The New York Times DealBook blog just put out a fine piece on the collapse of MF Global: A Romance with Risk That Brought on a Panic by Azam Ahmed, Ben Protess, and Susanne Craig (December 11, 2012). It’s the most comprehensive summary of the events that led to the firm’s bankruptcy that we’ve seen so far. Until the investigations are done and the books are written, it’s a good source for thinking about the credit risk lessons to be learned in MF Global’s sad story.
MF Global seems to have made a lot of risk management mistakes. It took on a big dose of market risk with its $6.3 billion exposure to the European debt crisis – over the objections of its senior risk manager. With only $1.4 billion in capital, the company could barely afford to take any losses.
There may have been operational risk issues at the firm as well. About $1.2 billion of client money has gone missing, and after weeks of searching the company’s records it’s still not clear where most of it is. At best, this is a serious systems failure; at worst, it could be a lot more sinister.
But the fatal risk at MF global was a form of credit risk called counterparty risk. That’s the risk that a firm involved in a trade fails to pay what it owes. Counterparty risk is where market risk and credit risk intersect: as a company’s trading losses grow, its ability to pay decreases.
Moody’s downgraded MF global from Baa2 to Baa3 in October, citing exposure to European sovereign debt, a regulatory capital shortfall, and poor risk management. The market had been concerned about MF global for months, but the downgrade made it official. MF global was a poor counterparty.
That triggered collateral calls, with trading partners and clearing houses demanding that the firm either close out its positions or post more collateral – usually in the form of cash. The firm that cleared MF Global’s sovereign debt trades demanded $300 million in cash collateral alone. MF Global’s risk challenges suddenly became a fatal liquidity problem.
That should not have been surprising. The same thing happened to Lehman Brothers and even to Goldman Sachs. The combination of collateral demands and a runoff in repurchase financing drove Goldman’s liquidity pool from $120 billion to $57 billion in just four days after Lehman went under.
MF Global’s counterparty credit squeeze suggests the likely explanation for the missing client money: desperate to meet collateral calls, the company used client funds to try to plug the holes in its liquidity dike. That changes the nature of MF Global’s operational risk problem. It’s worse than a mistake; it’s a crime.
There are a number of worthwhile accounts of the firms that failed or nearly failed in the great financial market collapse of 2007-2009. Too Big to Failby Andrew Ross Sorkin and All the Devils Are Here by Bethany Maclean and Joseph Nocera are two strong general accounts. House of Cards by William D. Cohan tells the Bear Stearns story very well, and Fatal Risk by Roddy Doyle does the same for AIG.
But the stories they tell are mainly about the personalities of the men who led those companies and the blunders they made in strategy and funding. None of them focus closely on risk management. So to correct that, we thought we’d start building a timeline of major risk management mistakes made in the run-up to the crisis.
It’s not comprehensive. It’s just a series of anecdotes drawn from books, press coverage, government reports, and other sources. Please help us build it up by contributing more stories. We’ll add them to the timeline as you do.
Sometime in 2005
Stan O’Neal, CEO at Merrill Lynch, puts Ahmass Fakahany, his Chief Administrative Officer, in charge of risk. Over the next two years, Fakahany dismantles Merrill’s risk organization, firing senior managers and moving the remaining risk managers off the trading floor. Merrill’s exposure to the mortgage market grows to $55 billion, and the company is forced to merge into Bank of America in 2008.
June 2006
Citigroup demotes Richard Bowen, a senior risk manager for raising the alarm over the decline in underwriting standards in its mortgage business. His report reaches Vice-Chairman Robert Rubin, but Rubin fails to act on it. The next year, Citigroup takes over $20 billion in mortgage loan losses.
September 2006
The Market Risk Committee at UBS warns of the high concentration of risk in the firm’s CDO warehouse program, which at its peak reached $50 billion. Management ignores the warning, and UBS takes $40 billion losses on its CDO portfolio in 2007 and 2008.
August 2007
Joseph Cassano, head of AIG Financial Products, refuses to give Joseph St. Denis, an internal auditor, information on potential collateral calls on Financial Product’s $420 billion portfolio of credit default swaps on mortgage backed securities. St. Denis leaves the firm, but Cassano stays. Just over a year later, when collateral calls peak at $80 billion, AIG is rescued from bankruptcy by a government bailout.
September 2007
Lehman Brothers removes Madelyn Antoncic, its well-regarded Chief Risk Officer, from the Executive Committee, clearing the way for the company’s disastrous Archstone-Smith acquisition in October. That drives Lehman’s commercial real estate exposures to $40 billion just as the market peaks. To avoid write-downs, Lehman tries to spin off its real estate portfolio. But that only highlights the company’s weakness, and Lehman fails in September 2008.
