Lecture on executive pay and the crisis

[Sheilbh]

Recommended from Buttonwood's blog at the Economist:

Reforming banks
The wrong numbers

Oct 25th 2011, 10:07 by Buttonwood

ANDREW Haldane, the executive director for financial stability of the Bank of England, has given many excellent speeches but his latest effort is a must-read for anyone who wants to know how we got into this current financial mess.

The speech deals with "the flaw" that so confounded Alan Greenspan, how private sector investors failed to control risks in the banking sector. The problem was not that bank executives had no skin in the game; in 2006, the managers with the largest bank stakes were Dick Fuld of Lehman Brothers, Jimmy Cayne at Bear Stearns, Stan O'Neal at Merrill Lynch, John Mack at Morgan Stanley and Angelo Mozilo at Countywide. All lost substantial chunks of their wealth as share prices collapsed.

The fundamental problem is that tax rules (the deductibility of interest) and regulations encouraged banks to gear up their balance sheets. Effective control of the banks, however, rested with the shareholders. As Haldane puts it

    Ownership and control rights are vested in agents comprising less than 5% of the balance sheet.

Thanks to limited liability, the losses of these shareholders are also constrained. In the early 19th century, liability was unlimited, prompting banks to run with much more conservative balance sheets. But that was deemed to deprive industry of much-needed capital so limited liability was brought in.

That transferred the responsibility for monitoring bank managers from shareholders to depositors. Haldane points out that, in the 19th century

    Depositor flight and bank runs came thick and fast, operating as an effective disciplining device on managers and shareholders

The Great Depression illustrated the economic damage that could be caused by widespread bank runs, so deposit insurance was brought in. With the liability of shareholders and depositors now limited, the disciplinary role fell on the holders of debt. But they proved hopeless in the task. In the run-up to the crisis, Haldane points out that

    Credit default swap premia for all banks fell dramatically between 2002 and 2007, on average by around three-quarters. Market perceptions of risk were falling at precisely the time risk in the system was building.

The problem seems to be that while, in theory, creditors would bear the pain if banks collapsed, in practice creditors doubted that they would.

    Having debtors assume pain on paper is fine in practice. But crisis wars are not fought on paper. And if debtors recognise that risks in contracts will not be enforced, they will no longer have incentives to price risk and exercise discipline themselves. So it has been for well over a century.

So that left bank executives to their own devices. As is well known, they were incentivised by share options, a process that in theory aligned their interests with equityholders. Again, this did not work well in practice. Those investors who bought bank shares in the early 1990s have lost money in real terms. But investors are not long-term holders any more. The average holding period for US and UK bank shares fell from 3 years in 1998 to around three months by 2008.

Those short-term investors were hoping to ride the ups and downs of the market, and thus welcome volatility. They thus allowed a system to develop where bank executives used return on equity as their target measure. The easiest way to increase return on equity is to take on more debt; you have more capital to pursue profitable opportunities for the same amount of equity.

That brings in the killer statistics of Haldane's speech.

    Imagine that in 1990 bank CEO pay had been indexed to bank ROE. By 2007, CEO compensation would have reached $26 million. That is precisely in line with their actual payouts. If you believed ROE were a reliable performance metric, US bank CEOs would have had a watertight defence back in 2007.

Instead, of course, we had ludicrously leveraged banks that were too big to fail and brought the economy down with them.

But there is an alternative measure, return on assets (ROA), which allows for both debt and equity. As Haldane notes

    Imagine if the CEOs of the seven largest US banks had in 1989 agreed to index their salaries not to ROE, but to ROA. By 2007, their compensation would not have grown tenfold. Instead, it would have risen from $2.8 million to $3.4 million. Rather than rising to 500 times median household income, it would have fallen to around 68 times.

In other words, if we had used the right risk measure, the worst of the recent mess might have been avoided, and bankers would not have grown so obscenely rich.

On that note, I am off to New York for the Buttonwood conference. Hope to report back on Friday.

http://www.bankofengland.co.uk/publications/speeches/2011/speech525.pdf

There's lots more of interest in the speech (including charts!), but I'm still chundling through it to be honest.  But it does seem very interesting.

