Howie Hubler: The $9 Billion Subprime Blunder

Key Takeaways

• A correct $2 billion subprime short was buried under a $16 billion bet on triple-A safety.
• To fund the short's premiums, the desk sold protection on supposedly safer AAA CDOs.
• When AAA mortgage tranches cratered too, the loss is reported near $9 billion.
• The lesson: hedges that share the same risk are not hedges at all.

Background

By the mid-2000s, Morgan Stanley wanted a bigger share of the booming trade in subprime mortgage derivatives. Howie Hubler had built a strong record on the firm's bond desk, and in 2006 the firm gave him his own unit, the Global Proprietary Credit Group, to trade the firm's own capital. According to the Shortform summary of Michael Lewis's "The Big Short," by April 2006 Hubler's group was generating an estimated 20 percent of Morgan Stanley's profits, which made him one of the most valued traders in the building.

Hubler shared the skepticism that a handful of outside investors had about subprime mortgages. He believed the riskiest slices of mortgage bonds, the lower-rated mezzanine pieces packed into collateralized debt obligations, were likely to default. A collateralized debt obligation, or CDO, is a security built by repackaging other debt, in this case mortgage bonds, and slicing it into tranches that pay out in order of seniority. The safest tranches carried triple-A ratings; the mezzanine tranches sat lower and absorbed losses first.

So Hubler did what the famous shorts of the era did. He bought credit default swaps on the mezzanine subprime CDOs he expected to fail. A credit default swap works like insurance: you pay a recurring premium, and if the underlying bond defaults, the seller pays you the loss. The reported size of this short was about $2 billion, and on its own it was a sound, even prescient, position.

The trouble was the cost of carrying it. Buying that protection meant paying out premiums month after month while waiting for a default that might be years away. Hubler needed a way to cover that bleed, and the solution he chose is what turned a good trade into a catastrophe.

What Happened

The trade had two legs that pulled in opposite directions, and only one of them was a real hedge. The first leg, the subprime short, was correct. The second leg, sold to pay for the first, is what broke.

To offset the premiums he was paying on his short, Hubler had his desk sell credit default swaps on triple-A rated CDOs. Selling that protection brought in premium income, the mirror image of what he was paying out. But the premiums on AAA paper were tiny, because almost no one thought the safest mortgage tranches could ever default. To raise enough income to cover his short, he had to sell protection on a far larger amount of AAA bonds than he had shorted.

• 2006: Morgan Stanley creates the Global Proprietary Credit Group under Hubler; by April 2006 it reportedly generates around 20 percent of firm profits.
• 2006 to early 2007: The desk buys roughly $2 billion of credit default swaps shorting mezzanine subprime CDOs, and sells roughly $16 billion of credit default swaps on triple-A CDOs to fund the carry.
• Mid-2007: Subprime delinquencies surge; the value of the short rises, but the much larger AAA position starts losing far more.
• July 2007 onward: As the crisis deepens, the AAA tranches the desk had insured fall in value alongside the junk they were meant to be safer than.
• October 2007: With the scale of the damage clear, Hubler leaves Morgan Stanley; reporting says he was allowed to resign.
• December 19, 2007: Morgan Stanley discloses $9.4 billion in mortgage-related writedowns for its fiscal fourth quarter, including $7.8 billion tied to U.S. subprime trading positions.

Because he had to sell about ten times as much AAA protection as the junk he shorted to balance the cash flows, the position was wildly lopsided. Selling protection on a bond is economically the same as owning it: you collect a small premium and you are on the hook for the full loss if it goes bad. Hubler had effectively bought, in size, the very kind of mortgage risk he was betting against in smaller size elsewhere. The two legs were not opposites. They were the same bet, in different sizes, pointing the same way.

Why It Happened

The failure was not a failure of the short. The short was right. The failure was the structure built around it, and three assumptions inside that structure.

The first assumption was about correlation. Hubler's desk treated the AAA CDO tranches as a different, safer asset than the mezzanine tranches it was shorting. In a normal market, that distinction holds, because senior tranches only take losses after the junior ones are wiped out. But the AAA and mezzanine tranches were carved from the same pools of subprime loans. When defaults rose across the whole housing market at once, losses tore through the junior tranches and kept going into the senior ones. The thing that was supposed to make the AAA bonds safe, the cushion of lower tranches beneath them, was too thin for a nationwide collapse. Correlation went to one, and the safe leg and the risky leg fell together.

