Financial engineering as a career: Part 1
The other day the new MSFE students showed up at Columbia for orientation and I had to welcome them. These are some notes from what I said.
Several years ago, my son, who did a PhD thesis on the reception history of Max Weber, the founding father of sociology, introduced me to two influential essays by Weber, entitled respectively Science as a Vocation and Politics as a Vocation. In them Weber discusses what problems you have to face, and what personality and character you have to own, if you decide to make these fields your calling, and he’s surprisingly thoughtful and yet practical about it.I thought it would be interesting to begin to think about the same questions with respect to entering the field of Quantitative Finance, particularly from a practitioner’s point of view.
According to Zvi Bodie, financial engineering is the application of science-based mathematical models to decisions about saving, investing, borrowing, lending, and managing risk. I think that’s a reasonable definition.
Science – mechanics, electrodynamics, molecular biology, etc., – seeks to discover the fundamental principles that describe the world, and is usually reductive and analytic. Engineering is about using those principles, constructively and synthetically, for a purpose. Thus, mechanical engineering is concerned with building devices based on Newton’s laws, suitably combined with heuristic or empirical rules about more complex forces (friction, for example) that are too difficult to derive from first principles. Electrical engineering is the study of how to create useful electrical devices based on Maxwell’s equations and solid-state physics, combined with similar heuristics. Similarly, bio-engineering is the art of building prosthetics and other biologically active devices based on the principles of biochemistry, physiology and molecular biology.
So what is financial engineering? In a logically consistent world, financial engineering should be layered above a solid base of financial science. Financial engineering would be the study of how to create functional financial devices – convertible bonds, warrants, synthetic CDOs, etc. – that perform in desired ways, not just at expiration, but throughout their lifetime. That’s what Black-Scholes does – it tells you, under certain assumptions, how to engineer a perfect option from stock and bonds.
But what exactly is financial science?
Canonical financial engineering or quantitative finance rests upon the science of Brownian motion and other idealizations that, while they capture some of the essential features of uncertainty, are not finally very accurate descriptions of the characteristic behavior of financial objects. (You should perhaps even object to my use of the word ‘characteristic’ since it’s not clear that financial markets even have time-invariant characteristics.) Markets are filled with anomalies that disagree with standard theories. Stock evolution, to take just one of many examples, isn’t Brownian. We don’t really know what describes its motion. Maybe we never will. And when we try to model stochastic volatility, it’s an order of magnitude vaguer.
So, the point I want to make, for those people who consider coming into the field from one of the hard sciences, is that financial engineering rests on a shaky basis. That’s not to say that it isn’t worth doing. In one sense it makes it more interesting. If you’re going to work in this field, you have to understand that you’re not doing classical science at all, and that the classical scientific approach doesn’t have the unimpeachable value it has in the hard sciences. You have to ask yourself if you can live with that.
The Value Of What You Do Is Often Not Deep
A personal story. In 1985, bond options were the hot new product, the synthetic CDOs of the era. At that time most trading desks used Black-Scholes style models for bond options, in which each bond was treated as a weird kind of stock. Our major theoretical problem was how to consistently model the yield curve and all the bonds that defined them in unison, so that we could then value options on them. Eventually we came up with BDT, which, while it had its problems, was at least self-consistent and perhaps more usable than what came before.
But, surprisingly, it was the new user interface to the model, that I built myself in those primitive pre-mouse days of screen-based programs, that had the biggest impact. It worked well because I made it do what the traders needed, which I learned from working with them. My new version saved them countless keystrokes compared to the command-line interface that drove the previous FORTRAN version. All my model’s input and output were visible on one screen. There were also fields for storing information about the client and the trade. And best of all, you could save a possible trade under discussion and retrieve it the next day for continued development and discussion.
Though primitive by today’s standards, this interface was astonishingly better than what the desk had used before, and the traders and salespeople were overjoyed. By creating and saving the most common types of option trades as templates in files at the start of each day, they could respond rapidly to clients, accommodating many more requests much more efficiently.
Ever since then, it’s been impossible for me to overlook the difference a simple and well-designed piece of software can make to a business, no matter how good or bad the model underneath.
If you work in this field, then, despite the genuine glories of quantitative finance, you may have to face the fact that you can have the most dramatic effect by improving the ergonomics of trading and sales. Is that fact something you can live with?
Crude and Approximate But Often Useful
Financial Engineering is a multidisciplinary field. It involves financial knowledge, business knowledge, mathematics, statistics, and very importantly, computation, because there’s little you can achieve without computation. There are very few analytic solutions that apply to the markets and products you actually deal with, so you must approximate all the time, and decide what complexities to ignore.
Because of this you need experience to be genuinely useful, and so there are very few young geniuses in the field, unlike mathematics or chess. Experience and some wisdom is often necessary, because you’re dealing with people and their quirks, and a large part of it is a social science. Hard science assumes there is a stable world underlying the observed phenomenon; in social sciences that stability is much less obvious, perhaps even non-existent, because you’re playing with people and they keep changing the rules.
The methodology of quantitative finance is different from large parts of physics or chemistry or even biology. Financial models are crude, and are mostly analogies. They say something like “it may be useful to think that people value bonds by discounting them over the paths of all future short-term interest rates.” This isn’t true, but it’s just possibly useful. In contrast, in physics you can say that the quantum mechanical world behaves as though a particle really does take all future paths to its target, with interfering probabilities. This isn’t merely useful, it’s actually true.
So, if you work as a practitioner, you will have to live with the fact that you are going to have to make crude, false but hopefully useful approximations.
If you go to work in a big investment bank, you’ll soon discover that traders and salespeople order you about and often make more money but have less technical skills than you. That’s less true today, when more trading is technology and algorithm based, and when products are more complex, and when the buy side offers many different opportunities, but it’s still often the case.
Many practitioners or programmers gets weary after a while, and want to become “one of them.” But they may not have the skill or more importantly the personality to do that. There are more opportunities these days, especially at hedge funds, but nevertheless I’ve seen many people get disillusioned by having to continue in their mainly technical role. Can you change to be what you want? Can you live with being who you are?
Part 2 to follow