[grumbler]

Interesting assessment, and it does answer a lot of my questions about how executives could have acted so irresponsibly.

It seems like it would take a regulatory regime to enforce responsibility, since the market has failed to do so (and, indeed, if Haldane is correct, could not do so).

But I wonder how he accounts for the fact that investment banking essentially merged with retail banking, at least in the US (as I read the history, anyway).  That, in my mind, was a key contributor to the crisis, since investment bakers were then able to gamble with insured money.

[Admiral Yi]

He omits my favorite explanation for the crisis (along with easy money and high leverage): everybody getting the risk on subprimes wrong.

[Sheilbh]

Aren't those more causes of the crisis than explanations?

[Admiral Yi]

Aren't those more causes of the crisis than explanations?

You've lost me.

[DGuller]

That, in my mind, was a key contributor to the crisis, since investment bakers were then able to gamble with insured money.

I don't think you can do that, at least not since the S&L crisis.

[DGuller]

He omits my favorite explanation for the crisis (along with easy money and high leverage): everybody getting the risk on subprimes wrong.

He is 100% correct to do that, though, because subprime crisis was to the financial meltdown what Archduke's assassination was to WW1.  The only people who keep harping on subprime mortgages in these days and age are hopeless Republican hacks.

[Admiral Yi]

 

[DGuller]

And to the extent that subprime mortgages did cause the meltdown, the causes are pretty much the same as the ones already enumerated.  Subprime mortgage risk wasn't mispriced because of stupidity, or because of government mandates ( ).  It was mispriced because all kinds of risk was mispriced or ignored.

[Admiral Yi]

And to the extent that subprime mortgages did cause the meltdown, the causes are pretty much the same as the ones already enumerated.  Subprime mortgage risk wasn't mispriced because of stupidity, or because of government mandates ( ).  It was mispriced because all kinds of risk was mispriced or ignored.

Credit cards?  Car loans?  Small business loans?  US Treasuries?  Muni bonds?  Commercial paper?  Corporate bonds?  Emerging market sovereigns?  Prime mortagages?

[alfred russel]

I think this is flawed for a lot of reasons. Among them, executive comp is usually filled with incentives which are linked to measures relating to equity, but achieving previously unmet incentive targets hasn't led to the explosion in financial executive pay. Boards have simply decided to pay more for executives, presumably because they think that is needed to attract and retain top talent. This is not a phenomena limited to financial companies, and has been a general trend over the last few decades which has produced a lot of literature and commentary. Basically:

Imagine that in 1990 bank CEO pay had been indexed to bank ROE. By 2007, CEO compensation would have reached $26 million. That is precisely in line with their actual payouts. If you believed ROE were a reliable performance metric, US bank CEOs would have had a watertight defence back in 2007.

doesn't make much sense because bank CEOs did not agree in 1990 to index pay to ROE, and there was turnover at most positions, when Boards presumably sought out top candidates not based on an ROE formula but what consultants told them was the "market" rate. And the market rate was largely determined by surveys of what everyone else was doing, and again not a formula involving ROE.

Plus, incentives based on equity tend to make much more sense than for assets. Incentives are set (presumably) by equity owners looking to increase their own profitability. Yes that produces incentives for owners to increase leverage and risk to generate returns, but there are methods to take that into account and the board is also supposed to be overseeing risk management. Equity is by nature a higher risk investment, and equity owners of a firm are not likely to be so concerned with systematic risk their decisions put into the economy if such is the price of increasing their own profit opportunity. That conflict of interest needs to be moderated by regulators.

Also, I didn't read the entire article, or even most of it, but it sounds as though the author is referring to book numbers for equity. If that is true, I think that is somewhat bizarre, as those numbers have very little connection to reality.

[Sheilbh]

You've lost me.

I think DGuller's comparison is pretty apt.  It's like blaming WW1 on the Black Hand.  That may be strictly true but I don't think it really explains anything.

