Tuesday, May 07, 2013

The True Cost of (Equity) Capital

I know it's nice to discount any complaint or warning issues by bankers that some new regulation will negatively impact the economy, but I think when it comes to concerns around increasing (equity) capital requirements, they are correct. We should increase capital requirements anyway.

Kwak discounts the bankers' concerns with a number of assertions which are, factually, not true:
Some of the arguments against higher capital requirements are simply incoherent, like the idea that banks would be forced to set aside capital instead of lending it...
Some contradict basic principles of corporate finance, like the idea that adding more equity capital increases banks’ cost of funding.* Yes, equity is usually more expensive than debt (meaning that investors demand a higher expected rate of return) because it is riskier (the range of possible outcomes is greater). But as you add equity, both the debt and the equity become less risky (since the firm is less leveraged), which reduces the cost of debt and the cost of equity...
In the real world, debt has a tax advantage (interest on debt is tax-deductible, while cash that is paid to shareholders or reinvested in operations is not), so increasing debt can reduce the overall cost of financing. But that’s a government subsidy...
I'll be gentle and not cite the part where he brings up Modigliani-Miller. Kwak is likely familiar with some corp fin, but is not well versed in bank operations or regulation. A banks ability to lend is primarily constrained by capital requirements (not reserve requirements as many thing) which means there is a limit to how many loans it can extend, and the riskiness of those loans, based on how much equity it has. All else equal, the more a bank lends, and the higher an interest rate in can charge on those loans, the more money it makes.

Therefore, higher capital requirements limit how much a bank can lend, which impacts its ability to make money, and also causes economic harm to the extent that less bank lending is harmful to the non-bank economy.

Some important additional elements: We want bank lending to be prudent, which means putting private capital in first loss position against bad loans. The equity that investors put into banks is exactly the capital which should be in first loss position, so having more of it is not a bad idea as it will be able to absorb more loan writeoffs before the bank needs to go to the Government for a hand out.

Beyond that, the system can still be gamed by mispricing risk on the asset side -- if you can classify a risk asset as a safe one (as the banking system did with mortgages) you can end up undercapitalized on a risk adjust basis even with larger capital requirements. Higher thresholds deal with this in a crude way, but I'm not sure what better tools there are to model risk more accurately and make sure that this sort of covert leveraging is not allowed. Better to accept it's a black box and forbid securitization -- keep loans on the books of the originator and make those who invest in the originator have their necks on the line for performance.


Blogger Ralph Musgrave said...

“A banks ability to lend is primarily constrained by capital requirements..”

Clearly if a bank’s capital is for some strange reason permanently fixed and impossible to alter, then capital constrains lending. But of course capital is not fixed: that is, if there are viable lending opportunities out there, a bank will be able to attract the capital to carry the risk involved in making those loans.

Next, what exactly is wrong with Modigliani-Miller? They made the perfectly valid point that adjusting capital requirements has no effect on the cost of funding a bank of a given size and involved in making loans of a given amount and type. That is, the reason shareholders demand a higher return than depositors is that shareholders carry a risk. But if that risk is shared between a large number of shareholders or shareholdings, there is no effect on the total risk being carried. Ergo the cost of carrying the risk remains unaltered. Ergo capital ratios are near irrelevant. So why don’t we just up the ratio to 10 or 20%? That would vastly improve the safety of banks, while having no effect on the cost of funding banks.

And if you want proof of the latter point, consider the fact that British mutual building societies in effect have a shareholder to asset ratio of 100%, yet they compete very effectively with private banks. Those building societies have no shareholders: only depsoitors. That means that those depsoitors are in effect shareholders. As Mervyn King pointed out in reference to those depositors, “in a mutual organisation they are, in effect, the shareholders.” See:


1:49 AM  
Blogger winterspeak said...

Hi Ralph

You are right, a bank can raise more capital and lend more. But it needs to raise more capital first and make the loan second (to stay in regulatory compliance) and raising capital is non-trivial and expensive. Eventually, the marginal cost of that additional capital equals the marginal profit expected from the next loan. This is why capital constrains lending. Reserves do not operate in this way at all.

MM simply states that, under certain conditions, how a firm is financed makes no difference to the value of the firm. It is not relevant to the mechanic where a bank levers up its balance sheet to make loans.

I would love to understand how British mutual societies actually work.

9:56 AM  
Blogger NeilW said...

And there's the small matter of deposit insurance which lowers the cost of deposits as a funding stream. You don't get that with capital bonds.

To be fair banks make loans first and then raise capital. Or more accurately they start looking for capital early in the loan sales cycle based upon conversion projections. If there is a lot of demand coming in it kicks off the capital process so that the capital is in place by the time the loans actually materialise. So demand for loans still drives the capital process and the limiting factor is the price of loans and funding as it always is.

Given the bailout of the banks you could argue that the central bank is forced to provide capital as well as reserves and therefore it doesn't really operate as much of a constraint. :)

Permanent Building Societies are very simple. They borrow short and lend long. It shouldn't work but it does - mostly.

Traditionally they have lots of little savers and borrowers - all of whom are risk members but the savers are deposit protected. The profit accumulates to the society usually and interest rates paid out or charged may be altered depending upon how big a buffer the society needs to deal with bad loans.

Traditional building societies are 'shadow banking' organisations that clear via a normal commercial bank account with a clearing bank.

Again traditionally building societies had a set amount of money that they could lend per month depending upon the savings. Building Societies like Halifax (which is where I live) pioneered more advanced funding techniques with computer technology that got around that restriction with a more dynamic funding model.

The road to ruin was then set.

11:50 AM  

Post a Comment

Subscribe to Post Comments [Atom]

<< Home