The second assumption was about the ratings. The desk relied on the triple-A label to mean the bonds were close to riskless, which is why the premium income from selling protection on them was so small and why the desk had to sell so much of it. The FCIC's final report describes how the most senior tranches of subprime ABS CDOs, the so-called super senior positions that banks kept on their own books, were widely assumed to be safe and were valued off market indicators rather than the loans inside. When those assumptions broke, the super senior paper that everyone had treated as money-good repriced sharply lower.

The third problem was the negative carry that started the whole chain. A short on subprime CDOs cost money to hold, in the form of ongoing premiums. Rather than simply pay that cost and size the short to survive, the desk tried to make the position self-funding by writing protection on AAA bonds. That decision, made to solve a cash-flow problem, is what converted a modest, correct bet into a giant, wrong one. The hedge against the cost of the trade became many times larger than the trade it was meant to support.

There was also an incentive backdrop. The group had been a major profit center, and a large, steadily-earning book of AAA premium income looked like low-risk earnings right up until it was not. Positions that print small, consistent gains while quietly holding rare but enormous downside are a recurring pattern in trading blowups, and this was a textbook case.

By the Numbers

• Subprime short (mezzanine CDOs): roughly $2 billion of credit default swaps purchased. (Earn2Trade; Shortform, per Michael Lewis. Reported estimate.)
• AAA protection sold (long risk): roughly $16 billion of credit default swaps written on triple-A CDOs. (Earn2Trade; Shortform, per Michael Lewis. Reported estimate.)
• Cash-flow ratio: about 10 to 1, the amount of AAA swaps sold per unit of subprime swaps bought. (Shortform, per Michael Lewis. Reported estimate.)
• Sold to a single counterparty: about $4 billion of swaps to Deutsche Bank. (Shortform, per Michael Lewis. Reported estimate.)
• Reported trade loss: around $9 billion net when Morgan Stanley exited. (Shortform; Earn2Trade; CBS News, per Michael Lewis. Reported estimate.)
• Disclosed mortgage writedowns, fiscal Q4 2007: $9.4 billion total. (CBS News; Deseret News, reporting Morgan Stanley's December 19, 2007 disclosure. Documented.)
• U.S. subprime trading portion of the writedown: $7.8 billion. (Reported from Morgan Stanley's December 19, 2007 disclosure.)
• Fourth-quarter net loss: about $3.6 billion, the first quarterly loss in the firm's 73-year history. (CBS News; Deseret News. Documented.)
• Outside capital raised: $5 billion from China Investment Corporation, for up to a 9.9 percent stake. (CBS News; Deseret News. Documented.)
• Hubler's reported exit pay: about $10 million in deferred or back pay. (Earn2Trade. Reported.)

Note the distinction. The $9.4 billion writedown, the $3.6 billion quarterly loss, the 73-year record, and the China Investment Corporation stake are documented figures from Morgan Stanley's December 2007 disclosure as carried by contemporaneous reporting. The $2 billion short, the $16 billion AAA position, the 10-to-1 ratio, and the round $9 billion trade-loss number trace largely to Michael Lewis's account and the summaries of it, and should be read as reported estimates rather than figures Morgan Stanley itself published line by line.

Aftermath

Hubler left Morgan Stanley in October 2007, and reporting describes him as being allowed to resign rather than be fired, departing with several million dollars of pay the firm had previously promised him. He was not charged with any crime, and no regulator has alleged fraud in connection with the trade. The losses came from a bet that went wrong, not from a finding of wrongdoing.

The firm absorbed the damage publicly on December 19, 2007, when it disclosed $9.4 billion of mortgage-related writedowns and a fourth-quarter net loss of about $3.6 billion, the first quarterly loss in its 73-year history. To shore up its capital, Morgan Stanley sold a stake of up to 9.9 percent to China Investment Corporation for $5 billion. Chief executive John Mack accepted public responsibility and gave up his bonus for the year, and the firm attributed the bulk of the loss to positions held by a small trading group in one part of the business.

The episode became one of the defining set pieces of Michael Lewis's 2010 book "The Big Short," which framed Hubler as the trader who got the subprime call right and still managed to lose more than almost anyone. In Lewis's telling, the loss was the single largest trading loss in Wall Street history to that point, though other accounts rank it second; either way, it sits among the largest losses ever traced to one trader's book. Morgan Stanley survived, unlike Bear Stearns and Lehman Brothers, but its near-miss fed directly into the wider reckoning over how banks valued and held the senior tranches of subprime CDOs.

