The next addition to our Masters of Fraud series is former Morgan Stanley trader Howard Hubler. Hubler doesn’t exactly fit the mold of others in the series, because he didn’t actually commit fraud, at least not knowingly. What he did manage to do was incur an approximate $9 billion trading loss- the largest in the history of Wall Street.1 It is another stunning example of how important risk management techniques are and how even long-established policies can break down if not reviewed properly.
Housing Bubble Revisited
New Jersey native and former football Montclair State football player, Howard Hubler made his way to Morgan Stanley in the mid-1990’s.2 Hubler soon became a bond trader on Morgan’s asset backed securities desk. The types of securities traded on this desk are known as structured products, most notably mortgage backed securities (MBS). Mortgage backed securities are financial instruments comprised of batches of residential mortgages of various credit quality.
Many of the individual mortgage loans were ‘subprime’, meaning they were made to less credit worthy borrowers (typically with FICO scores under 650).3 If you recall, the mortgage lending standards of the last decade had gotten very relaxed. Financial companies, banks and mortgage originators were offering mortgages to people that had no business buying houses.
Some were known as NINJA loans (an acronym standing for No Income No Job).
Many loans were being made at the ‘request’ of authorities to underprivileged borrowers. But the theory went that by grouping all these mortgages together, the benefits of diversification reduced the overall risk of individual subprime loans.
Once the mortgages were originated, most of these financial companies would sell them off to be packaged and resold to investors, often institutions such as pension funds and life insurance companies. The main point is that the majority of loan originators didn’t keep the loans on their books and didn’t bear the default risk. That risk was passed on to investors.
When that’s the case, credit quality slips. And as long as there was institutional demand for the products, helped along by risk-free stamps of approval from credit rating agencies, the MBS machine kept running. This easy credit was fueling the housing bubble.
This type of trading done by asset backed traders was described by Michael Lewis in his book, The Big Short, as a low stakes poker game rigged in their favor.4 The MBS market was considered rather sleepy, until the housing bubble of the mid 2000’s. After all, the housing market slowly and steadily went up a few percent every year.
The underlying sales pitch was the national housing market has never gone down.
And it worked-low interest rates and securitization helped the MBS market become enormously profitable. Hubler’s team was generating hundreds of millions in profits for Morgan Stanley from this operation. By 2006, Hubler and his group of 8 bond traders accounted for roughly 20% of Morgan Stanley’s profits, by their own estimates.5
But a growing portion of these mortgage backed securities were starting to worry some. During the relatively short time Hubler wholesaled the mortgages (the difference between when they bought the mortgages from the originators and sold them off to investors as MBS) they were exposed to the risk that the loans could go bad. Subsequently, the credit default swap (CDS) was created to hedge the positions they were wholesaling.
A CDS is basically an insurance policy on the MBS in case it defaulted and, like an insurance product, the premium must be paid annually to continue ‘coverage’. You can see how these financial securities might be confused with insurance products (many regulators wanted them classified as such). Others view CDS more like a put option. Regardless, Hubler and his team saw the warning signs and smartly bought some of this protection on the riskier, BBB-rated tranches of MBS.
The Hedge Gone Bad
The mistake came when Hubler and the team got a little too cute and tried to fund the CDS purchase by selling credit default swaps on the triple-A rated MBS. When selling the CDS, you take in the premium, like the insurance company, but are obligated to payout if the insured event occurs. And since the price to buy insurance on the ‘safer’ securities was low, he had to sell that much more of it to offset the cost of buying protection on the riskier securities.
By January of 2007, Howie Hubler had sold credit default swaps on $16 billion worth of triple-A rated securities. In other words, he insured the value of the instruments he was betting against.
But by the summer of 2007, the mortgage market was beginning to unravel. New Century, the largest subprime lender in the U.S., had filed for bankruptcy.
Eventually, the pristine triple-A securities collapsed right along with the junkier BBB tranches. Those MBS, priced at 100, were now trading at 7.6 Needless to say, this was a major black eye for the credit rating analysts at Standard & Poor’s and Moody’s. Even though both types of securities collapsed, there was a massive net loss of $9 billion for Hubler since he had disproportionately sold more insurance on the AAA MBS.
To put this loss in context, this is roughly twice as large as Jerome Kerviel’s trading loss and 12 times larger than Nick Leeson’s rogue trade that brought down Barings Bank. Had this kind of a loss been at any other bank, it surely would have gone under. But Morgan Stanley had around $70 billion in capital, so the loss represented around a 13% loss of total capital-severe but not catastrophic.7
Where were the Risk Managers?
Once again, there appeared to be a major lapse in oversight. How could Hubler have amassed a gargantuan $16 billion position in these securities? Well, since the securities contained essentially risk-free assets (triple-A credit ratings), the trades came across in the AAA ‘bucket’ and caused no alarm. They were treated like U.S. government bonds. Of course, these ‘risk-free’ securities that were priced at 100, would lose 93% of their value.
Another lapse was the risk management calculation that Morgan Stanley was relying on, ‘Value at Risk’ or VAR. This calculation, still used today, is backward looking and uses historical volatility to predict future volatility. In ‘normal’ periods it’s a good measure of inherent risk but not during a major disruption. In other words, the system works well until it doesn’t.
Many sources believe that the higher ups at Morgan Stanley simply didn’t understand the risk on the AAA-rated securities, including Hubler and his group. Since the VAR is backward looking, and the housing market had never experienced anything like this since the Great Depression, everything seemed fine.
The lesson is learned is that in a strained financial situation, correlations converge toward one, negating the benefits of diversification, ironically when they’re needed the most. Today’s risk managers must be more aware of backward-looking risk detection techniques and take a more proactive role before any position sizes can get out of hand.
After Morgan Stanley
“After the extent of the damage was realized by Morgan Stanley’s management and risk teams, Hubler was given the option to resign instead of being fired in October 2007. He was paid $10 million in back pay upon his departure².”
4,5,6,7Lewis, Michael. The Big Short. New York City: W.W. Norton & Company, 2010. Print.