How to Reform the Banking System

Good financial reform would accomplish two goals 1) prevent systematic meltdowns of the entire credit sector (and prevent runaway credit expansion) 2) allow for bottom up, entrepreneurial decision making about who gets credit. We do not want a top down system where a couple private agencies or government regulators set everyone’s credit rating. In that world the big corporations game the system, get cheaper credit, outbid the small players for real estate, and litter the country with cookie cutter box stores.

The current financial reform bill accomplishes neither goal. The reforms allow banks to make up the risk weighting of their assets themselves. The bills preserve the pro-cyclical nature of the system. Banks are still making mortgage loans with ridiculously low down payment and income payments. Regulations and “too big too fail” ensure the biggest corporations still get cheaper credit.

If the system continues to function without further catastrophe, it’s only because the players are smiting from the recent crash and the cultural shift and increased scrutiny will keep things cool for a while.

So what does good financial reform look like? The following is a plan that accomplishes the goals above (no systematic risk, bottom up decision making) with the least amount of changes to the existing system.

1) Extend FDIC insurance to cover all deposits, of any size. Limit the insurance to covering only 95% of the deposit, not the full 100%. Extending FDIC coverage to deposits of any size will kill the money market funds, which were a key cause of the latest crisis. The 95% limit is merely to make depositors at least pay a bit of attention to the reputation of their bank, rather than shopping purely based on the highest interest rate.

2) Eliminate the Fed discount window. Make banks responsible for both credit risk and maturity risk/interest rate risk. If the bank cannot redeem a deposit according to its promised terms, the bank goes into default and the equity holders get wiped out. Allowing access to the Fed discount window is a huge back door subsidy to banks. It allows banks to game their books and pretend they’re not really insolvent, even though they do not have cash to make depositors whole. The discount window absolves banks of the responsibility for managing duration/maturity/interest rate risk, and therefore makes the system as a whole much more cyclical and prone to periods of credit over expansion and then over contraction.

3) Ban securitization of loans. Banks must keep all loans on the books. If institutional investors want to invest in a portfolio of long term loans then they should buy a bank CD. Eliminate Fannie Mae, Freddie Mac, etc. Allowing banks to get their loans of the books defeats the entire purpose of having a regulated, insured banking sector. A separate bond market should still be allowed to exist. But there should be a very clear line between the regulated, FDIC insured banking sector, and the rough and tumble uninsured bond and credit markets.

4) Convert capital requirements into a cash escrow requirement. The cash escrow would come out of the bank’s own equity, it is not a reserve of deposits. The cash in the escrow account is senior to all of the banks obligations other than insured deposits. Capital is an accounting construct that can always be gamed, especially when executives have a strong incentives to fudge the numbers. Cash is real and verifiable. The cash escrow would be set as a percentage of insured deposits loans - perhaps at around 5%. Capital requirements also make banks overly cyclical. A bank may have made great lending decisions, and fully able to meet all its promises to depositors. But due to a market crunch the current market value of its loan portfolio might be depressed and thus the bank is viewed as insolvent when in reality it is healthy. Conversely, a market boom might make a bank look very healthy when in fact the bank is extremely exposed to interest rate risk.

5) Banks should be banned from buying back their own stock. Dividends must be held in escrow for five years. The FDIC can also regulate bank executive compensation, perhaps requiring that compensation be held in escrow for X number of years, and adding clawbacks into executive pensions if the bank goes under. The FDIC can tweak compensation so executives are not playing a game of heads I win, tails the government bails the bank out.

6) For every $100 in insured deposits a bank accepts and loans out, it must accept $5 in uninsured deposits. The FDIC can monitor the delta in interest rate between the insured and uninsured deposits. If it gets too high, that’s a sign the bank is making dangerous loans and perhaps needs stricter scrutiny or regulation.

7) The FDIC will not overly worry about top down regulations about the kinds of loans the banks make. Rather it will make cash escrow requirements high enough, make the bank stockholders have the first loss position, and design stock holder and executive compensation in a way to encourage long term thinking. Then it will be up to the bank executives and lending officers to use their own personal judgment about what is a worthy loan. There will be no top down, nationwide credit rating agencies. Only if the interest rate on uninsured deposits diverges rapidly from the insured deposits will the FDIC step in and implement more heavy handed regulations.

8) The FDIC is split into seven separate organizations, one for each region of the country. The FDIC must financially break even - it should redeem depositors with the insurance premiums it accepts from banks. Each regional FDIC has a completely independent board, management, and complete regulatory discretion. Forcing the FDIC to break even forces to the FDIC to be disciplined, and forces it to make had decisions about the trade offs between encouraging lending and limiting risk taking.

9) Each FDIC is allowed to set regulations (cash escrow requirements, divided escrow requirements, compensation claw-backs for executives, limits on the riskiness of loans, down payment and income standards for mortgage lending, etc) for its member banks. The FDIC executives have full power to use their own personal judgments in regulating its members banks. The FDIC VP of Bank Scrutiny can look at the books of Springfield Savings & Loan and say, “Your loans technically fit the risk models, but these models are way too complex, and I think there is something fishy going on. Clean up your books, use a simpler, transparent model, or we’ll withdraw FDIC backing from your bank.”

9) If a regional FDIC goes under, the Fed steps in and makes depositors whole. But all the executives of that regional FDIC get fired and their pensions are slashed.

10) The compensation of the FDIC executives and/or board can be tweaked to set the overall amount of risk in the banking sector. Here are several ideas for making compensation reward conservatism and low risks: a) all executive compensation over $50K is held in escrow for seven years, and clawed back if the FDIC needs to be bailed out. b) executives get reverse options (stock options with $0 exercise price, with a payoff price that is the lesser of the stock price at issuance and the stock price when exercised/sold). c) if the FDIC goes under executives have their pension slashed in half. These might actually make the executives way too conservative. The exact levers should be tweaked to get the risk level that policy makers want.

