06 December 2007
Paul Myners on Moral Hazard in banking
Following my recent entry on the moral hazard in banking, I was very interested to read Paul Myners comments on a similar theme. These comments were made in a substantial interview reported on ft.com 4th December 2007.
The full interview can be found at (subscription maybe required):
http://www.ft.com/cms/s/0/7272fc1c-a28b-11dc-81c4-0000779fd2ac.html
Specifically on the moral hazard point Mr Myners said:
"I think there’s an inbuilt moral hazard in banking. I think that isn’t just confined to banking. I think it includes private equity and other forms of investment in which the incentives to take on risk in pursuit of reward are not symmetrical to the consequences of getting that wrong. So there’s a natural inclination to stay with risk for too long. Chuck Prince embraced this in his famous statement about the music’s still playing, but I think that’s true to some extent to equity investors as well. While the market is rising, it is better to stay in than to seek to anticipate a fall, get that wrong for a short period of time, underperform against an index and benchmark and other managers, and run the risk of losing the account. So, the economic rationality that should lie behind equity and other forms of investment, doesn’t always work in the way that the economist assumes because the economist has not factored in the agency risk for the agent underperforming against expectation."
It seems to me that the problem that Mr Myners describes is not so much "inbuilt" but rather a direct consequence of the way fund managers are paid and of performance being measured over too short a timescale.
If I had my own money invested in a market that was still rising, but expected to fall significantly at some point in the next year, then I would clearly be looking to sell [Interestingly I did move significant personal money out of equities in June 2007!].
Surely then if I employ a fund manager to look after my interests then the fund manager should also be looking to sell my assets for me in this scenario. But Mr Myners points out that this strategy does not work for the fund manager because, if the market continues to rise for several months then the funds performance will look poor over this period, and the fund manager will lose his bonuses.
[I do worry that American stock values are currently propped up by fund managers looking at each others behaviour and all desperately hoping that they can get through to year end without a big fall. Should we expect a big fall early in the new year?]
It seems top me that there are two possible ways round this problem:
1) Stop all performance related pay for fund managers. They could be paid a flat salary and encouraged to take seriously their fiduciary duty to the underlying owners of the assets.
or
2) Require all performance related pay for fund managers to be linked to long term (five years?) performance measures, using a structure similar to the FILLIP described in Performance and Reward.
Obviously it is important that the people who hire fund managers also restrict their attention to long term performance. If they move their money too often they undermine the long term thinking of the fund manager. Mr Myners comments about developing the expertise of Pension Fund Trustees are important in this respect. However it is important that any "professionalism" in pension fund trustees stops short of giving them a vested interest in outcomes, or does anything to undermine their duties to the underlying owners of the funds.
The full interview can be found at (subscription maybe required):
http://www.ft.com/cms/s/0/7272fc1c-a28b-11dc-81c4-0000779fd2ac.html
Specifically on the moral hazard point Mr Myners said:
"I think there’s an inbuilt moral hazard in banking. I think that isn’t just confined to banking. I think it includes private equity and other forms of investment in which the incentives to take on risk in pursuit of reward are not symmetrical to the consequences of getting that wrong. So there’s a natural inclination to stay with risk for too long. Chuck Prince embraced this in his famous statement about the music’s still playing, but I think that’s true to some extent to equity investors as well. While the market is rising, it is better to stay in than to seek to anticipate a fall, get that wrong for a short period of time, underperform against an index and benchmark and other managers, and run the risk of losing the account. So, the economic rationality that should lie behind equity and other forms of investment, doesn’t always work in the way that the economist assumes because the economist has not factored in the agency risk for the agent underperforming against expectation."
It seems to me that the problem that Mr Myners describes is not so much "inbuilt" but rather a direct consequence of the way fund managers are paid and of performance being measured over too short a timescale.
If I had my own money invested in a market that was still rising, but expected to fall significantly at some point in the next year, then I would clearly be looking to sell [Interestingly I did move significant personal money out of equities in June 2007!].
Surely then if I employ a fund manager to look after my interests then the fund manager should also be looking to sell my assets for me in this scenario. But Mr Myners points out that this strategy does not work for the fund manager because, if the market continues to rise for several months then the funds performance will look poor over this period, and the fund manager will lose his bonuses.
[I do worry that American stock values are currently propped up by fund managers looking at each others behaviour and all desperately hoping that they can get through to year end without a big fall. Should we expect a big fall early in the new year?]
It seems top me that there are two possible ways round this problem:
1) Stop all performance related pay for fund managers. They could be paid a flat salary and encouraged to take seriously their fiduciary duty to the underlying owners of the assets.
or
2) Require all performance related pay for fund managers to be linked to long term (five years?) performance measures, using a structure similar to the FILLIP described in Performance and Reward.
Obviously it is important that the people who hire fund managers also restrict their attention to long term performance. If they move their money too often they undermine the long term thinking of the fund manager. Mr Myners comments about developing the expertise of Pension Fund Trustees are important in this respect. However it is important that any "professionalism" in pension fund trustees stops short of giving them a vested interest in outcomes, or does anything to undermine their duties to the underlying owners of the funds.