Lessons for Investors

1. A hedge that shares the same risk is not a hedge. Hubler's AAA position was sold to offset his subprime short, but both legs depended on subprime mortgages performing. When the housing market broke, they moved together instead of against each other. Before you call something a hedge, ask what single event would sink both sides at once, and if there is one, you are not hedged.

2. Negative carry tempts you into the wrong fix. The short cost money to hold, and the desk solved that cash-flow drag by selling far more protection elsewhere. Paying to hold a correct position is uncomfortable, but manufacturing income by taking the opposite risk in size can quietly invert your whole bet. Size a position so you can afford to carry it, rather than funding it with a bigger exposure.

3. A rating is an opinion, not a guarantee. The whole structure rested on triple-A meaning near-riskless, which is why so little premium changed hands and why the position grew so large. Senior tranches built from the same weak loans as the junk beneath them were never as safe as the label implied. Treat a credit rating as one input to verify, not a fact to build your downside around.

4. Small, steady income can hide a fat tail. Writing protection on AAA paper looked like reliable, low-risk earnings until correlation spiked. Strategies that earn a little most of the time while exposing you to rare, enormous losses are a classic trap. Judge a position by its worst plausible outcome, not by the smooth returns it shows in calm markets.

5. Being right on the thesis is not enough. Hubler's core view, that subprime mortgages would default, was correct, and he still produced one of the biggest losses on record. The structure around a correct idea, the sizing, the funding, and the assumptions about what moves with what, can matter more than the idea itself. A right call implemented wrong can lose more than a wrong call ever would.

Frequently Asked Questions

Who was Howie Hubler and what did he do? Howie Hubler was a Morgan Stanley bond trader who ran the firm's Global Proprietary Credit Group from 2006. He correctly shorted about $2 billion of risky subprime CDOs but funded that bet by selling roughly $16 billion of protection on supposedly safer AAA CDOs, which then collapsed too.

Why did the Howie Hubler trade lose money if the short was right? The short was small and correct, but the much larger AAA position was effectively a long bet on the same subprime mortgages. When the whole housing market fell, the safe-looking AAA tranches cratered alongside the junk, and the larger losing leg overwhelmed the smaller winning one.

How much money did Morgan Stanley lose on the trade? The trade loss is widely reported at around $9 billion. Morgan Stanley separately disclosed $9.4 billion of mortgage-related writedowns for its fiscal fourth quarter of 2007, including $7.8 billion on U.S. subprime trading positions, and posted a roughly $3.6 billion quarterly loss.

Was Howie Hubler charged or convicted of anything? No. Hubler was not charged with any crime, and no regulator alleged fraud over the trade. Reporting says he left Morgan Stanley in October 2007, allowed to resign, and departed with several million dollars of previously promised pay.

What is the main lesson from the Howie Hubler loss? The core lesson is that a hedge built on the same underlying risk is not protection at all. Selling AAA protection to fund a subprime short only worked while the two were uncorrelated, and when correlation went to one, both legs lost together.

Sources

1. CBS News. Morgan Stanley Reports $9.4B Writedown. December 19, 2007. https://www.cbsnews.com/news/morgan-stanley-reports-94b-writedown/
2. Deseret News. Morgan Stanley posts loss after $9.4B 4th-quarter writedown, gets $5B investment from China. December 19, 2007. https://www.deseret.com/2007/12/19/20060179/morgan-stanley-posts-loss-after-9-4b-4th-quarter-writedown-gets-5b-investment-from-china/
3. Financial Crisis Inquiry Commission. The Financial Crisis Inquiry Report, Chapter 8: The CDO Machine. 2011. https://fcic-static.law.stanford.edu/cdn_media/fcic-reports/fcic_final_report_chapter8.pdf
4. Shortform. Howie Hubler: The Blunder That Cost Morgan Stanley $9 Billion (summary of Michael Lewis, The Big Short). https://www.shortform.com/blog/howie-hubler/
5. Shortform. Morgan Stanley's 2008 Crisis: How the Bank Lost $37 Billion (summary of Michael Lewis, The Big Short). https://www.shortform.com/blog/morgan-stanley-2008/
6. Earn2Trade Blog. Howie Hubler: One of the Largest Losses of the 2008 Crisis. https://www.earn2trade.com/blog/howie-hubler/
7. CBS News (MoneyWatch). 6 Ugliest Trades in the History of Investing. https://www.cbsnews.com/media/6-ugliest-trades-in-the-history-of-investing/

Disclaimer

This article is educational content only and is not financial advice. Nothing here is a recommendation to buy, sell, or hold any security. Consult a licensed advisor before making investment decisions.