[First, a couple housekeeping items. Due to changes in my circumstances, there are good news and bad news this week. The bad news is that this blog will slow down a bit — I’m currently targeting a rate of one to two posts a month. The good news is that MaverickQuant is now free, including its archived content. (If you have a paying subscription, payment is now suspended, so you don’t have to do anything. Thanks for your support!). Stay tuned, and spread the word! — JB]
[Jérôme BUSCA] This week, I am interviewing my colleague Robert Benhenni, who is a seasoned professional on Wall Street. I met Robert in New York where we collaborated on several projects involving quantitative investment strategies across asset classes. He has broad experience in the financial industry, with nearly 20 years working in New York with investment banks and hedge funds. He now runs his own consulting and product development firm QFA Inc., which services the financial industry in the US and Europe across risk management, quantitative research, and investment strategies. Hello Robert and welcome to Quant Talk. First of all, can you tell our readers about your background and your career?
[Robert Benhenni] Thank you for such a nice introduction. Glad to be here. After my Ph.D. in Applied Mathematics (Statistics) at UCLA, I first joined Bell Labs, where I worked on the application of queueing systems to telecommunication networks. Simultaneously, I enrolled in the MBA at the University of Chicago Booth School of Business. Thereafter, I moved to First Chicago bank (now JP Morgan) to work in the Global Derivatives division as a desk quant for commodities derivatives where I developed more accurate American and barrier options models for pricing and hedging commodity derivatives. An outcome of this project was its publication in Derivatives and Financial Mathematics 1. Going beyond my quantitative focus, I tried to get an economic sense of the derivatives industry which led to a publication of an article I wrote in Global Investor2 to explain how the derivative industry really functions. Shortly after, I moved to New York where I worked for different financial institutions.
[JB] Could you tell me about your work in New York?
[RB] I first joined CDC Ixis (now Natixis) at a time where CDOs were in vogue. I had the chance to work on one of the hottest CDO that came to the market. It had the appealing name Synthetic CDO BISTRO issued by JP Morgan. Some of the investors I met who were investing in the different CDO tranches relied mostly on the rating by the rating agencies, but, being a quant, I developed my own CDO analyzer. This synthetic CDO was different from previous traditional CDOs as it relied on Credit Default Swaps and was highly leveraged. The nuances of this BISTRO CDO may have escaped some of the investors and I could certainly understand a lack of a full analysis being exacerbated by the new Credit Default Swap market and excessive reliance on rating agencies. As a consequence of the quantitative analysis of the BISTRO deal, we purchased one of the tranches, but with an insurance wrap that turned out to be the right move.
[JB] That must have been very exciting being in the middle of all the Synthetic CDO deals. What were your next exciting projects at Natixis?
[RB] It just happened that Natixis had a hedge fund on the side that was planning to launch a new quantitative corporate credit fund. I was called in to build the credit quantitative strategy and be the portfolio manager for the new credit fund. I would say that this was probably my best time in Wall Street as the fund was successfully launched based on the credit investment strategy I developed, and did very well. As we all know, the good times in Wall Street don’t last too long! As the parent company based in Paris decided to acquire a large US traditional asset management firm which was asked to consider our small hedge fund in New York, it was decided that our quantitative hedge fund did not fit the traditional asset management framework and was therefore liquidated. During the time from the US asset management firm acquisition to the time of liquidation, I was put in charge of all portfolio risk management and headed the risk management committee. As one expects, any liquidation leads to conflicts and tensions including senior management and I found myself playing the Henry Kissinger diplomacy of small steps to calm the tensions which was a role that I didn’t plan for my career. This is an experience I would not want to repeat, but was a learning opportunity.
[JB] Wow! Going from great times to unfortunate times! What was your next move after you left the hedge fund?
[RB] First, I decided to try entrepreneurship and launched an LLC with an old colleague of mine to build a novel Capital Structure valuation model for use in investment strategies such as Capital Structure Arbitrage. Both Goldman Sachs and Credit Suisse were interested in this new approach and used it. I then moved to Morgan Stanley where I built novel corporate default statistical prediction models for both non-financial and financial firms covering the full US market that had high forecasting power. Lastly, I created for a US credit asset manager the first-ever Middle-Market company performance index providing early investment insights into major stock indexes: S&P 500, S&P SmallCap 600 and Russell 2000, that the media (WSJ, CNBC) have discussed. This was a great satisfaction, as my work was discussed in the media.
[JB] How important is quantitative modeling to your job?
[RB] Quantitative modeling is a big part of everything I do. It is intellectually stimulating and practical at the same time, and should be undertaken only when there is a clear added value to the business. Most importantly, it should follow the KISS principle — “Keep It Sophisticatedly Simple” — as Prof. Arnold Zellner, whom I met at U. of Chicago, always says. In my experience, the best models that work in practice are always simple.
[JB] As we know, quantitative equity investment strategies have been around for some time. How can you explain that it is not the case for quantitative corporate bond investment strategies?
[RB] Historically, corporate bond trading has relied on traditional credit analysis involving fundamental ratios, among others. Unlike equity investments, where there is a lot of readily available data facilitating a quantitative approach, corporate bond data sets are more complex, with different maturities and coupons per issuer and not as easily accessible and hence hindering a quantitative approach. But it is not too late! With advanced software and more available electronic trading platforms, asset management firms and hedge funds have been allocating more resources over the last few years building bond data infrastructure allowing them to build systematic corporate bond investment strategies, departing to some extent from traditional credit analysis.
[JB{ What’s your take on machine learning/AI techniques?
[RB] Generally, the corporate credit market has been slow to adopt quantitative methods for investment purposes. This is now changing as we speak. Clearly, AI techniques — specifically LLMs — can help analyze vast amounts of textual data. In particular, correctly interpreting covenants at scale is crucial for bond valuation. Another application would be predicting firms’ creditworthiness using multi-factorial models and machine learning techniques. I believe neural networks have tremendous potential for forecasting yield spreads and liquidity.
[JB] What do you think students should study to thrive as quants?
[RB] I teach several courses in quant finance at NYU in the department of Finance and Risk Engineering. I had the opportunity to observe that the students who were offered the most promising jobs on Wall Street are those who excel in statistics, coding, and financial engineering.
[JB] Fantastic! Thank you for coming to our interview segment.
[RB] Thanks for having me.
“Lattice Methods for Exotic Options”, Li A., B.R., Derivatives and Financial Mathematics, Nova Science 1997
The Structure of the Derivatives Industry”, B.R., Global Investor, February 1998