In defense of banking …

… but not the banks we have.

Opacity is absolutely essential to modern finance.” “Societies that lack opaque, faintly fraudulent, financial systems fail to develop and prosper.” Steve Waldman

This is a response to three posts at Interfluidity that argue that banking is essentially a con job.   The quotes here are from those posts except as noted.

While I agree on the issue of “opacity” (bankers lend money and it is difficult or impossible for depositors and creditors to monitor bank assets), I also think that Waldman is confusing two things:  modern finance and the financial systems that facilitated economic development.  They have/had very different characteristics:

  • Modern banking:  limited liability, government-guaranteed deposits, high risk of loss to bank creditors (in the absence of government intervention).
  • Before 1930:  at least double liability, deposit “contracts” that were always implicitly conditional, and comparatively low risk of loss to bank creditors (in the absence of government intervention).

In short, prior to the 1930s the assets of the bank-owners themselves were very much at stake with the result that their leverage ratios were much, much lower than in modern times.  (Econometric analysis indicates that  in the late 1800s neither the leverage ratio nor the size of banks was a predictor of either bank profitability or stability of profits.   Lamoreaux, Insider Lending, pp. 97-98.)

History:  Unlimited liability and the nature of the deposit “contract”: 

“A banking system is a superposition of fraud and genius”  “Banks guarantee all investors a return better than hoarding, and they offer this return unconditionally, with certainty,”  “First and foremost, they offer an ironclad, moneyback guarantee.”

Waldman’s model of banking has these properties, perhaps, because he’s referencing a world with deposit insurance instead of bank failures – that is, a world with explicitly government backed deposits:  this form of banking has been around for just three quarters of a century.  Historically banking is built on a conditional promise to depositors to return the money unless the bank closes its doors.  When the bank closes its doors, “time” is lost, but because nobody would entrust his money to a banker who wasn’t wealthy and subject to liability (either unlimited or capital calls on stock owners to pay debts), once the banker’s horses, etc. are sold, you get your money back.  Depositors understood that “time” could be lost and that only a fool would leave his money with someone who wasn’t wealthy.  The contract was conditional on the risk of temporary illiquidity and, in practice, conditioned on the assets of the banker.

To the degree that Waldman’s concept of banking is based on the Diamond and Dybvig model, even they acknowledge that the 19th century solution to the problem of bank runs, suspension (i.e. closing the bank and liquidating) also solved the problem.  It is only dominated by deposit insurance if the insurer has the ability to target a tax to those who withdrew money during a run (or to achieve the same result via inflation).

History:  Bank losses were put to the owners of the bank:  While crises and bank failures were common, there are many historical examples of major bank crises that didn’t result in losses to depositors’ principal.  (See for example the failure of Overend and Gurney in England, which was the Lehman’s of 1866.)

“Second, they point to all the other people standing in front of you to take the hit if anything goes wrong.”

This is a pure invention of modern “limited liability” banking.  Even through the early years of the 20th century, bank shareowners were subject to capital calls in the event the bank failed.  In Britain bank stock was usually purchased by paying only a fraction of par, leaving stockholders subject to a capital call that could be a multiple of the original payment.  In the US the norm was double liability, so stockholders who bought shares with a par value x, were liable for a “second” capital call of x. (Double liability was part of the National Bank Act of 1864 and also enacted by most state legislatures.  It was the law in Massachussets even earlier.)  It’s not clear that losses of principal (i.e. excluding “time”) to depositors due to bank failures were either common or large.  It’s my understanding that there were no such losses in England from 1850 through 1900 – and probably very few between 1800 and 1850.  In the US there were most likely some losses, but how significant were they after the whole legal process was complete?

 “ ‘Shadow banks’ are nothing new under the sun, just another way of rearranging the entities and guarantees so that almost nobody believes themselves to be on the hook.  This is the business of banking.”  “Since no one (most especially the financiers) believes themselves to have agreed to be the bagholder, we are left in an ocean of conflict over who must bear what costs.”

