Tuesday, April 27, 2010

What I want to see out of Financial Reform, v 1.2

The financial reform argument is very strange - it's a hard enough problem to solve for people who actually get the banking system. It's almost impossible for Congresspeople.

That said, I do have a list of things I'd like to see and a list of things that I think will be problematic. You'll notice some trends - I'd much rather see a system where the consequences for bad behavior are strongly negative but not destabilizing (numbers 1 and 3 on the ideas I like, for example), issues are made transparent (number 2, 8 and 9 of what I like), markets are more equal and less prone to manipulation (4, 5, 7, 11, 12, 13 in what I like) and destabilizing regulations are avoided (6 and 10 , and also sort of 5 and 7 on things I like, and 1, 2 and 3 on things I don't). In other words, make sure that people acting in their own interests contribute to a stable system - the definition of decentralized. I dislike a system that requires regulators to be smart (2 and 4 in stuff I don't like) or risks political pressure (12 on things I like, 4 and 5 on things I don't). I'd rather see a degree of corporate efficiency and competition that benefits consumers (1 on things I don't like.).

In other words, this is a thoroughly "conservative" approach to the banking system - not conservative in the political spectrum sense, but conservative in the "let's keep things that we know definitely work, instead of throwing babies out with bathwater". The Democrat vision of centralized and regulated everything is something I definitely don't agree with.

Ideas I like:

1) I think I really like pre-funded bailouts and living wills. Every bank would determine how it would be broken up, in real time, with mandatory and frequent regulatory reviews. Unlike the other "regulation" forms, this is something which the government actually should be decent at. Each bank would have to fund its own pre-funded bailout pot, because otherwise there's an incentive to take extra risk and be bailed out by your competitors, which is even better than getting bailed out by the government. They'd also have to be structured so they don't encourage banks to rely on them  - perhaps banks pay a substantial interest rate on the funds they draw from their pot, and once the fund is used, banks can't book any positive earnings until the pre-funded bailouts are replenished to the levels they should have been at.

2) I love the idea of contingent convertible debt, as proposed by Mankiw, in which banks issue debt that is converted to equity of a regulator deems them insolvent, would be tradeable, and thus provide an excellent leading indicator of bubbles. It won't be perfect, but it will be generally pretty good, and can tip off regulators where to look.

3) I've mentioned an accelerating bailout tax before. I've never seen anyone promote it other than me. More here: http://tfideas.blogspot.com/2010/01/instead-of-bank-tax-why-not.html. Any bailouts paid will have interest rates that accelerate over time - a bank caught up in a storm, but no worse than its peers, would hopefully be able to recover reasonably fast, and so while they'd pay for their bailout, they'd recover. The really irresponsible entities would eventually default on their bailout interest and die, but it would happen with lots of prior warning and after everyone's recovered a bit.

4) I like that flash trading was banned because it was unethical. If flash trading is gone, dark pools aren't nearly as necessary and should probably go, as well.

5) I think the short-sale uptick rule should come back.

6) I think mark-to-market rules need reforming. People don't like the idea that mark-to-market can accelerate a liquidity crunch, but it can. In many cases, market is not representative of actual long term value (esp in a liquidity crunch, because long term value and spot value diverge). Accurate representation of value is good, but mark-to-market isn't necessarily that much better at that than historical.

7) I'd love to see some way of reducing high-frequency trading. High frequency trading has little merit (expounded here: http://tfideas.blogspot.com/2009/11/problem-with-high-frequency-trading.html).  I've mentioned a "high-turnover tax" before, and I still think it will be a good both in reducing HFT and also making Wall Street as a whole a little less schizophrenic. http://tfideas.blogspot.com/2010/01/how-to-deal-with-high-frequency-trading.html. Again, I haven't heard anybody promoting this. Another idea I have heard elsewhere is to tax cancelled transactions - UHFT are the biggest creator of cancelled trades, by far, and this could deter them.