11) If the government wishes to step in and lower interest rates in order to encourage credit, it should simply subsidize lending directly by offering matching deposits for every loan issued.

12) If the government wishes to control the inflation rate it should do so via balance sheet effects. To generate prevent deflation/stimulate aggregate demand the government should have a mechanism to multiply everyone’s deposit account by a percentage. To prevent excessive inflation, the government should raise taxes without increasing spending, especially taxes on capital gains, real estate sales, and gold sales.

UPDATE: A couple points I forgot - the existing money market funds should be converted to standard, FDIC regulated banks. The money market funds are doing the same thing as banks, and they have received government backing, so they should have the same regulations and incentive structures as proposed above. Also, the FDIC in my proposal would act much more like a real company. Executive incentives would be set in a way to reward conservatism, but the overall compensation would be much higher than it is now, so it could better attract top talent.

UPDATE - a response to commenters:

A few commenters have pointed out that my plan has no chance of happening. Of course they are both correct. I am a software developer, not a policy director for the Obama administration. I create my proposals as foils by which you can judge other proposals, including the policies actually being enacted.

What’s interesting about my proposal is that in theory it should be politically possible. It’s in the interests of the majority of the American voters. It’s quite compatible with both conservative and progressive principles (conservatives may not love the part about increasing FDIC insurance, but overall I think the plan is much more free market than the status quo). Banking interests may or may not have liked the plan, but Obama was not beholden to banking interests, since his campaign support was so broad based. Congress may have posed an obstacle. But public anger over the crisis and the bailouts was large enough, that Obama could have given a rousing speech and then enacted the entire reform via executive order. No one would have stopped him, and his approval ratings would have shot through the roof.

Instead we still have a banking system that’s an unholy mix of socialism and kleptocracy. The government is propping up an insolvent banking sector and originates practically every mortgage loan in the country. Yet the bonus money still flows at Goldman-Sachs and a stream of MBA’s still make their way from Wharton to Wall St. in search of earning riches as a reward for mis-allocating capital. The financial reform bill has few defenders, and many biting critiques.

The big question is why can’t a sensible reform be passed? Why is sensible reform not even being talked about? I have my own theories. But I’d be curious to hear others, especially from Foseti who works in this area, or from any of my readers who still believe in the American political system.

Now let me respond to a couple other points by Aretae:

The banking firms are smarter (collectively) than any possible regulator or regulators. The financial firms have more money at stake, and they are necessarily smarter. Read Kling on the regulatory chess game.

Aretae must have misunderstood the core of my proposal, because I basically agree with him. My proposal ditches the whole regulatory whack-a-mole game. I dump the ratings agencies, risk weightings, centralized regulations on lending, accounting based measures of bank capital, etc.

The essence of my proposal is that a) banks fail when they cannot pay their depositors and b) the incentive scheme of the bank executive holders and equity holders is designed so they cannot play the game of paying out excessive dividends and bonuses just before the bank blows up. I do this by clawing back dividends and bonuses, seizing a cash escrow posted by the equity holders, and cutting executive pensions. Aretae’s proposal to return to double equity exposure would also work.

I do have one separate point of disagreement with Kling. Kling seems to think that the regulators are smart and well intentioned, by due to human limitations they cannot have enough foresight to see how their regulations will create unintended consequences.

In my take, regulators should have known better at the time. For instance, fixed rate mortgages became a crushing burden to banks when interest rates rose in the 70’s. The policy response should not have been to get rid of fixed interest mortgages. It should have been to keep inflation rates and interest rates stable and predictable. There is simply no excuse for the kind of monetary instability the U.S. has experienced over the past forty years. It’s not an issue of smart people acting with too much hubris. It’s an issue of a political and intellectual system that’s so messed up its incapable of enacting and sticking to any sort of sensible, well thought out reform.

If you were doing something…abolishing Freddie, Fannie, and the FDIC would be a start.

I agree 100% with abolishing Freddie and Fannie.

But abolishing the FDIC is absolutely not the place to start. Right now the U.S. banking system is basically insolvent. Two hundred eighty banks have been allowed to fail in the last two years. The major Wall St. banks would have failed if they hadn’t been propped up. If FDIC insurance was removed, the depositors at those banks would have faced significant losses. If a depositor in one bank takes a loss, depositors at the other banks start running to their latest branch. Banks in 2010 simply do not have the reserves to redeem even a small fraction of the panicking customers. The entire system would collapse, people’s banks accounts would be frozen or disappear, spending would collapse, profits plummet, and unemployment would shoot even higher. Daily life would repeat the experiences described in Benjamin Roth’s diary of the Great Depression (which is required reading for anyone trying to understand the Great Depression).

If you really wanted to abolish the FDIC, it would have to happen as the last step of my proposal. Once the rottenness is purged from the banking system, once executives and equity holders have been given the proper incentives, once banks have operated for a while under the knowledge that they will actually fail for making bad loans, once banks have gotten used to not having a reserve window and having to match maturities, and once natural selection has taken its course and allowed the sound banks to flourish and the bad banks to fail, then as a last step you could remove the federal backing of the regional FDIC’s, letting them free as a fully privatized, free market insurance company.

But abolishing the FDIC as a first step would be utter catastrophe. This is a general problem with libertarianism. Libertarians make great critiques of the bad routes policy has taken over the last century. But they have no idea which color wires to cut, and in which order, in order to defuse the government apparatus without blowing everything up. Progressives then (accurately) point out that the libertarian policies would leave the economy in a pile of rubble. Unfortunately, this has the effect of discrediting all the things that libertarians actually got right.