Saturday, March 14, 2009

A coherent framework for Regulation

This financial crisis thing has been going on for quite a while now.

Many proposals have been made to solve the problems facing the financial system. Most of them have failed. This is a good thing, as the mainstream proposals were and remain uniformly poor.

Basically, most proposals have been designed to bring "bad" assets off bank balance sheets so that somebody else (first: other banks, now: the taxpayer) bears the cost of paying ridiculous compensation packages to bankers who made paper, ponzi profits by taking on risk without limit.

Of course we have to recapitalize the financial system. To not do so would be negligent. However, banks should offer something in return for getting using public wealth, and regulators should do something to impede this sort of thing from happening again (remember: savings and loan crisis happened in the '80's, and regulation was continuously eroded throughout the time since).

This should be the first point, obvious, but not respected in recent years:
> The regulator must be strong, independent and willing to regulate.
Throughout the western world, regulator positions have been filled by senior bankers. It doesn't work. If somebody has spent his life living for one system, he is unlikely to seriously criticize the system, especially as his social network will largely be among the regulated. Self regulation is also largely ineffective, for obvious reasons. Perversely, in places like the UK, the Financial Services Regulator was apparently pressured to adopt a "light touch" i.e. not to intervene in the free market! It is apparent that regulators need to take a step back and do some serious thinking about what their job actually is, and should be.


> If entities are systematically fundamental, they should not be private. If private entities are "too large to fail" they must be made smaller or nationalized.
In a system where companies are compensated from taking on too much risk, and know that the taxpayer will bail them out should the risk go bad, then the system bears a large incentive to abuse that position, and take on more risk than an entity would take on would they not have an implicit public insurance. The obvious answer to this appears clear: BREAK UP THE BANKS! Private risk taking, markets are great at allocating resources quite efficiently. They brought us everything from the Nintendo Wii to cruise ships to ... Therefore, any even semi capitalist system must retain a system where private risks can be taken and compensated accordingly. I think the area of private risk is currently adequately regulated: Investors can invest in stocks, bonds, alternative investment vehicles. If they make the wrong choices or trust the wrong people, the lose, and inversely, if they make the right choices, they win. The most important element is that the entities can lose and that bad investment choices are penalized. In a coherent framework for markets, the most important element is that there are no bailout, nor any need for bailouts. Private, non-state insured risk taking should therefore be encouraged.
Conversely, entities which hold insured deposits should not be allowed to take on excessive market risk. It appears obvious, but this is not the way things are being handled today. Large banks have taken on very large risks (RBS-ABN...) and when these risks have gone foul they have asked the state for help. Many have argued that the issue is with short term executive compensation. This may have contributed to excessive risk taking, but impeding this appears lees direct and more difficult to manage than impeding banks from taking on excessive risk in the first place.
What is excessive risk, and how to stop it? The easiest, naive way to stop excessive risk taking is with so called "capital adequacy ratios" whereby regulated banks are not able to take on more leverage/gearing than a particular amount. As a guideline, this should be closer to 3x than 30x. It should appear obvious, but apparently nothing is. Along the same lines, off balance sheet accounting which serves to hide risk from shareholders and regulators should be forbidden and recognized in law for the fraud that it is. Accounting rules should generally be revised to close loopholes and increase transparency.
Banks, and other risk takers should be given a choice: do they want to be systemically important small business lenders or do they want to be free to take on risk? I envisage a system whereby there is a regulated financial sector and an unregulated financial sector. In the regulated sector entities will have access to central bank liquidity and state capital guarantees. They will take deposits and may make loans to businesses at "market" rates, or invest in certain free market opportunities albeit in a highly regulated manner. Arguably these entities should be owned by the state due to their systemically fundamental nature and the strictly regulated way of operating. Another reason why these entities should not be freely owned is that they should not be subject to pressures to make short term profit by shareholders maximizing their short term wealth accumulation at the expense of future crisis. Basically, it would be difficult to make these regulated banks competitive with the free banks and therefore it may be best to exempt them from competition. It may be a good idea to have an artificial "market" whereby regulated bank executives have some of their compensation tied to their effectiveness in "competing" with the other regulated banks. This would however unlikely reach levels of star traders etc. in the free sector, except for cases of exceptional effectiveness or cost cutting in the regulated sector.

On the other hand, you will have the "free" unregulated financial system. "Banks" operating in this space will not be subject to regulation, will not have access to central bank liquidity and deposits will not be guaranteed by the state. It is very likely that deposits in the free sector will pay significantly higher interest thank insured deposits. This is desirable, as interest will reflect the higher risk taken on by free entities. Conversely, companies will face the choice of borrowing from the regulated or the non-regulated sector. This will be achieved by market mechanisms. Enterprises deemed to be less risky will be able to borrow from the regulated sector at lesser risk premia than risky enterprises forced to borrow from the free sector. Although the regulated sector will be compelled to take on some risk, this will be negligible compared to the risk taken on by the free sector.
The free sector will basically be characterized by funds (which can be called banks or investment vehicles or whatever). These will engage in lending, structuring, borrowing, trading, brokerage etc. These entities will be in competition with each other, will take on risk and compensate accountholders/investors according to their abilities in this highly competitive investment marketplace. The only role of the regulator in this market is one of atomization or the prevention of excessive consolidation. To impede the clout risk if entities becoming systemically important or "too large to fail" the regulator must break up entities or stop consolidation in the "free" financial sector. Although this is an obvious contradiction to the term "free", it appears the only way to impede the risky sector from consolidating risk and endangering public welfare. This also implies limits of financial technology - with the knowledge we have today, it appears clear that diversification is useful only on a single portfolio and not on a systemic level. Excesses such as "bond insurance" to generate cheap borrowing will not be attractive if insurance is sold at attractive premium. Rating agencies should be disenfranchised, or given some liability connected to ratings they give.

Implications to be noted:
Cheap liquidity caused the financial crisis. Private banks had such easy access to Central Bank liquidity that everybody with a business card could get "leverage" on their "diversified" investment products. This distorted risk premia, and basically rewarded short sightedness and led to the entire economy resembling a pyramid scheme. Should my suggestions be implemented, risk premia would no longer be distorted by excessively cheap liquidity, which in my eyes was the main factor contributing to the current malinvestment crisis.
This will make monetary policy less effective. This is a good thing, as recent history has shown that Central Bankers do not possess the foresight necessary to wield a tool as powerful pas determining financial system liquidity for any particular currency. This is an area where the market should be given more space, so that risk premia once again reflect risks taken.

What should be done right now:
Bankers want their firms to remain private so that they can continue to benefit from their privileged position in the financial system to syphon off public wealth for themselves. This is unacceptable.
The financial system is bust. For recapitalization to be durable, the financial system should once again reflect true values, and prices should not be determined in a political process (lest the financial sector mutate into an industry with soviet incentives with corruption, inefficiency and waste on an even higher scale than today). To achieve this, the entire system should be put into administration, and all liabilities laid bare to the regulator. Bankrupt firms should be identified, their equity and debt holders wiped out.
In the short term, operations could be consolidated into a mega bank which would then start organizing itself into free and regulated operations. Free operations would be sold off in small tranches to ensure competition and some upside for the taxpayer. Regulated operations organized as elaborated previously. Financial institutions which are not bust should generally be left alone, subject to the constraint that they would be required to remain smaller than "too large to fail". Certain deposit holding entities may elect to consolidate in the regulated sector.

That's about it for a short summary. I welcome debate on these issues, and look forward to any feedback anybody may have.