Most of the phds that seem to post in this forum seem to come from physics and maths,but i am aware that electrical engineers sometimes follow this path...i'd love to learn a bit more about the quant world from an EE's perspective.in a couple of months of will have a phd in electrical engineering (signal processing and source coding/compression)on the side i have undertaken a small quant finance research project with a Prof from the finance facultyi'm interested and very eager to learn more:(1) what aspects of EE pop up in quant work?(2) is it common for EE phds to make such a transition?(3) do EE's generally make good quants? i could imagine that their maths wouldnt be as good as a maths phd (obviously) but maybe better software and practical problem solving skills?(4) it would be good just to hear about the experiences of people with a similar background..thanks in advance!

I am nominally an EE PhD, with a specialty in stochastic control. I too have left EE for quant-fin, and I am still undergoing the transition. Bearing that in mind, here are some points:1) Stochastic control shares some of the same methods of quant finance, if studied on the level of e.g. Fleming and Rishel's book (see Amazon). As for the transition from control to quant-fin, Merton (of BSM fame) was an optimal controls guy in undergrad, and a couple of the FE-faculty at Stanford are also of a control background.2) I realize the above might not answer your question, as it sounds like you are not a control guy. Well, truly few aspects of EE pop up in quant work (sorry). The reason is that in signal processing, basic control and communication curricula one assumes that the noise is additive and not multiplicative, an assumption routinely violated even in the most basic BSM formula. Another difference is that quant-fin models are formulated -- indeed approximated -- as continuous-time systems (beloved by physicists), and then discretized for numeric solution; quant modelling skills are really needed in the first step, while the second step is largely an exercise in numerics. Thus signal processing-type skills are of little help in modelling: EEs (as such) make great quant-programmers not great quant modellers. Additionaly, I can see signal processing skills as a help in trading software (again a distinct task from modelling).3) However the basic skills learned in EE are a good springboard for studying a new animal, stochastic diff-eqs, which is really what you need to do if you're gonna model (or at least understand existing models).Hope I helped, and good luck.

I am in a similar position, I am graduating soon with an EE PhD (Data Converters and circuit design). I would like to go into Quants upon graduation, particularly Model Validation/Development or other research oriented roles. However, with my lack of relevant experience, I presume going through one of the investment banks training programs is the best route! I have already been through one investment banks graduate training program in the past and I see little point in repeating the exercise. Is there an alternative to this?

Just a thought, why not try and get into a quant role at a commodity trading house(or utility) that trades electricity. I know the industry isn't getting the best rap at the moment post Enron, but it is still around. It may help you get the edge by you abilty to undestand the physcial commodity better than applicants from other fields.

I am an EE MSc and am doing my PhD in Finance and working as a Quant in a banka lot of what I am doing is related to EE (Optimization, Filtering ...)

I'm an MSEE with a BS Math/Physics. Electrical Engineering is a wide field ranging from chip design to power engineering and building electronics. When I got out of school I worked at one of the firms that did Kalman Filtering and had something to do with "Applied Optimal Estimation." Most of the people in the firm were advanced degreed scientists/engineers with a heavy analytical (math) bent. Programmers were not even considered members of the technical staff. If your EE is in stochastic signal processing as arises in communications and control theory then the techniques for handling time series cross over in a multi-disciplinary sense. If you do target tracking (another time series) or radar and image/video processing and understanding then you get exposure to non-parametric techniques as well as the rigorous error bounded estimates for prediction and pattern recognition that you would expect and the Bayesian people insist on. I've done Times Series (Hamilton) as well as Asset Pricing (one of the courses LO, MckInlay, and Campbell's book was gleaned from) for graduate courses. The economists certainly look at things differently. EE is not really a good basis for economics (models of consumption still baffle me.) Keep in mind that for example doing undersea signal processing, noise is not additive or white and all of those assumptions lead to poor accuracy in estimates. Non-linear, non-stationary, stochastic volatility is more close to reality in that environment and solutions quickly become intractable. I'm not aware of applications of Ito's law and Levy processes with jumps for characterizing signals or target tracks. Particle filters and sequential Monte Carlo methods are important for tracking. Similarly, you usually don't get to generate and solve BS PDE's obviously. The computational techniques (numerical methods, Monte Carlo etc.) cross over however.I have had the pleasure of doing some bond pricing analytics and system development for high yield and convertibles and continue to do product development but I'm not sure that I would be quick enough with VB and Excel to survive as a desk quant! I can think of several of my former colleagues who now work in stat arb for hedge funds or work in NY at major banks as quants.Apologies for the rambling...

Keep in mind that for example doing undersea signal processing, noise is not additive or white and all of those assumptions lead to poor accuracy in estimates. Non-linear, non-stationary, stochastic volatility is more close to reality in that environment and solutions quickly become intractable. Now you know why all those stochastic-volatility, jump diffusion calculators ain't worth squat. Welcometo second-generation finance.

Well, just to add to the collection of specialities - I have just finished (defended in January) my phd in Bioinformatics. My dissertation was about exceptions in protein sequences from complexity point of view, and now I am doing IR derivatives. So, it seems there are very few people who came to finance from finance