I disagree.  This is the modern business of banking.  When our financial system developed the ones who were most at risk were the bankers themselves.

Bagehot’s Lombard Street is in many ways about the growth and contraction of a shadow banking system.  The “bill-brokers” — of which Overend was by far the largest — grew up as undercapitalized shadow banks that relied extremely heavily on the lender of last resort facilities of the Bank of England in crisis.  The Bank objected to the quantity of Overend bills it had discounted and told Overend/the bill-brokers that they would have to find their own source of capital in the next crisis.  In 1866 the Bank of England kept its word and Overend failed.  As noted above, the Overends had to sell their personal assets (and there was a capital call on shareowners), but every penny was repaid to creditors.

After this crisis, British finance continued to flourish.   The lesson I draw from this is that the key to central banking and a strong financial system is knowing when to close the liquidity spigot.

In fact, it is possible that the problem created by bank failures in the depression may not have been losses to depositors (although they were certainly the more visible), but the wiping out of the stratum of society that had the reputational and financial assets to be bankers or invest as shareholders in banks (i.e. the capacity to answer and the willingness to risk a capital call).

Opacity, confused creditors and growth

“Societies that lack opaque, faintly fraudulent, financial systems fail to develop and prosper.”
“Industrialization occurs in societies with corrupt and fragile big banks.”

From a historical perspective this is unsupported by the evidence.  Societies that develop and prosper have banks that pay their depositors back from the assets of the owners of the bank, should the bank fail.  This is the reason bankers were trusted with “other people’s money” in Venice, in Amsterdam (each, the center of world trade in its time), in 18th and 19th c. England and in the US through the 1920s.  Only in modern times did anyone imagine it is possible to have financial system populated by entities with complete limited liability.  (Observe that banking systems will full liability also can collapse – usually because the flourishing economic activity that made it worthwhile for bankers to take the risk of losing all their property evaporates for external reasons.)

I think the truth that underlies Waldman’s argument is that it’s impossible to have a truly competitive financial system, without destroying the financial system itself – and this is one of the greatest errors of modern finance. Why?  Historically, bankers must take risks and their assets must be at risk; bankers will only do this if they make a hefty return.  In modern times, they must be highly compensated simply to keep them from taking advantage of their access to other people’s money and their abundant opportunities to profit from activities that tend towards fraud.  They will be wealthy, they will control the payments system, and they will favor policies that are in their own interests, not those of the general public.  They will be the target of populist complaints.  If you try to address these complaints by introducing too much competition, you will compete away the profits that will compensate them for their risk (or their access to pilferable funds).  Theory tells us that given too much competition, bankers will decide they’re better off stealing people’s money.

(cf. “Information asymmetry (the source of opacity) is built into the very fabric of the division of labor in complex societies. …  insiders’ privileged position and knowledge enable them to extract rents.”  )

Banking and capitalism

Bankers have historically been the “capitalists” of their time, but only if one understands “capitalism” in the Marxist sense of the word.  Bankers are the wealthy making profits off their wealth.  In modern times where non-wealthy individuals enter the banking profession, the bankers are profiting from their position in a company that has vast amounts of money at its command.  Because these companies have limited liability — and abundant government support systems — however, society, not the owners of the bank are the true sources of their capital.  When profits are being made off of capital provided by the public, a better word for what the bankers are doing is probably best characterized as socialism.  (cf.  Prof. Varma on safe assets, h/t Ashwin Parameswaran)

“Part of what makes an FDR different from a Mitt Romney is that an FDR understood his power to be derived from more or less arbitrary privilege, while a Mitt Romney imagines himself to have “eaten what he killed” in brutally efficient markets.”  Yes.

“They persuaded themselves, long before they persuaded the rest of us, that any games they played for their own enrichment would necessarily lead to social gain over the long term.”  This is very true too.  One thing I’ve learned from blog discussions is that traders think it’s socially efficient for them to make money off of the sorry, uninformed creature who thinks that asset is worth what he’s paying for it.  (They  obviously missed the Econ 101 class on asymmetric information and efficiency.)

