The large Franco-Belgian bank Dexia went effectively bust last week and had to be bailed out for the second time. Two numbers: the Belgian retail business was nationalised by the Belgian government, with deposits of €60bn. The total balance-sheet of the bank is over €500bn: over eight times the size of the retail bank.
The recent Vickers report on UK banking draws a similar picture of British banks. The Vickers proposal is to ring-fence “£1.1tn-£2.3trn” of retail operations, within total bank balance sheets of £6trn (report, paragraph 3.40, page 52.)
What are the banks doing with the other £4-5 trn? Lending it abroad perhaps? The latest BIS number for the total international claims of British banks is £1.45trn. That leaves £2.5-3.5 trn of domestic non-retail assets to be accounted for.
The FT’s fine pundit Professor John Kay has the explanation :
Most of the £6,000bn total of UK banks’ assets and liabilities represents financial institutions trading with each other.
The UK banking system is an extreme case of hypertrophy, but it’s not alone. The BIS gives the total of international interbank claims (2011 Qtr1) as $8.4 trn.
Most economists are subject to what you might call reality illusion. They think the driving forces of the economy lie in real economic transactions and decisions by households and firms over borrowing, investment, asset allocation and spending. Interbank and other purely financial transactions are unimportant housekeeping.
Really? As Paul Krugman wrote recently on banks in another context:
This is, after all, the 21st century. Things have moved on a bit.
Here’s my wee contribution to postmodern financial theory. Thesis: much of the the financial industry has no connection with the real economy at all, except through its own superprofits, perverse risk creation and huge demands on capital.
A thought experiment. Consider four “investment” banks with no retail operations and no loans to nonfinancial companies. Let then be:
- the South Sea Bank Company
- la Banque Royale du Mississippi
- Banco Ponzisssima da Corleone
- Münchhausen Bankgsesellschaft AG
They have access to their central banks. Its CB lets the South Sea Bank have £100m of monetary base. It can now make a loan to Münchhausen Bankgsesellschaft, which in turn lends it on to Banco Ponzisssima, which lends it on the Banque Royale du Mississippi, and finally back to the South Sea Bank. By the magic of fractional reserve banking, the £100m swells to £1bn of loans. But nothing has happened in the real economy. The circuit of such pure banks is a black hole: it sucks in capital, talent, and base money, and produces nothing except its own profits.
Ah, you object, how can the banks make any money on the turn? The answer is that at each stage they lengthen the maturity and add risk in other more advanced ways, with options and swaps and so on. They charge high fees for this negative expertise.
Now of course the thought experiment is unrealistic. It should be widened to nonbank financial corporations, such as hedge funds and financial insurers. This increases the scope for profit and risk generation without necessarily or even probably engaging the real economy in any other way. Eventually, a small proportion does leak out into boring loans to households, businesses and governments - all of which adds the risk, as it’s not just that Greece may default, you have to worry about your loan to Dexia which holds a lot of Greek debt. Still, can you show me that my financial bubble network is impossible?
If the model is even partially true, it has explanatory power over the liquidity trap: it’s not households and non-financial businesses preferring cash to bonds, the cash never reaches them. Monetary expansion in current circumstances is pushing on a cosmic string.
By the magic of fractional reserve banking, the £100m swells to £1bn of loans. But nothing has happened in the real economy.
What’s confusing is this: how can something that amounts to “nothing has happened in the real economy” when it is done, cause real damage to the real economy when it comes undone?
If I owe you a billion pounds, and you owe me a billion pounds, and we suddenly discover that neither of us can pay the other back, how does that affect the price of cheese?
I am absolutely NOT denying the real-world usefulness of “fractional reserve banking”; I’m only trying to understand how a poker game among a group of bankers can result in ALL of them losing real money. By “real money” I don’t mean “big money”; I mean money on this side of the event horizon.
-TP
Keep in mind that, as James points out, in the real world, there are other actors besides the banks. A part of what happened in 2008 is that companies that did need to be backstopped, most critically AIG, owed money to parties that were outside the regulatory system, such as hedge funds. As others have pointed out, there are ways that the money can simply disappear, but it can also end up in places where it only comes out from behind the event horizon in ways that it is laundered and not obviously the same money.
There is nothing new under the sun. James Wimberley is recapitulating Henry Thornton’s demolition of the real bills doctrine-in 1802. It’s as true today as it was then-a ziggarut of financial debt can be built on a small base of real debt.
My first question would be “are corporate payments clearing through retail?” Because a lot of corporate transaction accounts aren’t (in the US system) technically retail accounts, although they function just like retail accounts.
Tony P: In my model, it’s true that all four can go bust and good riddance. In the real world, not so because the poker-game side is enmeshed with retail functions: the payments system, trade finance, mortgages and so on. That’s why it was a mistake letting Lehman, a straight investment bank, fail. I remain sceptical whether Glass-Steagall redux, as proposed by Vickers, will allow regulators to wash their hands of the investment side. For one thing, as I argue here, they need monetary policy instruments that impact retail more directly.
Ebenezer Scrooge: thanks for Thornton, I’ll look him up - but I’m off to Brazilian beaches in a few hours. I won’t be able to monitor this thread either. Stay cool all you commenters!
My own preference would be a sort of Glass-Steagall Plus. Of these four industries — commercial banking; asset management and research; underwriting offerings of securities; proprietary trading — a business entity would be permitted to participate in one only. Neither it nor any of its affiliates could engage in any of the other three businesses. Oh, and any entity engaging in the last two would have to be structured as a partnership, and not the limited-liability kind. They could play with their own money only, if they’re prudent. Or they could lever up, but then if they screwed up they’d lose their house (instead of thousands of innocent middle-class bystanders losing theirs, which seems to be the result under the current dispensation).
Oh, and sorry, just can’t help myself:
Consider three “investment†banks…
>> the South Sea Bank Company
>> la Banque Royale du Mississippi
>> Banco Ponzisssima da Corleone
>> Münchhausen Bankgsesellschaft AG
Our three weapons are fear, surprise, and ruthless efficiency, and an almost fanatical devotion to the Pope!
It’s not just that the poker game is enmeshed with retail functions. Two other things: 1) Bankers like to buy houses and yachts and champagne (or maybe now it’s “escorts” and cocaine) and so pay themselves salaries and bonuses that siphon off a bunch of that money. So after a few years that hundred million has all gone into transaction fees, and the remaining pile is completely empty. 2) Because they’re playing a shell game, they can offer better (apparent) returns on average than any transaction involving real-world operations (this is true by construction: if the deal doesn’t offer better apparent returns, no one buys it, it doesn’t happen, and you cobble together another deal that does). So pretty soon people operating in the real world will want in on the game, and whee!
The whole thing reminds me more than a little of renormalization-style physics, where you have infinite clouds of virtual particles all almost cancelling one another, unless something catastrophic happens (e.g. the disintegration of a black hole). In renormalization finance, you have clouds of multibillion-dollar obligations, almost all existing for some brief time window so that I can pay you 10 grand tomorrow if interest rates go up and you can pay me 11 grand if they go down, but the $10 billion on each side of the transaction is just a way of making the math work out. Unless something catastrophic happens.
Tony,
Don’t forget that in the post the South Sea Bank was “given” £100 million by the central bank. In real life that would come from depositors and investors in the bank. Their feta consumption will likely drop when the bank goes bust.