Ludwig von Mises has this definition:
Banks borrow money in order to lend it; the difference between the rate of interest that is paid to them and the rate that they pay, less their working expenses, constitutes their profit on this kind of transaction. Banking is negotiation between granters of credit and grantees of credit. Only those who lend the money of others are bankers; those who merely lend their own capital are capitalists, but not bankers.
Why bring this up? Well Tim Harford applies the Modigliani – Miller theorem to banks.
M&M applies to banks, too, but with a twist. Banks that get into financial trouble cause systemic damage, so even if M&M applies from the point of view of investors, society would prefer less debt and more equity. But bank investors want the opposite, because the “too big to fail” subsidy means that shareholders enjoy successful gambles while creditors are bailed out if things go wrong. This subsidy means that debt-laden banks are more valuable to investors. If M&M holds, the taxpayers’ loss is the bankers’ gain.
Bankers have tended to argue that equity is scarce and expensive and too much equity means that banks will make fewer loans at higher rates. M&M shows us that this argument is wrong in theory.
In practice, M&M roughly holds: as leverage falls, equity becomes substantially cheaper. Banks are tempted to take on more leverage not because debt is efficient but because debt is the route to an implicit government subsidy.
Banks should be obliged to use more equity funding – or in the misleading jargon of the industry, to “hold more capital”. But equity is not “held”. It’s perfectly good money, provided on flexible terms. It can be lent to businesses and homebuyers just like debt – and with far more resilient results.
I’m a huge fan of the M&M theory – but I’m not convinced that it applies to banking.
The thing is that M&M start their analysis under conditions of perfect markets and them derive their propositions. As exercises in logic they are correct. Then we usually go through the exercise of relaxing various assumptions and observe the changes in the results of the theorem. When you do this you find that transactions costs (costs of bankruptcy for example), agency costs, corporate taxation, asymmetric information etc. matter in the capital structure decision.
So too with banks. Under the M&M base case of perfect markets, banks would not exist at all. That is because banks would earn a zero-profit. The household sector would be able to borrow and lend in debt markets. So banks add value in a world where the M&M perfect market assumptions fail.
Now that doesn’t mean that banks are immune from poor business decisions. It also does suggest that the business of banking is particularly risky – that is why they earn those high profits. Excessive leverage will bring down a bank, just like any other business, if bad debts accumulate. Requiring a bank to hold more equity creates a buffer against poor lending decisions, it doesn’t actually prevent those poor decisions being made.
Bottom line is that pointing to the M&M theorem doesn’t add much value when thinking about banks. That theory predicts banks wouldn’t exist at all.
I recently read Cochrane wanting more equity in banks and he thinks the cost is worth it for the risk reductions. Someone else too, but I can’t remember who.
Pedro
11 Mar 13 at 3:06 pm
Maybe – but I can’t see M&M being the justification.
Sinclair Davidson
11 Mar 13 at 3:12 pm
Under that definition of a bank, do we have any banks at all?
Isn’t a bank today an institution that lends money created out of nothing?
Driftforge
11 Mar 13 at 3:28 pm
It wasn’t, but I thought you might be interested the find the article. It was a review of this book
http://www.amazon.com/The-Bankers-New-Clothes-Banking/dp/0691156840
Pedro
11 Mar 13 at 4:06 pm
The other review was Felix Salmon.
Pedro
11 Mar 13 at 4:07 pm
Pedro – I’ll check it out. That is the same book Harford points to.
Sinclair Davidson
11 Mar 13 at 4:36 pm
I read Harford’s article. I think his point is one you agree with and that your claim banks wouldn’t exist, while potentially true, is not relevant to his point.
Harford uses MM to show that the implicit subsidy increases the preference for debt funding of banking businesses. That is hardly a startling revelation and I suspect he was just trying to make the obvious sound more clever. But I suppose we all sometimes try to hard to find something interesting to say.
Still, there is an interesting question you put, in a world with no capital structure bias for debt, would banks exist? I think the answer is yes.
Pedro
11 Mar 13 at 5:24 pm
Errr….the Austrians have more to say about Fractional Banking than deposits v loans + expenses.
