Thursday, April 26, 2018

An unleveraged banking experiment.

I have written previously about my own confusion regarding the issue of bank capital.  The issue seems largely rhetorical to me.  It comes down to whether you call deposits capital, in which case you're an unleveraged money market fund, or you call deposits liabilities, in which case you're a bank.  The difference seems to be that insuring deposits turns them into liabilities.

There are so many debates in finance that seem to me like they could be solved simply.  Too big to fail could be solved by using private deposit insurance instead of public insurance, which would lead to prices that reflect risk.  Even with public insurance, I'm not sure what's stopping us from simply pro-rating insurance fees to reflect changing capital levels or size.  Money market funds seem to do just fine.  To the extent that there is some risk in NAVs falling below $1, it seems to me that a standard contract for investors could have a clause that if NAV ever falls below $1, then withdrawals must pay an additional 1% fee.  This would be a fairly insignificant amount, it would reduce panic withdrawals, and to the extent that there were still withdrawals, they would naturally push NAV back above $1.  In the rare event that this happens, it seems like this would be a stabilizing policy with little cost to investors.

I am sure that I am naïve on these matters and there are good reasons why some of these ideas aren't used.

But, regarding bank deposits, I would like to imagine a system with 100% capital requirements.  Instead of making deposits, depositors would just buy shares in the bank.  The returns depositors earn would not change much, because today depositors make up a very large portion of the capital available to banks, so the returns that currently go to equity holders would be spread pretty thin when shared among the new depositor/shareholders.

But, depositors want certainty.  In this regime, they could get certainty by selling at-the-money puts on their shares.  Depositors would make deposits or withdrawals by buying or selling shares.  The bank would mediate their asset base by buying and selling shares on the open market.  So, sometimes, withdrawers would be selling shares to depositors and sometimes they would be selling to the bank.

This would be a 100% capitalized banking system.  The put sellers would basically be taking the role that today's equity holders take, but with much less volatility because even failed banks usually only have capital shortfalls of a few percentage points of their assets.  In today's system, equity in a failed bank would fall to $0 if the value of assets fell below the value of liabilities.  This system would be more like a money market fund.  A failing bank whose shares had sold at $100 might now sell for $98.  Depositor/shareholders would exercise their puts, and the put sellers would now be shareholders.  The depositor/shareholders wouldn't lose a penny, and they would be free to reinvest their $100 back into the same bank at $98 per share or into another bank.  The main factor determining that decision would be how high the put premium was.  If a lack of confidence led to a run on shares, the bank could recapitalize by buying shares at the market price if that price fell below NAV.  The only losers would be the put sellers.

There could even be a public agency through the Fed that was a major put seller.  This would create a natural method for recapitalizing banks during nominal financial crises, because depositors would buy shares in other banks, and when the Fed funded exercised puts, it would be a natural monetary injection into the system that was automatic and didn't require discretionary decisions about which institutions to support.  Of course, this whole system would work better with some moderate inflation so that share prices tended to have an upward trajectory and exercised puts weren't triggered frequently.

In a way, this wouldn't be much different than today, where the Fed owns a bunch of treasuries and also holds reserves that they pay interest on, and monetary policy comes from managing both of those quantities.  In this system, they would earn income on treasuries they own and on puts they sell, and they would manage the quantities of treasuries and bank shares that they own from exercised puts.  It would sort of be a nationalization of the commercial banking system, because as a put seller, the Fed would basically be taking the risks that current bank shareholders take.

Today, banks are induced to take risks because the upside flows to shareholders.  In this system, that upside would still exist, but it would have to be earned through put premiums as income.  Riskier banks could offer shareholders higher dividends, but it would come with higher put premiums.  Maybe seeing that bank share prices rarely declined by more than a couple percentage points, few depositor/shareholders would even bother buying puts.

Since upside profits would be retained by the depositor/shareholders, put sellers would have less potential upside than today's bank shareholders do, but that would also translate into less downside risk.  They would mainly be trading a regular income stream for the occasional shock.  The main regulatory issue there would be how much leverage put sellers would be allowed to utilize when they sold puts.

Anyway, this is all academic.  But, I like to think about these sorts of things using a different rhetorical framework to try to think more clearly about these issues in a way that separates rhetorical factors from real factors.  Limiting ourselves to the rhetorical frameworks we generally accept seems like it leads to limited solution sets and to solutions that solve rhetorical problems when really, what we need are solutions to real problems.

I hope you haven't found this brief post to be a waste of time.  I welcome comments that point out how ill informed this post is.

10 comments:

  1. Check out George Selgin's excellent writings. He has a lot on 'free banking', ie banking regulated like any other businesses without special pleading.

    As for deposit insurance (I haven't read the whole thing): it's totally unnecessary and counterproductive. Here's how:

    If a customer wants a secured deposit, they should look for a bank that invests all deposits in government bonds only. No moral hazard, all very boring.

    ReplyDelete
  2. Well, I liked this post.

    John Cochrane, erstwhile free-marketeer, says commercial banks should have no leverage (and by law, too) and should only lend out equity.

    Ordinary depositors would go to other financial organizations, which would only invest in short-term government paper.