February 2008
Oliver Wyman issues an internal report criticizing Bear Stearns’ risk management for being a low priority. CEO Alan Schwartz says the report doesn’t indicate any substantial deficiencies. Within two weeks Bear Stearns’ funding evaporates, and the firm is forced to merge into JP Morgan.
Last post, we argued that Lehman’s Repo 105 balance-sheet-management tactic was not the worst thing Lehman Brothers did on its way to extinction. Volume 8 of Anton Vakulas’s Bankruptcy Examiner’s report details a bunch of blunders with far more serious consequences.
Here are a few of our favorites:
Although management was aware of the growing problems in the mortgage markets as early as 2006, Lehman decided to take on more risk “to pick up ground and improve its competitive position.”
It chose to do that by “making ‘principal’ investments – committing its own capital in commercial real estate, leveraged lending, and private equity investments.”
And it sacrificed liquidity along the way, so that “less liquid assets more than doubled during the same time from $86.9 billion at the end of the fourth quarter of 2006 to $174.6 billion at the end of the first quarter of 2008.”
But in our view, Lehman’s biggest missteps were in risk management. Lehman’s opportunistic push for growth changed into an aggressive push to offset declines in its commercial mortgage business late in 2007 and early in 2008, and that led the company to ignore its own risk controls.
This chart from the Examiner’s Report shows how Lehman’s risk appetite grew through 2007 and into 2008. Risk appetite was Lehman’s estimate of the amount it could lose without jeopardizing employee compensation or shareholder returns.
The chart also shows how in the last two quarters of 2007 the firm exceeded its own risk appetite limit with hardly any restraint, and then in the first quarter of 2008 solved the problem not by reducing risk but by raising the limit.
There are plenty of other instances of management’s undisciplined approach to risk management:
Management played games with key risk measures and “omitted some of the largest risks from its risk usage calculation” until June 2008 and “did not revise its stress testing to address its evolving business strategy.”
“Lehman rewarded its employees based upon revenue with minimal attention to risk factors in setting compensation,” in spite of the fact that managers were supposed to “use risk-weighted metrics such as return on risk equity…to determine compensation.”
After she resisted an increase in the firm’s risk appetite in 2007, Lehman replaced its highly regarded Chief Risk Officer, Madelyn Antoncic, in the third quarter of 2007 with the Chief Financial Officer, Rick O’Meara, who had no risk management experience.
Lehman did not see risk management as a crucial discipline or as a vital safeguard for a company operating with little capital and limited liquidity in troubled markets. For Lehman, risk management was a public relations tool.
It was a show put on to reassure clients, regulators, lenders, and the rating agencies but with no substance or bite. And that misuse of risk management was far more misleading and much more damaging than Repo 105.
A few weeks ago, the world was shocked to learn that Lehman Brothers was guilty of “window dressing” its balance sheet throughout 2007 and 2008. As the New York Times’ Dealbook put it, ”In Lehman’s Demise, Some Shades of Enron.”
The outrage is based on information in the examiner’s report on Lehman’s bankruptcy. We’ve read the report, and we think there is less to be upset about than the press coverage suggests.
The charge against Lehman is that they used accounting tricks to move assets of its books, making leverage ratios seem lower than they really were. The amounts were $38.6 billion in the fourth quarter of 2007, $49.1 billion in the first quarter of 2008, and $50.4 billion in the second quarter of 2008.
These seem like big numbers. They certainly meet the test for materiality set by Lehman’s auditors, Ernst and Young, which was “any item that moves net leverage by 0.1 or more (typically $1.8 billion).”
But let’s look at their impact not from a reporter’s point of view, or even a lawyer’s. Let’s look at this the way a credit risk analyst would.
This chart shows Lehman’s net leverage ratio as reported (without the Repo 105 assets) and as adjusted (with the Repo 105 assets). It was a measure the company emphasized in its public reporting and in its presentations to the rating agencies.
Net leverage is total assets less restricted cash, securities held as collateral, securities held to resell, borrowed securities, and intangible assets divided by stockholders’ equity less intangible assets.
When we analyze financial ratios we like to look at two dimensions of the data: level and trend. Let’s take a look at both for Lehman’s net leverage ratio.
Repo 105 does not make a big difference in the level of Lehman’s leverage. It changes from 12.1x to 13.9x in May 2008, for instance, but that’s not a decisive difference. When the examiner told the rating agencies about Repo 105, none of them said it would have made them change their ratings.
And Repo 105 does not change the trend in the company’s leverage. From November 2007 to May 2008, it fell by 25% with Repo 105 and by 22% without it. Either way, Lehman was cutting its leverage significantly.