Try explaining FB to you neighbour, he will refuse to believe it.
Alfonso
11 Mar 13 at 5:26 pm
I think I’m going to have the read the underlying book.
Yes – probably. His statement though strikes me as bwing wrong – but I don;t disagree with, say, Cochrane’s review that I subsequently read.
Sinclair Davidson
11 Mar 13 at 5:28 pm
Sinc
As I understand it MM suggest banks need to increase their equity position. They are already doing that under Basel 3 with the banks considered most risky requiring just under 10% of risk weighting capital. Are MM suggesting even more is needed?
And how do they assume that as banks become more highly capitalized their cost of equity will fall. I would in fact argue the opposite. Investors don’t have to buy bank stocks. There is a huge suite of sectors choices available.
If banks normalized ROE begins to fall with additional equity infusions more investors will avoid the dilution and head elsewhere.
Leveraged banks have served us well. Raising equity is okay, but lets not carried away.
JC
11 Mar 13 at 5:44 pm
JC – Not M&M themselves. Rather Harford seems to invoke their theorem to suggest that banks should do so. As I say in the post, I’m not sure that works well as an argument.
Sinclair Davidson
11 Mar 13 at 5:50 pm
see http://marginalrevolution.com/marginalrevolution/2005/05/fischer_black_a.html
Black’s economic thought is centered around the view that all profit opportunities will be exploited. what happens if the central bank decides to add zeros to the accounts held at the Fed?
In Black’s view banks were already holding all the dollars they wished to.
One reaction is for banks to borrow less money at the discount window, or perhaps borrow less from each other. Money will leave the system as quickly as it entered. Another reaction is banks to sit on the new money.
Jim Rose
11 Mar 13 at 6:19 pm
That’s fine Sinc.
I was basically caught out by this glaringly silly comment.
First thing… Leverage in the banking system has fallen in the US over the past 5 or so years, yet equity hasn’t got cheaper. It became more expensive.
Furthermore, the Banks are trading at far cheaper multiples as Basel 3 looked like a done deal.
Secondly, I can’t see how existing shareholders would sit around get diluted and we’d see banks stocks rise as a result. That theory is nuts.
The secret of banking, if it’s indeed a secret is run a wafer thin equity line and leverage up while avoiding charge offs.
JC
11 Mar 13 at 6:25 pm
Sinclair,
To me the most important reason for having a bank is risk moderation through aggregation.
If you were going to try to rely on debt markets to provide the lending / borrowing inter-mediation then you would need an awful lot of individual instruments per loan to provide the sort of risk dissemination that a bank provides.
The problem with that is that you then also have problems paying and receiving interest and capital repayments with fractions of a cent. If you have (for example) $1m to invest in loans then even if you have $1 bonds to allow for 1 million borrowers to spread the risk around then to receive monthly payments you have very, very small amounts coming from each borrower.
Roll that up so you are getting $1 from each borrower per month and you are now looking at more concentrated risk of default.
You also end up with only allowing whole basis point moves in interest rates.
To try to create such a perfect market you would need to assume not only perfect information you would also need to assume near zero costs of money transfer and, crucially, money is nearly infinitely divisible so that the payments could be handled.
Anything else and you end up with concentrated risk with little upside in that risk – a strong disincentive to invest.
Driftforge – show me a bank that can create money out of nothing and I will show you a government institution. No commercial bank can do this.
Andrew Reynolds
11 Mar 13 at 8:15 pm
please government — put a super profit tax on banks . we guarantee them ,and they made 5000 profit for every aust man woman and child last year . (according to marcus paddely)
sunshine
11 Mar 13 at 8:43 pm
Sinc the major interest in discussing MM is why it does not apply in the real world.The answer is that its assumptions(eg.no taxes,no bankruptcy costs,no other imperfections)are not satisfied in the real world.Also,
this is not a new view-it came up when Basle1 was introduced.
Tom Valentine
11 Mar 13 at 8:49 pm
Commercial banks create money (deposits) through lending.
sdfc
11 Mar 13 at 9:19 pm
Andrew – yes. I had a two week module on the economics of banking in my financial economics subject and we would run through all that stuff. Starting with the no banks under a perfect market scenario.