Incentives are set (presumably) by equity owners looking to increase their own profitability. Yes that produces incentives for owners to increase leverage and risk to generate returns, but there are methods to take that into account and the board is also supposed to be overseeing risk management. Equity is by nature a higher risk investment, and equity owners of a firm are not likely to be so concerned with systematic risk their decisions put into the economy if such is the price of increasing their own profit opportunity. That conflict of interest needs to be moderated by regulators.

I think he sort of addresses this - I'm still reading through to be honest:

The story so far. Ownership and control rights for banks are vested in agents comprising less than 5% of the balance sheet. To boost equity returns, there are strong incentives for owners to increase volatility. Those risk-taking flames have been fanned by tax and state aid. As stories go, this one sounds grim.

But this story also contains a puzzle. Long-term shareholders in banks have not obviously reaped the benefits of these distortions. The purchaser of a portfolio of global banking stocks in the early 1990s is today sitting on a real loss. So who exactly is it extracting value from these incentive distortions? The answer is twofold: shorter-term investors and bank management.

It is not difficult to see why shorter-term investors in bank equities stand to gain from volatility. Institutional investors in equities are typically structurally long. They gain and lose symmetrically as returns rise and fall. Many shorter-term investors face no such restrictions. If their timing is right, they can win on both the upswings (when long) and the downswings (when short). For them, the road to riches is a bumpy one – and the bigger the bumps the better. As in Merton's model, all volatility is good volatility.

Perhaps reflecting that, there is evidence of the balance of shareholding having become increasingly short-term over recent years. Chart 8 shows the average duration of holdings of US, UK and European bank shares. Average holding periods for US and UK banks fell from around 3 years in 1998 to around 3 months by 2008. Banking became, quite literally, quarterly capitalism. Today, the average bank is owned by an investor with a time-horizon considerably less than a year.

A second piece of evidence on equity short-termism, which is specific to banks, comes from trends in banks' capital planning. Since the late 1990s, there has been increasing focus on return on equity (ROE) as a performance target. Indeed, major banks have recently set explicit numerical targets for ROE as a guidepost for shareholders and managers. These targets are revealing about investor and managerial incentives.

ROE is an entirely equity-focussed concept. As such, ROE targets provide strong incentives for banks to increase equity returns, either by boosting asset volatility or gearing up their balance sheet. ROE targets hard-wire in the North-Easterly drift evident in Chart 5. Instead of adjusting for risk, ROE is flattered by it. This is fine for short-term investors who thrive on the bumps. But for longer-term investors they are a road to nowhere, as recent experience has shown.

What we have, then, is a set of mutually-reinforcing risk incentives. Investors shorten their horizons. They set ROE targets for management to boost their short-term stake. These targets in turn encourage short-term risk-taking behaviour. That benefits the short-term investor at the expense of the long-term, generating incentives to shorten further horizons. And so the myopia loop continues.

These incentive problems do not stop with owners. Under joint stock banking, ownership and control are divorced. This generates what economists call a principal-agent problem: managers (agents) may not do what owners (principals) wish. In an attempt to solve it, shareholders have sought over recent years to align managerial incentives with their own. That has meant remunerating managers in equity or using equity-based metrics such as ROE.

These trends were all too apparent in the pre-crisis data. The wealth of the average US bank CEO increased by $24 for every $1000 created for shareholders in 2006 (Fahlenbrach and Stulz (2011)). The typical bank CEO pocketed over $1 million for every 1% increase the value of their firm. These are rather potent pecuniary incentives on bank managers to keep shareholders sweet.

These facts also help explain the evolution of large US banks' CEO pay between 1990 and 2007. Imagine that in 1990 bank CEO pay had been indexed to bank ROE. By 2007, CEO compensation would have reached $26 million. That is precisely in line with their actual payouts.

If you believed ROE were a reliable performance metric, US bank CEOs would have had a watertight defence back in 2007. Indeed, it was the L'Oreal defence: because we're worth it. But experience since suggests this performance was cosmetic. ROE flattered returns, and hence compensation, in the upswing.