8) Putting things on exchanges if possible is not the panacea everyone thinks but it is a useful step in cases. It will certainly reduce the opacity of the system somewhat, which is a good thing. It also alleviates, somewhat, the "too interconnected to fail" problem. This is not the case for more specialized derivatives, but for some standard ones, it is a big improvement - you just need to worry about the exchange's viability instead of everyone's individually.

9) Similarly, forcing all contracts to be on-balance sheet would be helpful. This would entail increasing what actually gets counted in capital regulations. In a sense, there is a "shadow banking" system which should be brought into the open.

10) Reduce the role of the ratings agencies in official calculations. The ratings agencies are being demonized by everyone, and probably unfairly - they did a bad job, but according to AK Barnett-Hart's masterful thesis, there doesn't seem to be a statistical link between who pays the agencies and how everyone is rated, implying that outright corruption wasn't an issue. Additionally, there are places where their ratings don't have a great substitute. All that said, relying on two or three (largely young) analysts for the valuation of an entire class of securities is probably not so smart. Ratings should be a piece, but not all, of how these things are regulated.

11) Subsidized and lenient mortgages need to go. Down payments on Federally-subsidized mortgages need to go up. If the government is paying for you, you'd better be responsible.

12) No way this will happen, but a Fed and an SEC that are more independent of the whims of Congress would help avoid the "Phil Gramm/Barney Frank" effect, where stupid congressmen assert that whistleblowers are wrong because they're enjoying the party (Phil Gramm famously proclaimed that the Fed call for higher rates was ridiculous because tech stocks were still undervalued in February 2000 and higher rates would hurt urban teen employment, and Barney Frank argued that for affordable housing, it'd be fine to "roll the dice" a little bit with subprime mortgages at Fannie and Freddie.)

13) I'd like to see corporate board reform tacked onto the bill. It's nominally related in the public eye, even if it's different from a true economic or legal sense. This would entail a lot of what is listed here: http://tfideas.blogspot.com/2010/04/fixing-executive-compensation.html and here: http://tfideas.blogspot.com/2010/03/perfect-government-in-trevors-world.html. The biggest factor is to force the board to think like owners instead of like the CEO's buddies. The CEO should not be allowed to be the board's chairman, the boardmembers should probably not be paid and they certainly should be forced to have large personal positions in the company. Proxy voting (anyone who doesnt vote is voted by the board) should not be legal. Elections should be frequent, with publicity for board challengers, and compensation formulae should be a) public and b) very long-term. If this means paying the CEO after he or she is gone, fine, but the CEO should get paid based on the results of that year's accomplishments over the next few years, instead of just a snapshot-in-time approach. We might consider punishment for litigation (probably excluding patent litigation, which is more benign).

14) Mandatory waiting periods for consumer lending. 7 days to back out, for example, would be extremely useful in reducing rash decisions and would force people to take time to look at documents and understand them.

15) Some sort of mechanism for getting banks to hold onto junior tranches of deals they originate. This is not to loan against, but to incentivize more responsible underwriting. This would have to be done carefully to minimize distortion of deals to create high reward, low probability option value.

16) A close look at annual fees vs mutual fund loads. Reducing annual fees in exchange for increasing back-end loads to mutual funds and hedge funds would prevent "churn" and chasing performance, which have the effect of driving people into strategies just as they begin to overheat, worsening bubbles. A "lockup" period once you enter a fund would be an inferior version of this, but increasing the cost of churn without increasing the cost of funds - as would happen if annual fees were partially replaced with back-end loads - would be particularly useful. I don't know how this would be implemented but I imagine it would not be too difficult. This is a third piece of this list that I haven't heard anyone promoting.

17) More radically, I'd be interested to see what would happen if we re-widened spreads from pennies to nickels, or partially ECN-ed the market, where matching and execution are not continuous but instead execution happens at certain predetermined points (perhaps every minute, or 5 minutes, or something like that) and only matching can happen in the interim. These would help to reduce a lot of the egregious HFT and flash trading practices.
18) Making the market circuit breakers time-dependent. Clearly, a 10% drop in 10 minutes is more serious than a 10% drop in a day - see May 6, 2010.

