This site has moved to
The posts below are backup copies from the new site.

July 18, 2009

Economist's View - 3 new articles

"Financial Invention vs. Consumer Protection"

Robert Shiller says that while a Consumer Financial Protection Agency is a good idea, it wouldn't have prevented the housing bubble:

Financial Invention vs. Consumer Protection, by Robert J. Shiller, Commentary, NY Times: James Watt, who invented the first practical steam engine in 1765, worried that high-pressure steam could lead to major explosions. So he avoided high pressure and ended up with an inefficient engine. It wasn't until 1799 that Richard Trevithick ... created a high-pressure engine that opened a new age of steam-powered factories, railways and ships.

That is how innovation often proceeds — by learning from errors and hazards and gradually conquering problems through devices of increasing complexity...

Our financial system has essentially exploded... We need to invent our way out..., and, eventually, we will. That invention will proceed mostly in the private sector. Yet government must play a role, because civil society demands that people's lives and welfare be ... protected from overzealous innovators who might disregard public safety and take improper advantage of nascent technology.

The Obama administration has proposed a ... Consumer Financial Protection Agency, which would be charged with safeguarding consumers against things like abusive mortgage, auto loan or credit card contracts. The new agency is to encourage "plain vanilla" products that are simpler and easier to understand. But representatives of the financial services industry have criticized the proposal as a threat to innovation...

If a consumer agency had been set up 20 years ago, would the subprime mortgage crisis have been prevented? We don't know, but it seems improbable. Such an agency would most likely have slowed some abusive practices... That ... would have reduced the severity of the crisis, and that is no small thing.

On the other hand, unless these regulators were extremely vanilla in approach and just said no to any innovation, or unless they had an unusually deep understanding of speculative bubbles, I think they would have allowed most of those subprime mortgages. And they probably wouldn't have had the detailed knowledge they would have needed to halt the decline of lending standards on prime mortgages in a timely way. In all likelihood, we would still be in this financial crisis.

In short, the new agency seems a good idea, and, if it is created, it should ... support innovation and staffed by people who know finance..., including some who appreciate that human behavior must be understood and factored into financial design.

But that leaves us with the deeper quandary: Our society needs financial innovation, and still seems vulnerable to ... speculative bubbles that create truly big problems. Even if they can be mitigated, periodic crises may not be preventable, at least not by banning abusive credit cards or even by throwing the bad guys in jail. ...

The effectiveness of our free enterprise system depends on allowing business people to manage the myriad risks — including the risk of asset bubbles — that impinge on their operations in the long term. And this process needs constant change and improvement.

Complexity is not in itself a bad thing. ... A laptop computer is an immensely complex instrument... Yet it can be designed well so that it seems plain vanilla to the ultimate user.

And as for steam engines, the modern turbine high-pressure versions are not plain vanilla in any sense. They are sophisticated triumphs of engineering. They help generate most of our electric power with very few accidents.

I'm not sure his example works. If a modern turbine engine fails, it doesn't threaten the broader economy. If the engines were interconnected, so interconnected that the failure of one could bring them all down (beyond a single set at a given geographical location), then they might threaten the entire economy and be more like financial innovation.

The point is, because the costs of a steam or turbine engine blowing up are mostly localized, we can allow innovation to occur with very little regulation within the private sector without too much concern. Of course, we need to make sure that, say, a steam engine blowing up in a garage doesn't harm the neighbors, or harm any employees who might be there, and we also want to protect the inventors from themselves to some extent, but since the threat from an explosion is localized, we can allow innovation to proceed in the private sector under relatively light regulation without incurring great risks.

Suppose, however, that the turbine engines were interconnected and the failure of a single engine anywhere in the system could bring the whole system down, and not just for a day or two, but for months and months, and that the loss of so much power for so long would wreck the economy. In such a case, how much trust would you be willing to place in an unregulated private sector development of a new engine type for the grid? How much complexity would you be comfortable with? How much testing would you want the engines to undergo before being allowed in use? Would the fact that they have "very few accidents" as Shiller notes be of comfort?

When the dangers are great, we need to be careful. The financial grid is interconnected in just this way, and we need to do all that we can to ensure that new innovations do not become engines of destruction yet again.

"Let The Good Times Roll Again?"

The question of how to interpret Goldman Sachs' unexpectedly large earnings has been addressed here and here (with an addendum at the bottom of the second post). One lesson appears to be that the bad incentives underlying executive compensation are still be present. However, exactly why that is the case - other than the sheer size of the profits - has not been fully addressed. This argument from Lucian Bebchuk takes a step in that direction by noting that the Goldman Sachs' bonuses are a reward for short-term gains, when we ought to be tying compensation to long-run performance. Thus, the problematic incentive to take on large amounts of risk in pursuit of immediate profits is still present. (Though it's not included below, he also argues that compensation should be based upon the long-run value of a broad basket of securities, not just common shares as has been typical in the past, to avoid distorting incentives in a way that gives executives another reason to take on too much risk):

Let the good times roll again?, by Lucian Bebchuk, Commentary, Project Syndicate: Goldman Sachs announced this month plans to provide bonuses at record levels, and there are widespread expectations that bonuses and pay in many other firms will rise substantially this year. Should the good times start rolling again so soon?