Tom – yes. That is how I’ve always taught M&M. This year for the first time in over 20 years I won’t be teaching it. I’ve moved from financial economics to applied micro.
Sinclair Davidson
11 Mar 13 at 9:27 pm
Yes, even with perfect MM requirements banking would exist for depositors.
JC, I think he means that equity should get cheaper under perfect MM, which must be the case.
Pedro
11 Mar 13 at 9:51 pm
Welcome to micro Sinc.
Aliice
11 Mar 13 at 9:54 pm
Thanks – haven’t taught micro since the early 90s.
Sinclair Davidson
11 Mar 13 at 10:13 pm
Yes, but how? Why would existing and potentially new stockholders actually bid up bank stocks with a coming dilution and ROE falling?
JC
11 Mar 13 at 10:52 pm
Not really, it is perfectly viable to wipe out shareholders while still bailing out the company. Errr GM is the example that first comes to mind.
Then again, surely there aren’t too many people still believe that shareholders are influential in deciding company policy.
Tel
11 Mar 13 at 10:53 pm
Playing a Martingale. Looks like risk moderation, but actually just rejigging the risk profile. Something you really, really want to be doing with someone else’s money.
If you talk fast and wave hands, it does seem to many people to be the ultimate betting scheme.
Tel
11 Mar 13 at 10:57 pm
@ JC – isn’t your point really that it pays for a “large” bank to play chicken with the lender of last resort? That was what I took Harford’s point to be.
In terms of dilution – well, if competitive pressure (such as the credible withdrawal of implicit guarantees) drives their bank to have larger capital they can either forego dividends, or raise equity which, ex hypothesi, is a value-increasing exercise. Put blunly, having a half share in a solvent bank is better than owning all of one that is judicially managed.
@ Tel – being legalistic about who is a debt holder, who is equity only really shuffles the analysis around: “junk bonds”, or in local parlance, 2nd and 3rd charge or “mezzanine” lenders are really “equity” for this purpose, and create the same moral hazard issues if they stand to be bailed out. Take care in comparing US bailouts – - they tend to be means of overcoming otherwise “unbreathable” “contractual” commitments with state protection: one of the key benefits for GM was the ability to appropriate the value that had, by dint of state regulation, accrued to its dealers.
Pyrmonter
11 Mar 13 at 11:07 pm
!
unbreathable = unbreakable.
Pyrmonter
11 Mar 13 at 11:15 pm
“Yes, but how?”
Because the MM thesis is that the source of funding does not affect the value of the company. Real world advantages debt funding so the value of the stock should rise to reflect the higher profitability of the highly leveraged bank.
Pedro
12 Mar 13 at 2:35 pm
So why is a particular bank “too big to fail?” or any other company for that matter.
Let the business fail. The cost of bailing out the next round of mendicant businesses is far greater than that saved at the beginning.
Besides which, I thought that failure was the way we recirculated assets from underperforming businesses. Or is that not sophisticated economics any more?
Winston SMITH
12 Mar 13 at 2:46 pm
Another discussion of the book, worth reading too
http://www.slate.com/articles/business/moneybox/2013/03/bankers_new_clothes_anat_admati_and_martin_hellwig_argue_that_the_real_problem.html
And this comment about the trajectory of interest rates suggests this is the best time to make the jump.
“but the deeper explanation is that real interest rates have been steadily falling since 1983, from 7% to less than zero”
http://www.themoneyillusion.com/?p=19940
“So why is a particular bank “too big to fail?”
When it failing fucks up very large numbers of other people and businesses, which is something most failing non-bank companies will struggle to do.
Pedro
12 Mar 13 at 4:19 pm
@ Winston Smith re when is “too big to fail”. Tentatively, when its failure would lead to a material reduction in the outstanding credit stock, or more importantly, the supply of fresh credit. That was what I took to be the point of Bernanke’s financial disintermediation elaboration of Friedman and Schwartz’s monetary explanation of the great depression.
Pyrmonter
12 Mar 13 at 10:10 pm