That is hardly surprising since it puts risk ahead of return and short-term ahead of long-term performance. When the downswing came, the volatility of equity returns sent many banks to the wall. Equity-based incentive contracts helped propel them there. Firm-level evidence could not be clearer. In 2006, the top 5 equity stakes of US bank CEOs ran as follows: Dick Fuld (Lehman Brothers), James Cayne (Bear Stearns), Stan O'Neal (Merrill Lynch), John Mack (Morgan Stanley) and Angelo Mozilo (Countrywide) (Fahlenbrach and Stulz (2011)). We know how these disaster movies ended.

So having been divorced for almost two hundred years, ownership and control have been reunited. But they are an odd couple. Their marriage contract encourages both partners to behave in a volatile, short-term fashion. The marriage itself is destined to last less than a year. In the mid-19th century, unlimited liability was said to be deterring investors of "fortune, intelligence and respectability". For very different reasons, today's governance arrangements might be suffering the same fate.

Worth saying that I think as well as doing research for the BofE (for example he's previously tried to estimate the implicity subsidy received by the banks) he is one of the members of the Financial Policy Committee which, I think, is meant to be our main financial sector regulator.  I think so anyway - I'm not sure what's happening with that but it was certainly the government's intention.

[grumbler]

(snip) Basically:

Imagine that in 1990 bank CEO pay had been indexed to bank ROE. By 2007, CEO compensation would have reached $26 million. That is precisely in line with their actual payouts. If you believed ROE were a reliable performance metric, US bank CEOs would have had a watertight defence back in 2007.

doesn't make much sense because bank CEOs did not agree in 1990 to index pay to ROE...

I don't understand your objection.  A hypothetical is proposed, and the conclusion drawn that, had the hypothetical been true, "US bank CEOs would have had a watertight defence back in 2007."  That makes perfect sense to me.

The problem comes, of course, when equity value and asset and revenue value get out of whack.  That's called a bubble, and when executives get rewarded for creating bubbles, they create them.

[Sheilbh]

I just found the FT Alphaville's description of the lecture here which is better than the Economist blog:
http://ftalphaville.ft.com/blog/2011/10/25/711856/the-history-and-future-of-banking-according-to-andy-haldane/

The history and future of banking, according to Andy Haldane
Posted by John McDermott on Oct 25 21:42.

Andy Haldane's latest speech is a coherent, logically argued history of modern banking that ends with four intriguing policy ideas. The Bank of England's Executive Director, Financial Stability, is always worth reading but his Wincott Annual Memorial Lecture, delivered on Monday evening, is the best introduction to his views on banks.

The FT's Martin Wolf includes a cogent summary of Haldane's proposals as part of his formal response to the speech, highlights of which are available on his blog.

In essence, the speech is a history of the development of what Haldane calls "wonky risk incentives": the market failures that have led to the imbalance between privatised gains and socialised losses. We've tried to summarise the main points below (all quotes from the speech):

1. In the early 1800s, control rights over banks were in the hands of those "whose personal wealth was literally on the line". Most banks were unlimited liability partnerships. Equity capital was often half of banks' liabilities while cash and liquid securities accounted for nearly one-third of assets. "Banking was a low-concentration, low-leverage, high-liquidity" business.

2. But by the second half of the 19th century, "shareholder discipline was proving rather too effective as a brake on risk-taking". There were bridges to build, railways to lay, and countries to invade; parliamentarians lobbied to free bank capital and credit. About the same time, the City of Glasgow bank collapsed and some shareholders were left destitute. There seemed to be too much risk held by capitalists. Thus, the UK (following the US) moved to a limited liability model of bank governance.

3. Unfortunately (there is no blame attached anywhere in Haldane's speeches), it soon became apparent that the calls on reserve capital heightened rather than mitigated panic. Shareholders became too dispersed. "Their upside payoffs remained unlimited, but their downside risks were now capped." In the late 1990s, many banks ended their partnerships and went public, compounding the problem.