Labels: fund management, incentives, moral hazard, Myner
12 November 2007
Moral hazard, bank incentives and the credit crisis
In August, Mervyn King, Governor of the Bank of of England, made an important point about "Moral Hazard" which has not been taken nearly seriously enough. His point was that if the Bank of England helps banks that have taken imprudent risks then it encourages banks to take such risks in the future. Banks (perhaps imprudent banks?) complained that his stance was out of line with Europe and America and impractical. But now that the Bank of England has pragmatically made credit available there is a danger than Mervyn King's important point is forgotten about and lost.
The moral hazard problem on bank's behaviour stems from an even more important moral hazard problem in the way that bankers are paid. It is far too easy for bankers to get extremely rich on annual bonuses and other short term rewards, before the risks that they have taken properly come home to roost. When the risks finally do come home it is someone else (the banks shareholders, or public financial regulators) who carry the cost. Bonuses are not retrospectively deducted.
The problem is further exacerbated by the way that bankers are appointed or dismissed. Promotion decisions are influenced far more by personal performance in the last 18 months (before the risks taken have come to maturity) than on long term performance. Then when the risks go wrong the top man might lose his job, but he is paid handsomely for it! The message to potential future executives is clear - short termism pays both on the way up and on the way down.
To avoid moral hazard we need to take long term performance far more seriously than short term performance. In this context I was delighted to read Patrick Hosking's article in The Times, (10th November 2007, page 61, "Time to reform the way bankers are paid").
He is absolutely spot on to say that the way that pay works in the big investment banks is at the very heart of the current credit crisis. The incentives for senior bankers work over a time frame that is far too short.
As Patrick Hosking quite rightly says, shareholders should be demanding better practice on pay. But this is so obvious that I think we have to ask why it has not already happened.
I believe one reason is that shareholders are represented by fund managers and institutional shareholders who typically have exactly the same incentive problem in their own pay. They receive big incentives based on one year cycles, which are not aligned with the interests of the underlying owners of the funds. This is most clear in the case of hedge funds. A hedge fund typically charges very high fees for a good year of investment performance, but does not pay money back if the value of the fund falls. A investment strategy that makes good returns four years out of five, but which is occasionally disastrous, therefore works very well for the hedge fund manager, but not for the underlying owner of the fund.
We should be just as concerned about the way that fund managers are paid.
The book "Performance and Reward" (see link "View the book" in left hand column) suggests solutions to these problems. First of all, it sets out a form of executive pay called a FILLIP. A FILLIP depends on a far more rigorous alignment of pay with shareholders interests, such that performance for top executives is only about long term growth in shareholder value. It also, as Patrick Hosking's article suggests, holds performance related pay in a manner similar to an escrow account for, say, five years. This ensures that value lost latter in the risk cycle is properly reflected in the performance pay.
The moral hazard problem on bank's behaviour stems from an even more important moral hazard problem in the way that bankers are paid. It is far too easy for bankers to get extremely rich on annual bonuses and other short term rewards, before the risks that they have taken properly come home to roost. When the risks finally do come home it is someone else (the banks shareholders, or public financial regulators) who carry the cost. Bonuses are not retrospectively deducted.
The problem is further exacerbated by the way that bankers are appointed or dismissed. Promotion decisions are influenced far more by personal performance in the last 18 months (before the risks taken have come to maturity) than on long term performance. Then when the risks go wrong the top man might lose his job, but he is paid handsomely for it! The message to potential future executives is clear - short termism pays both on the way up and on the way down.
To avoid moral hazard we need to take long term performance far more seriously than short term performance. In this context I was delighted to read Patrick Hosking's article in The Times, (10th November 2007, page 61, "Time to reform the way bankers are paid").
He is absolutely spot on to say that the way that pay works in the big investment banks is at the very heart of the current credit crisis. The incentives for senior bankers work over a time frame that is far too short.
As Patrick Hosking quite rightly says, shareholders should be demanding better practice on pay. But this is so obvious that I think we have to ask why it has not already happened.
I believe one reason is that shareholders are represented by fund managers and institutional shareholders who typically have exactly the same incentive problem in their own pay. They receive big incentives based on one year cycles, which are not aligned with the interests of the underlying owners of the funds. This is most clear in the case of hedge funds. A hedge fund typically charges very high fees for a good year of investment performance, but does not pay money back if the value of the fund falls. A investment strategy that makes good returns four years out of five, but which is occasionally disastrous, therefore works very well for the hedge fund manager, but not for the underlying owner of the fund.
We should be just as concerned about the way that fund managers are paid.
The book "Performance and Reward" (see link "View the book" in left hand column) suggests solutions to these problems. First of all, it sets out a form of executive pay called a FILLIP. A FILLIP depends on a far more rigorous alignment of pay with shareholders interests, such that performance for top executives is only about long term growth in shareholder value. It also, as Patrick Hosking's article suggests, holds performance related pay in a manner similar to an escrow account for, say, five years. This ensures that value lost latter in the risk cycle is properly reflected in the performance pay.
Labels: Bank of England, long term, Mervyn King, moral hazard, Patrick Hosking, short term