Not without reform. Indeed, one key lesson of the financial crisis is that an overhaul of executive compensation must be high on the policy agenda.

Indeed, pay arrangements were a major contributing factor to the excessive risk-taking by financial institutions that helped bring about the financial crisis. By rewarding executives for risky behaviour, and by insulating them from some of the adverse consequences of that behaviour, pay arrangements for financial-sector bosses produced perverse incentives, encouraging them to gamble.

One major factor that induced excessive risk-taking is that firms' standard pay arrangements reward executives for short-term gains, even when those gains are subsequently reversed. Although the financial sector lost more than half of its stock-market value during the last five years, executives were still able to cash out, prior to the stock market implosion, large amounts of equity compensation and bonus compensation.

Such pay structures gave executives excessive incentives to seek short-term gains ... even when doing so would increase the risks of an implosion later on. ...

Following the crisis, this problem has become widely recognised... But it still needs to be effectively addressed: Goldman's recent decision to provide record bonuses as a reward for performance in the last two quarters, for example, is a step in the wrong direction.

To avoid rewards for short-term performance and focus on long-term results, pay structures need to be re-designed. As far as equity-based compensation is concerned, executives should not be allowed to cash out options and shares given to them for a period of, say, five years after the time of "vesting"...

Similarly, bonus compensation should be redesigned to reward long-term performance. For starters, the use of bonuses based on one-year results should be discouraged. Furthermore, bonuses should not be paid immediately, but rather placed in a company account for several years and adjusted downward if the company subsequently learns that the reason for awarding a bonus no longer holds up. ...

A thorough overhaul of compensation structures must be an important element of the new financial order.

The latest I'm aware of on this topic is draft legislation Barney Frank circulated today:

All publicly traded U.S. companies, and particularly financial firms, would face new executive-compensation requirements under draft legislation circulated Friday by a top lawmaker in the U.S. House of Representatives.

The measure proposed by Rep. Barney Frank, D-Mass., would give shareholders a greater ability to weigh in on compensation packages, require compensation consultants to satisfy independence criteria established by the government and would ensure that compensation committees are made up of independent directors. ...

Banks, broker-dealers, investment advisers and all other financial firms would be required to disclose any incentive-based compensation structures, while federal regulators would be able to limit "inappropriate or imprudently risky" pay practices.

The panel is expected to take up some form of the legislation next week.

This legislation would give regulators the power to limit executive pay structures that they believe are excessively risky, but exactly what types of compensation structures would be allowed or required under this legislation is a bit vague. It doesn't sound like it will be as restrictive as Lucian Bebchuk would like, but we shall see.

There are two related issues here. First, was the compensation fair in terms of being a reward for productive activities such as directing capital where it was most needed, or distributing and reducing risk? Before the crash, some people tried to make that argument, but for financial executives, it seems clear that the compensation was based upon bubble rather than fundamental values, and that this lead to inflated rewards relative to the value that was actually created. But even for non-financial executives, it's hard to believe that the generous compensation high-level managers and executives received in recent years is due entirely to their superior productivity. The "say on pay" part of the proposed legislation tries to give shareholders a greater voice in setting compensation levels, and is directed at this problem.

The second issue is whether the compensation structures that have been used in the past lead to incentives to accumulate more risk than is optimal. It appears that they did, and the second part of the proposal is an attempt to overcome this problem by giving regulators the ability to limit systemically risky practices.

There are two types of excessive risk to be worried about, but only one is a systemic risk, i.e. a risk to the overall economy. First, a particular executive compensation structure may give an executive the incentive to take on too much risk - more than shareholders desire - but there is no danger to the overall financial system. This type of excessive risk taking needs to be prevented, but it is not a danger to the overall economy, and regulators are generally concerned with microeconomics questions concerning optimal incentive structures. Second, there is excessive risk that endangers the entire financial system and the broader economy. This is systemic risk that is the target of the proposed legislation, the kind of risk regulators such as the Fed are concerned with, and the kind we must do our best to eliminate if we want to avoid a repeat of the present crisis.

To me, both elements of the proposed legislation - the part that gives shareholders the power to limit compensation levels and the part that gives regulators the power to limit risky compensation schemes - seem vague and rather weak. I hope that changes as the legislation develops.

links for 2009-07-18

View comments

No comments: