea was to be capital efficient.

As you all know I decry the use of efficiency as anything we should aspire to in economics or business. It speaks to a level of certainty simply unattainable in reality. Being capital efficient implies keeping the bare minimum, thus lowering your cost of capital to the lowest possible number. This leaves nothing in reserve for contingencies. After all contingencies are unnecessary in a world where outcomes are well known. The key to efficiency, then, is in knowing the future so that you can eliminate the wastage implied by having to hold redundant cash to cover contingencies. But the highest point of efficiency is also the least sustainable level in the face of uncertainty. This is because the presumption of sufficient foresight or knowledge about future events is ridiculously at odds with reality. We simply do not know enough to be efficient in this purist sense. Yet our banks, and many of our economists, build their strategies and models around just such a vision. For bankers the result was the delusion that they could get away with very low levels of capital. Their risk management models and techniques – riddled through with technical failures inherited from their neo-calssical economic roots – were supposed to protect theme sufficiently that they could rise to new levels of capital efficiency.

As I see the world, suffused with endemic uncertainty, capital needs are vastly higher. More in the range of 15% than 7%. That way the financial system can wallow in capital inefficiency so that it will not infect the real economy. It can absorb the inherent risks of lending without collapsing or calling upon taxpayers.

In the end the debate about bank capital resolves itself to having a view about uncertainty. The less you think we know about the future, the more bank capital you should advocate there is.

Then again, if you call for more equity in banks, you will get called out of date. Like me.

Oh well. Here I go:

More bank capital please!