A Professor, a Whistleblower, and Ethics For Quants
On December 7, I published an article entitled “Deutsche Bank: Explaining The $12 Billion Loss That Never Was.” The piece outlined a series of complaints filed by former Deutsche Bank employees. One of those employees, Matthew Simpson, claimed to have discovered “substantial anomalies” in the firm’s credit default swap book while working at Deutsche’s credit correlation desk. Deutsche -- of course -- denied the allegations but did fire a top derivatives trader after an internal investigation into the matter and ultimately paid $900,000 to settle a related SEC whistleblower case filed by Simpson. Reuters broke Simpson’s story in the summer of 2011.
On December 5, 2012, The Financial Times reported that according to another former risk manager, a Mr. Eric Ben-Artzi, Deutsche Bank failed to account properly for the so-called ‘gap option risk’ associated with its leveraged super senior (LSS) book. In leveraged super senior deals, the protection buyer (in this case Deutsche) purchases insurance on the so-called super senior tranche of a collateralized debt obligation. The seller of that protection however, only posts collateral representing a small portion (typically 10%) of the notional amount insured. If the protection seller simply walks away from the deal, the protection buyer is left with nothing but the posted collateral. The myriad risks involved with these types of transactions are discussed at length in my previous piece and in a follow up article, so I won’t delve too deeply into them here, but suffice it to say that according to Ben-Artzi, Deutsche was accounting for these trades as though they weren’t leveraged (i.e. as though they were fully collateralized), a practice which allowed the bank to hide some $12 billion in paper losses during the financial crisis.
As most journalists no doubt know, the best time to respond to readers’ comments is often not when they are first posted—it is easy to come across as unduly testy or defensive. As such, I like to revisit the comments section of my articles after some time has passed to see what was said and respond to any comments I find intriguing. As I was perusing the comments on my first Deutsche Bank article, I noticed the following post which I found quite interesting:
Hi I am Marco Avellaneda, from NYU. Some of you probably know me; others not. Great.Here is the thing. All banks under-reported their exposure to CDOs. Starting with the major US banks GS,MS,JPM and down the line.The point here is that SEC brough (sic) no convictions and no serious cases against any of the major Wall Street banks. I believe that the top 4 or 5 are ``exempt'' because they are too big, interconnected, or simply well-connected, to fail.But this stands a blemish (sic) to the Securities and Exchange commission, which will go after (of course) subsidiaries of European banks. The latter are not ``represented'' in government or the department of Treasury, etc. OK? The case against DB is very weak. For a short period of time, DB hid 12BB in marl-to-market losses and thus misled shareholders by omitting material information about the stock. Should we go back and look at Goldman, JP Morgan and Bof A's books? Don’t think so, ladies!
I must have seen this initially because I did post a one-sentence reply related to Goldman’s LSS gap option modeling. After rereading the response however, it occurred to me that perhaps this was someone I should know – indeed the commenter basically says as much. As it turns out, Dr. Marco Avellaneda is a prominent mathematics professor at NYU’s prestigious Courant Institute which, according to U.S. News & World Report, is the number one applied mathematics school in the country.
Although Dr. Avellaneda told me in our e-mail correspondence that his “stature was quite normal,” a bit of research tells a different story. Avellaneda penned a 336 page book on the quantitative modeling of derivatives and was named “Quant of the Year” by Risk magazine in 2010 for “his groundbreaking work on the effect of short-selling restrictions on price dynamics.” Interestingly, he is also a partner at Finance Concepts, which certainly appears to be a hedge fund (click here to read an amusing piece by Avellaneda about the inception of that fund).
Given all of this, I found it peculiar that Dr. Avellaneda would so publicly trivialize the allegations made by two former Deutsche Bank risk managers. Furthermore, I was surprised that someone of such high repute would openly accuse the SEC of protecting politically connected U.S. institutions while pursuing investigations against politically defenseless foreign firms (of course the idea that Deutsche Bank is not every bit as well connected as any other large bank is laughable, something Matt Taibbi made abundantly clear in his 2011 piece entitled “Is The SEC Covering Up Wall Street Crimes”). Finally, I wanted to know if Avellaneda had any evidence for the contention that Goldman Sachs deliberately mismarked its crisis-era CDO book. Ben-Artzi once worked for Goldman Sachs after all, so it would certainly seem odd that he would deride Deutsche Bank and not Goldman if both firms were actively engaged in the mismarking of their respective derivatives books.
I e-mailed Dr. Avellaneda to see if he would be willing to do an interview with me. He responded promptly and indicated that he would be happy to speak with me about his comments. He also informed me that Eric Ben-Artzi was once his PhD student. It occurred to me that all things equal, one would expect that Dr. Avellaneda would be generally supportive of a former student who attempted to use what he learned in the classroom to protect the interests of shareholders and to generally help ensure that the routine mismarking of CDS positions didn’t imperil the entire financial system. This is not to suggest that Dr. Avellaneda was not being supportive, only that his comments seemed to indicate that he harbored a bit of healthy skepticism about the significance of Ben-Artzi’s claim. Avellaneda would later confirm this perceived skepticism in our e-mail correspondence.
As I crafted my questions I began to wonder about the importance of including ethics courses in programs that train future quants. With the universe of innovative financial products expanding rapidly, the mathematician’s role in keeping firms honest cannot be overstated. The more complex the securities become, the more important the role of the quants hired to manage the associated risks will be. As such, quants hired to oversee complex derivatives books (the notional value of which numbers in the trillions) must be held to the highest ethical standards. Let us not forget that it was a ‘faulty’ VaR model that allowed JPMorgan’s CIO office to hide the extent of its losses in early 2012.