    This raises an interesting point:

    The Fed would no longer expand the money supply through reserves (that is commercial banks lending a multiple out of their reserves).

    I guess we would go to straight helicopter drops, or money-financed tax cuts (my fave). (Although I think Cochrane is in favor of a fixed-money supply, and deflation).

    BTW, back when I worked in the S&L industry as a very junior lobbyist, there were still "mutual savings banks"---owned by the depositors. That was the 1980s. Maybe there are still some left. I think credit unions are owned by depositors too.

    But as you suggest, the financial industry is not going to be reformed, even by the admirable and sensible application of sound free-market principles, or classic economics. The industry has captured the regulators (and legislators) and that is that.

    It is like my howling about property zoning (another type of regulation loved by the lending industries).

    We will wait many a moon to see an unzoned America and free banking.







    ReplyDelete
  3. Some countries already have freer banking (and less zoning) than others.

    We don't need to wait for perfection to arrive, steps in the right direction are good. (Either by political activism, or much more effective of an impact on your personal circumstances: vote with your feet.)

    ReplyDelete
  4. Kevin, especially George Selgin's historically overviews of historical episodes of free banking. He also does theory, but the historic stuff is endlessly fascinating.

    ReplyDelete
  5. See eg this piece about George Selgin's book about British industrialist's making their own coins at the height of the first industrial revolution, because the Royal Mint wouldn't provide enough change:

    https://marginalrevolution.com/marginalrevolution/2008/09/good-money.html

    ReplyDelete
  6. In theory what you said would be all true, but in practice I doubt they would be realistic. If we allow all banks to be equity financed purely and have depositors purchase puts themselves to get de-facto deposit insurance for 100% certainty, depositors simply won't purchase puts, because they would think "my bank has never gone through a write-down in the last 10 years". When some banks do go down, the depositor would demand federal money using political power, and effectively we are only shifting leverage from banks to the shadow banks, and shifting moral hazards from the more aggressive bankers who engage in risky positions to the more opportunistic depositors who deposit to riskier banks that funds the shadow banks.

    You see this dynamic already playing out in countries like China, where more middle-class deposits go to the shadow banking system instead, thanks to higher interest rates that the money market funds pay, and higher marketing exposures these funds get. However, no depositor go to the option market and price their "real interest": if a money market financial product promises 8% return, depositors flock to that fund instead of 4% at the bank, even if they want 100% safety and even if they see the assets backing the fund has large put premiums - people simply aren't sophisticated enough to price low percentage of probability of liquidity risk. This is all despite the omnipotent party telling everyone to not trust these funds as “safe”. If even the communist party couldn’t convince people that these deposits are not “real deposits”, I struggle to see how a democratic government would - in a actual crisis, the people would just elect whoever would give free bailouts to these banks.

    Fundamentally humans generally underestimate small-probability events and bank’s job in our current economic system is to maintain 1:1 par between deposit and money to make the economy efficient, despite the irrational and inefficient human psychology. Unfortunately this means we collectively pay a price in the form of subpar banking regulations - it’s hard for me to picture how we can get out of this.

    ReplyDelete
    Replies
    1. Thanks for the thoughtful comment.

      In the US, though, it seems like money market funds held up pretty well, and didn't really require much targeted federal support. The Reserve Primary Fund "broke the buck", and eventually repaid shareholders slightly less than a dollar. So, it seems that depositors in commercial banks in the US get lower yields and also create greater demand for federal intervention.

      Delete
    2. That's definitely a good point. I guess I might be a bit paranoid in this regard, but I suspect US mutual fund being relatively stable is precisely the result of the commercial banks shouldering greater risk - if we do move the commercial bank to a more 100% capital-financed system, the money market funds would take in all the risk, and potentially more risk than their bank counterparts due to lack of banking regulations. I do concede that this is all just random thoughts on counterfactuals at this point.

      Delete
    3. Good point. Somewhere in the financial sector, there has to be some maturity transformation. Major overhauls in the system would push those risks somewhere.

      This is one area where I wish the reviews of banking during the boom weren't so focused on leverage, risk-taking, etc. There were pockets within the finance sector, like in the investment banks that were underwriting the securities, where leverage was part of the business model, and peculiar risks developed. But, of the trillions of dollars that were flowing into the mortgage securitization markets, it seems that most of that capital was from individual savers, pension funds, etc. These were unleveraged savers who were matching their maturity needs with the maturities of the securities they were buying, and they were doing it outside the commercial banking sector. At the time, I was on an investment committee for a small endowment that had a significant holding of Ginnie Mae securities. Just spit out income for a fund with a long-term focus. No banks or leverage involved. That type of ownership doesn't get many headlines, but it seems like this describes a large portion of the market of investors.

      So, in part, I think that it could be the case that financial markets have become sophisticated enough that investors can invest in long-maturity debt instruments without requiring commercial banks to do it. Maybe if we didn't induce the commercial banks to engage in maturity transformation, we would find that they are useful for taking on local credit risks, but that we don't really need them to borrow short and lend long. Maybe they could borrow long and lend long, and savers would be able to deposit liquid capital into institutions that only had short term assets.

      Delete