“What the market will bear” is not a sufficient statistic for ones social contribution. Sometimes virtue and pay are inversely correlated. Really! People have always been greedy, but bankers have sometimes understood that they are entrusted with other people’s wealth, and that this fact imposes obligations as well as opportunities.

I’m not sure that the right approach to this problem is to place “obligations” on bankers.  I’d opt for incentives.  Jamie Dimon’s houses and financial assets should go on the block to pay creditors if J.P. Morgan fails.

Would this reduce risk-taking?  Absolutely.  But in a good way.  Bankers need to be incentivized to take reasonable risks, not unreasonable ones.  Keep in mind that industrialization and modern banking was born in an environment with precisely these characteristics.

Does (financially-supported) growth require misinformed bank creditors? No

“I claim we would forego a lot of plain booms, the kind that ultimately enrich investors as well as society at large, if we didn’t have a financial sector skilled at getting people to assume risks they’d not directly consent to take.”

I think that what Waldman is missing here is that historically bank accounts in industrial regions start out as loans.  That is they are discounts not accounts.  So no bank creditor is putting their savings at risk, instead the bank (as noted above most likely the wealthiest person in town) is lending to the local entrepreneurs and tradesman, but avoiding having to maintain much of a stock of gold by clearing the whole town’s transactions on his books.  In a well functioning financial system, the town banker can send those debts to the big city and sell them off if he happens to need cash.  There ends up being a large network so that long-distance transactions can be cleared in the big city.  Basically finance is a debt clearing system that requires, in theory, no capital or actual savings at risk at all — but only a bunch of entrepreneurs/tradesmen who are willing to lend to each other because an entity with reputation (the bank) is underwriting the loans. In practice, people with money to lose have to guarantee that the system is sound.  These bank-guarantors make it possible for the system of unsecured trade credit that makes the economic world go round to operate.

Can finance solve the problem of systemic risk?  No, and neither can government.

“Diversification and maturity transformation can protect us from idiosyncratic shocks, and Murphy is right to point that out. But they cannot protect us from systematic misfortunes.”

Finance is unstable, so if a basic precept upon which most banker’s decisions have been based (e.g. the Bank of England will maintain its peg to gold) a lot of bankers can fail and lose their fortunes, taking the economy down with the bankers. (The worst of the bank failures in the Depression took place around the same time or after the Bank of England went off gold in Sept 1931 — causing a big hit to balance sheets around the world.)  Bagehot recognized that banking is a system of credit, so by definition its always within the realm of possibility that the whole system collapses.  There’s no reason the system can’t be born again – as long as there’s someone with assets to lose who recognizes trading activities he’s willing to finance.  But, of course, the process is excruciating.

But on systematic risk allocation, I think it unquestionable that status quo finance is completely terrible.”  Agreed.  But this is not a problem with banking.  It’s a problem with limited liability banking where the people making the lending decisions can put the ultimate losses to someone else.


If regulation will be very intensive, we need regulators who are themselves good capital allocators, who are capable of designing incentives that discriminate between high-quality investment and cost-shifting gambles.

I’m not convinced there is an alternative to the threat of personal losses to keep managers of “other people’s money” honest.  The banks have too much incentive to infiltrate whatever regulatory system is put in place, so it’s probably close to impossible to keep it from being captured in one way or another.

Have we reached the apotheosis of finance?  That is, did we once have a functional (though always changing) financial system that evolved to the degree that it is now on the path to implosion?  Far from Waldman’s view that finance has always been this way and this is the way it must be for growth to take place, I think that finance once had very different characteristics from that which it has today and that the loss of those characteristics significantly increases the likelihood of the total collapse of which  Bagehot warned.

The questions, for me, are:  Is the financial system dying from the effort to turn it into a competitive industry?  And can we save it?

Note: I edited  a phrase that was too strongly worded, and tweaked the text.