4. These governance changes created an asymmetric risk profile. As Robert Merton explained, the equity of an LLC "can be valued as a call option on its assets, with a strike price equal to the value of its liabilities". And "because volatility increases the upside return without affecting the downside risk. If banks seek to maximise shareholder value, they will seek bigger and riskier bets."

5. As well as creating "volatility junkies", the changes encouraged banks to gear-up their balance sheets. "As unlimited liability was phased out, leverage rose from around 3-4 in the middle of the 19th century to around 5-6 at its close. As extended liability was removed, leverage continued its upward trend into the 20th century, by its close reaching over 20. Seven years into the 21st century, at its high-water mark, leverage hit 30 or more."

6. The "largely silent" changes to the relative treatment of debt and equity catalysed the growth of leverage. From the early 20th century debt interest costs became deductible from profits, unlike — at least in most countries — the cost of equity financing. New evidence suggests there's an empirical relationship between this and leverage: "a fall in the tax shield by 10 percentage points would lower the debt-to-asset ratio of a typical firm by 2.8 percentage points."

7. Debt doesn't have the disciplinary effects commonly assumed. Plus, there is a time-consistency problem in crises. "As much as bank management and the authorities may pre-commit to debtor's bearing risk ex-ante, they may be tempted to capitulate ex-post." Creditors realise this, and price and lend accordingly.

8. The time-consistency problem has in turn been made worse by the emergence of too-big-to-fail banks: creditors know that governments have to cover the risk of any one of these failing. "That is doubly unfortunate, as it means debtor discipline will be weakest among institutions for whom society would wish it to be strongest."

9. On top of all this, long-term bank shareholders haven't received the benefits from these flaws — it's all gone to short-term investors and bank management. "The purchaser of a portfolio of global banking stocks in the early 1990s is today sitting on a real loss." Average holding period for US and UK bank stocks fell to 3 months in 2008 from 3 years in 1998. ("Banking became, quite literally, quarterly capitalism.")

10. In the late 1990s return on equity (ROE) became more important as a bank's performance target. This creates a "myopia loop": investors become more short-termist, a focus on ROE encourages short-term risky behaviour, which further incentivises short-term investment. By the eve of the last financial crisis the principal-agent problem was writ large: "In 2006, the top 5 equity stakes of US bank CEOs ran as follows: Dick Fuld (Lehman Brothers), James Cayne (Bear Stearns), Stan O'Neal (Merrill Lynch), John Mack (Morgan Stanley) and Angelo Mozilo (Countrywide) (Fahlenbrach and Stulz (2011)). We know how these disaster movies ended."

Haldane is that rare beast: an economist that also makes specific policy recommendations. Based on the above, he has four sets of ideas, neatly summarised by Wolf (emphasis ours):

The first is to raise equity requirements substantially. There has been some progress in this direction. But it probably does not go far enough. David Miles, a member of the Bank of England's monetary policy committee, with two co-authors, suggests that the optimal capital ratio would be 20 per cent of risk-weighted assets. Such a shift would impose private costs, because of the favourable tax treatment of debt. The answer is to change that tax treatment, by making a normal return on equity tax deductible or, alternatively, withdrawing the tax deductibility of debt.

The second approach is make debt more equity-like. Mr Haldane argues that so-called bail-in debt would prove too costly, because of the damage done by bankruptcy. US experience suggests that this need not be the case. But I agree that the point of bail-in should be determined by market-based measures of capital adequacy. More broadly, bail-in debt can be accepted only if regulators believe they can carry out the conversion into equity, in a crisis. Otherwise, equity ratios should be raised.

A third possible reform is via changes in control rights. It is possible to envisage radical changes to voting rights, for example, that might give some votes to some creditors. Mr Haldane describes this as a hybrid of the mutual and joint-stock models.

A fourth possible reform, suggests Mr Haldane, is to change the target return from one on equity to one on assets. That would certainly have a powerful and attractive impact on incentives.

Wolf adds, as he did via the Independent Commission on Banking, that he'd like the separation of retail and investment banking. This is a good idea but it still seems one step removed from the actual incentives bankers and shareholders are faced with, which is why Haldane's ideas are especially appealing, even if you don't agree with his precise recommendations.