19) Coordinating the networks so when a primary exchange closes down or a circuit breaker is tripped, the secondary exchanges close, as well. This should apply both to closings due to volatility and also to coordinating market opens and closes, because executions on a secondary exchange before the NYSE opens can often disadvantage retail investors in a way that is not fair.

What I don't want to see:

1) I think "too big to fail" is stupid and potentially harmful, and the idea of capping bank size or splitting big banks will worsen things, not improve them. here's why: http://tfideas.blogspot.com/2009/11/too-big-to-fail-stupid-concept.html and especially this: http://tfideas.blogspot.com/2010/04/more-on-financial-reform.html
more here: http://tfideas.blogspot.com/2010/01/prop-trading.html <- more about bank size than prop trading
and the followup here: http://tfideas.blogspot.com/2010/02/on-volcker-rule-about-prop-trading-and.html

2) Clearly, there needs to be some look at capital regulations, but tighter capital regulations will be very hard to implement well, because they often have some sort of unexpected negative consequence (Basel II capital regulations reduced the number of banks that took junior equity slices of the products they created because it made them not count in the capital structure, even though the bank having a junior piece made the actual structuring of the products more risk-averse). Reducing the opacity of the system and increasing what actually needs to get counted in capital regulations is not a bad start.

3) A pure transactions tax or a bank tax seems to have perverse effects and could even work to worsen crises. More here: http://tfideas.blogspot.com/2010/01/instead-of-bank-tax-why-not.html (also linked above) and here: http://tfideas.blogspot.com/2010/01/how-to-deal-with-high-frequency-trading.html (also linked above)

4) I am very torn on the Consumer Financial Protection Agency, because I see it as a blunt instrument of Congressional political games.

My main concern is that if, for example, Congress decides it wants student loans offered for lower rates, student loans will get offered for lower rates regardless of whether that's a sensible level to offer them at. When they start defaulting badly, the taxpayer ends up paying. It thus becomes a mechanism of shadow redistribution at the whims of whoever is in control of Congress.

I don't believe consumers are quite the "victims" everyone says they are - firstly, people are on some level responsible for their own financial choices, and while I can see getting caught up in the moment with things like credit cards, if we can't hold people responsible for purchasing a house with a degree of intelligence and foresight as to their own ability to pay, then we're functionally arguing for no responsibility for anyone. There's also the issue that the consumers took the banks for a ride this time, not the other way around - banks lent consumers all the money required to buy a house, then consumers couldn't pay it off and left the bank with a depreciating asset. The banks lost the entire value of the house; consumers just lost a credit rating that'll go back to normal in 7 years.

That said, there are places where consumer financial protection is valuable. Credit card companies used to target kids, for example, and charge interest rates bordering on usury.

On the whole, I think I'd rather see that kind of power in a much less centralized, Congress-dependent entity.

5) I don't want to see Congress try and regulate what sort of derivatives can and cannot exist or be held by various different entities. China illegalizes shorts and foreign investing, so they have an asset bubble in stocks. Illegalize certain actors from doing things and at best you'll increase the cost of capital, and at worst you'll create a massive, massive imbalance. This, sadly, seems to be a major part of the Democratic approach. Financial innovation has value, despite uneducated media complaints about "zero-sum transactions". see here: http://tfideas.blogspot.com/2010/02/financial-innovation.html. For a quick example, think about an airline and a refinery. Both are highly capital intensive and highly sensitive to price movements of gasoline, but airlines do badly if gasoline goes up and refineries (usually) do better. The downside risk for refineries and airlines (potential bankruptcy) is a lot worse than the upside risk (a little extra profit that'll get competed away quickly enough) is good. Thus, if they engage in a hedge, where refineries pay airlines if gas prices go up and the converse if they go down, they're both better off. You have a zero-sum financial transaction with a substantial economic benefit. Thus, derivatives aren't riskier than not having derivatives, it's just that the types of risk change (from gas risk to counterparty risk) into something much more difficult to measure. If you overestimate how much risk you've reduced, that can cause a crisis because you take on too much risk elsewhere. That doesn't, however, mean that derivatives should go away just cuz they're harder to measure.

No comments:

Post a Comment