Repo 105 makes for exciting headlines and promising legal claims, but it’s just not that big a deal. The public dismay over Lehman’s window dressing is disingenuous: banks and investment banks having been doing it for years.
The fuss reminds us of the police captain in Casablanca who is “ Shocked to find there’s gambling going on in here!” just as he gets his winnings from the croupier. Lehman did plenty wrong, but Repo 105 was not the worst by any means.
We hope everyone is having happy holidays. Last time, we defended Lehman Brothers from Andrew Ross Sorkin’s attack in Too Big to Fail. We’ve stolen some time from the seasonal festivities to take another look at the numbers, and we still feel there’s a strong case to be made for the way Lehman Brothers managed its finances right before its fall.
We compared Lehman to Morgan Stanley from November 2007 through August 2008, the three reporting periods leading up to Lehman’s collapse in September 2008. We looked at the progress each firm made improving three key measures of financial strength: leverage, exposure to mortgage-backed securities, and liquidity reserves.
We calculated leverage as the ratio of total assets to shareholders’ equity, exposure to mortgage-backed securities as the book value of residential and commercial mortgage-backed securities as a percent of shareholders’ equity, and liquidity reserves as total cash and unpledged liquid securities. To gauge the improvement in each measure, we calculated its value in August as a percent of its value in November.
In Sorkin’s account, Lehman Brothers went bankrupt because it had too little capital to absorb the potential losses from its real estate investments. But Lehman did much more to reduce leverage and cut exposure to mortgage-backed securities than Morgan Stanley, yet Morgan Stanley survived.
It’s true that Morgan Stanley did a better job of boosting liquidity than Lehman, but it needed to. Morgan Stanley did much more prime brokerage business with hedge funds than Lehman. It was the run-off in prime brokerage funds that brought Bear Stearns down, and it nearly ruined Morgan Stanley too. In the week after Lehman failed, Morgan Stanley went through nearly all $180 billion of its liquidity reserves and had to go to the Federal Reserve for rescue.
So what’s the risk management lesson in all this? Did Lehman Brothers deserve to die? Did it commit suicide, or was it killed? Could it have done anything to save itself? Too Big to Fail makes Lehman’s fate seem inevitable, and we agree.
Like any investment bank, Lehman relied on confidence-sensitive financing, and in September of 2008 the financial markets lost confidence in Lehman – and Morgan Stanley and Goldman Sachs. Nothing they could do about leverage, exposure to mortgage-backed securities, or liquidity reserves mattered.
It’s nothing new. As Walter Bagehot observed more than a century ago, “Every banker knows that if he has to prove that he is worthy of credit, however good may be his arguments, in fact his credit is gone.”
Except convince anyone it was doing anything right at all. In his new book Too Big to Fail, Andrew Ross Sorkin portrays the firm’s fall as a tale of arrogance, blindness, stupidity, complacency, and greed – nearly every vice but gluttony and lust. But the numbers don’t bear that out: Lehman reacted decisively to the problems it was facing.
November
February
May
August
2007
2008
2008
2008
Assets
$691,063
$786,035
$639,432
$600,000
Equity
$22,294
$24,832
$26,276
$28,443
Leverage
30.7x
31.7x
24.3x
21.1x
Liquidity pool
$35,000
$34,000
$45,000
$42,000
Repo financing1
37%
35%
33%
NA
1 as a percent of total funding
For financial institutions in trouble the prescription has always been: reduce assets, increase equity, improve liquidity. What did Lehman do? It cut total assets and raised shareholders’ equity, driving leverage down by 31%. It boosted its liquidity reserves to $42 billion, and trimmed short-term repurchase funding.
Lehman’s $2.9 billion loss for the third quarter of 2008 was hardly good news, but it came from what the Wall Street Journal called “savage cuts” to the value of its real estate. The Journal went on to say, “even longtime bears on the stock thought the firm was finally marking its residential portfolio to realistic levels.”
By the end of August 2008, Lehman had written down its troublesome real estate assets by 25% from the beginning of the year to $46 billion. With $28 billion in equity, it could afford to write them down by another 39% without destroying its capital base.
So why did Lehman’s damage control efforts fail? What sank the company? We’ll explore that in future posts, but it’s clearly not because Lehman ignored its problems, failed to act, or didn’t make real improvements to its finances.
The New York Times DealBook blog just put out a fine piece on the collapse of MF Global: A Romance with Risk That Brought on a Panic by Azam Ahmed, Ben Protess, and Susanne Craig (December 11, 2012). It’s the most comprehensive summary of the events that led to the firm’s bankruptcy that we’ve seen so far. Until the investigations are done and the books are written, it’s a good source for thinking about the credit risk lessons to be learned in MF Global’s sad story.