Wolf's point about the omission of shadow banking is an important one. In some ways, Haldane's description of early 1800s banking could refer to the hedge funds and PE funds of today, and it would be interesting if he could incorporate some of the FPC's early ideas into the account above, especially to explain the rise of leverage and maturity mismatch. (Some chat about derivatives wouldn't go amiss either.)

Another thing to consider is the incentives for a group not discussed this time by Haldane: politicians. Presuming you agree with the thrust of Haldane's argument (which you may not) there are still problems with (political) feasibility. Strangely enough, given how unpopular "bankers" (loosely defined) are these days, there's still a reluctance to do anything radical to realign incentives along the lines Haldane suggests.

Any argument about the time-consistency problem should ideally include the one faced by those trying to get elected in a few years time. This means providing answers to, for example, what happens to bank lending, what happens to a source of UK comparative advantage, and so on. For although these ideas may be of huge long-term benefit, most politicians inhabit their own myopic system: the electoral calendar.

Martin Wolf's shorter reply is here:
http://blogs.ft.com/martin-wolf-exchange/2011/10/24/the-threat-of-the-volatility-junkie/#axzz1bhyOVSyd
It's also quite interesting and worth reading.

[alfred russel]

I think he sort of addresses this - I'm still reading through to be honest:

There are a few things there that bother me:

Ownership and control rights for banks are vested in agents comprising less than 5% of the balance sheet.

Ownership and control rights are the equity portion of the balance sheet. There are a few problems with looking at the equity section of the balance sheet, especially for financial firms. A balance sheet has a mathematical relationship: assets = liabilities + equity.

For a financial firm, a deposit is both an asset and a liability. Suppose you give me $100 to hold while you travel. In the first scenario, I put the money in a safe place and wait for your return. If I was to account for this as a bank, I have a $100 asset (the cash) and a $100 liability (the money I owe you) and no equity at all. I am infinitely leveraged, you would think that is very high risk, even though this is a very safe arrangement! Now lets suppose another scenario. Instead of holding the money in a safe place, I bet the money on a coin toss. Before the toss, I would show $100 in assets (my ticket giving me a 50% probability of winning $200 would be worth that much) and a $100 liability. I am no more leveraged than the very safe previous arrangement, but there is only a 50% chance you will be paid when you get back. The point here is that leverage alone is not a good way to judge risk--you need to look at the profile of the firm.

Very generally speaking, liabilities are based on the value of what is owed at the point in time the balance sheet is dated. However, many assets are not. Some assets are not valued at all, some are valued at some point in time previous, and some are only valued based on whether they were transacted. Two firms with identical assets could show very different asset values. This is a big problem with using ROA. Because assets = liabilities + equity, and liabilities are relatively consistently measured and equity relatively small, the variations in how to measure assets are amplified in how equity is measured. Book equity is even more problematic to use in any type of analysis. Very often in analysis, book equity will be tossed out, and market equity used. The problem there being that in a bubble economy a firm may have a large equity value and not look leveraged, only to be seen as vunerable once its equity value plummets.

But this story also contains a puzzle. Long-term shareholders in banks have not obviously reaped the benefits of these distortions. The purchaser of a portfolio of global banking stocks in the early 1990s is today sitting on a real loss. So who exactly is it extracting value from these incentive distortions? The answer is twofold: shorter-term investors and bank management.

I'm interested in the phrase a "real loss". The use of the word "real" implies inflation adjusted, I wonder if there was also a nominal loss? I had seen post crash analysis that investors in US public investment banks did okay since they went public (small gains), even though 3 of the 5 no longer independently operate. Regardless, I don't think any diversified losses were too bad, and certainly not when you consider this is seen as the worst financial crisis since the Great Depression. You could argue that the behavior of firms caused the crisis, but firms act independently. In a properly aligned governance model between shareholders and management, firms will do what is in their best interest even if when universally applied could harm the economy. To moderate this problem, you need regulators.