In any event, Dr. Avellaneda agreed to submit written responses to my list of questions after which I was to call him so I could ask any follow-up questions I might have. What follows is the list of questions I sent to him:
1. So just to give readers some kind of context, Eric Ben-Artzi was a student of yours in the PhD program at NYU, is that correct?
2. I want to give you a chance here to sort of, restate your opinion on this whole issue irrespective of your comment on my article on Seeking Alpha. How important is this issue in terms of both its impact on quantitative modeling of derivative risk and in terms of what quants’ ethical obligations are once hired at a Wall Street firm?
3. So obviously I have to ask you about the actual LSS gap option. Given your experience in this field as both a noted academic and as a practitioner (your fund) I’m interested to know if you agree with Mr. Ben-Artzi’s valuation methodology for the LSS? If not, can you point readers to an academic paper, white paper, etc. which might help them understand where perhaps Mr. Ben-Artzi could be mistaken?
4. My next question references your comment on my original article about this whole issue. In that comment you said “All banks under-reported their exposure to CDOs. Starting with the major US banks GS, MS, JPM and down the line.” I assume by “GS” you meant Goldman Sachs, Mr. Ben-Artzi’s former employer. So it seems pretty clear from what I’ve read that Mr. Ben-Artzi’s position is that Goldman did not inflate the value of its CDO portfolio. I would imagine Lloyd Blankfein would say the same thing. Can you give some specific examples that support your claim that Goldman Sachs inflated the value of its CDO portfolio?
5. Next, let me ask you about something you said to me in your e-mail response if I may. You said “Besides, it seems that is was retroactive in nature: Eric Ben-Artzi joined DB after all this took place and the gap was filled.” Can you say a little more about the “gap being filled?” In other words, can you tell readers the extent to which the $10-12 billion gap option risk identified by Ben-Artzi was filled by the time he arrived at Deutsche Bank? This would help readers understand why Ben-Artzi was acting “retroactively.” It should be noted here that many people have said that Ben-Artzi’s claims are retroactive and as such are sort of devoid of much substance. So maybe your comments could help bolster that position.
6. Theoretically, if there were say, $200 million of this risk still on Deutsche Bank’s books, would that, in your mind, be “material” and therefore worth mentioning by any quant worth his salt that was hired to model risk for the firm?
7. This is related to question number 6. It is my understanding that Mr. Ben-Artzi was hired at Deutsche Bank to model risk. Given that the leveraged super seniors were still at the top of the list in terms of being the biggest risks on the book at the time Ben-Artzi was hired, was he not just doing the very thing he was hired to do by investigating how that specific position was valued? In other words, some commentators (notably DealBreaker’s Matt Levine) have cast Ben-Artzi as a kind of annoying meddler, but it would appear that he was in fact doing what he was hired to do and what he had been doing for years at Goldman. Help me understand what Ben-Artzi should have done differently in terms of assisting his employer in managing the risk associated with their leveraged super seniors.
8. So I come from a family of academics. My father and step-father have both taught at the university level for decades. My intuition (now I want to be careful here to say that this is just my intuition and it may be completely wrong) is that NYU would be proud that their PhDs are out in the world “fighting the good fight,” if you will. In other words, isn’t Mr. Ben-Artzi an example of what quants are supposed to be doing in terms of identifying risks and going to whatever lengths are necessary to ensure that shareholders aren’t unduly hurt by the mismanagement of those risks?
9. Courant's math finance program is widely considered one of the top in the world. Do you offer an ethics course? As a leading professor in the program, do you encourage your students to report fraud or ethical lapses they may see when they work on Wall Street? Do you think it's an important issue?
10. This question also references your comment on my previous article. You said that “This stands as a blemish to the Securities and Exchange commission, which will go after (of course) subsidiaries of European banks. The latter are not ``represented'' in government or the department of Treasury, etc. OK?” So that’s pretty adamant. Can you cite any examples of any actual enforcement actions (as opposed to investigations) against Deutsche Bank that would support the idea that the SEC is truly “out to get Deutsche Bank” as a result of its being less well connected politically than say, a JPMorgan?
11. Do you feel that Mr. Ben-Artzi did the right thing in reporting his concerns? If not, why not? Have you reached out to him to discuss the matter?
Perhaps Dr. Avellaneda was surprised at the nature or the depth of the questions, but he ultimately declined my interview after reading them, offering instead to discuss the issue over beers at some future date. “I feel that it would not help Ben-Artzi, Courant, or myself to comment on this,” he told me. This seemed like a particularly ironic turn of events to me given that it was Dr. Avellaneda’s comments on my article that inspired me to request an interview in the first place. That is, he had already commented on the issue in a public forum – I was simply asking him to clarify and support his comments and to perhaps offer some guidance regarding the ethical obligations of quants working in the financial world.
Ultimately, I believe these questions are important and I certainly imagine there are quite a few readers who would be very interested to hear them answered. Did all U.S. banks underreport their CDO exposure? Does the SEC go easy on the most well connected of firms (I guess that one is more rhetorical than anything)? Was the gap option filled at Deutsche Bank when Ben-Artzi was hired? How much of that position remains on the books today and is that amount material to shareholders? What are the ethical obligations of quants? Consider this an open letter to anyone with knowledge of these issues who might be willing to share their thoughts. The shareholders of major U.S. financial institutions would appreciate the candor, I can assure you of that.
Note: Mr. Ben-Artzi could not be reached for comment on this piece, similarly, my e-mails to Dr. Marco Avellaneda requesting comment were not returned. Should either individual decide to comment I will update the piece.