I applied for a job a while ago and just got a call back for an interview. The qualifications section of the application listed about ten points that the firm is looking for. I met nine of the ten but have no experience with “equity risk models and portfolio optimization.” As a level 3 candidate (awaiting results from June), I know what those terms mean, but I don’t have any relevant experience with them. The job is an Assistant PM at a firm that uses quant based allocation strategies. Could someone with relevant experience please point me toward some resources? For ex., what are the commonly used risk models? What software are most firms using for optimization? Can I download the software and teach myself in a weekend? etc. Thanks.
Northfield has optimization software as well as a risk model. Probably cannot download the software. Pretty sure it is quite expensive. Search Northfield. Axioma has one as well I believe.
Thanks, Rydex. Do you have any experience with that kind of software program? Is it easy to use and learn?
I would suggest using Matlab if you decide to play with utility functions, develop your own risk models and perform statistical analysis.
Northfield is easy to use but you shouldn’t worry about that. Your answer in the interview is “I am familiar with lots of the issues with portfolio optimization but I don’t have practical experience with any software”. At an interview, nobody wants to check off the box that says you have used Northfield (which is a little childish anyway IMHO). They want to talk to you about portfolio optimization. You have a big start with the CFA stuff. Presumably, you know about Treynor-Black (he-he), Markowitz, mean-variance, frontier, etc… That’s a huge amount and more than most people in finance know. The practical things you should know are: a) Instability of estimates - All these things fail because you are optimizing on projections that are likely to be incorrect. You want a model that incorporates market data (option data would be great but nobody does that) and is robust to your errors in projecting means, variances, correlations, etc… b) Liquidity - PM’s always scoff at people who say “You need to buy another 10000 shares because my optimization engne says so” c) Mathematically the optimization is easy. Portfolio optimization rarely requires much more than some pretty simple math - in theory you could write any portfolio optimizer in MatLab or Mathematica pretty easily. d) Risk factors - Programs like Northfield work on factors (Automobile, Gaming, Electronics, etc) and manage portfolios based on the factors not the securities (sort-of). That means there is the step of mapping securities to factors. You gain stability from factors but error from mapping. Expand on that kind of stuff in your interview and you will be fine.
maratikus Wrote: ------------------------------------------------------- > I would suggest using Matlab if you decide to play > with utility functions, develop your own risk > models and perform statistical analysis. Which brings up the point that risk management and portfolio optimization are blood brothers.
Great post, Joey, thanks. That’s encouraging. Joey (or anyone else): I’ve posted a portion of the description below. How much math do you think is involved? I am pretty sure that I can do this job having cleared at least 2/3rds of the CFA and having 2+ years of relevant experience (equity research and portfolio management), although I would note that my BA is not in a quantitative subject. I am not afraid of math, but I would hardly call myself a quant genius, either. Also, do you have an opinion about how relevant this position would be for my career given that I would like to become an institutional PM at some point? This is a beta approach to investing, whereas my experience has had more of alpha orientation (i.e., try to pick outperforming stocks and sectors) so I think this job would be helpful in rounding out my perspective. The company focuses on building “structured portfolios” that minimize cost and risk under the following assumptions: Equity markets are reasonably efficient—out-performance is difficult and hard-won. Taxes matter—they can be the biggest cost a private investor faces. Private clients have unique needs—and may benefit from a customized portfolio. Duties and Responsibilities: Construct optimized equity index portfolios for separate accounts Rebalance and transition portfolios to maximize tax-efficiency Construct custom benchmark data Generate after-tax return data for both portfolios and benchmarks Performance attribution Qualifications and Requirements: Strong mathematical and analytical skills Familiarity with academic financial theory and applications Working knowledge of equity risk models and portfolio optimizer Analytical writing ability Bachelor’s (4yr) degree w/ concentration in finance, accounting, engineering or related quantitative focus Thanks for all helpful answers.
JoeyDVivre Wrote: ------------------------------------------------------- > maratikus Wrote: > -------------------------------------------------- > ----- > > I would suggest using Matlab if you decide to > play > > with utility functions, develop your own risk > > models and perform statistical analysis. > > Which brings up the point that risk management and > portfolio optimization are blood brothers. true
JoeyDVivre Wrote: ------------------------------------------------------- > Which brings up the point that risk management and > portfolio optimization are blood brothers. amen.
bump – any thoughts JDV? bromion Wrote: ------------------------------------------------------- > Great post, Joey, thanks. That’s encouraging. > > Joey (or anyone else): I’ve posted a portion of > the description below. How much math do you think > is involved? I am pretty sure that I can do this > job having cleared at least 2/3rds of the CFA and > having 2+ years of relevant experience (equity > research and portfolio management), although I > would note that my BA is not in a quantitative > subject. I am not afraid of math, but I would > hardly call myself a quant genius, either. Also, > do you have an opinion about how relevant this > position would be for my career given that I would > like to become an institutional PM at some point? > This is a beta approach to investing, whereas my > experience has had more of alpha orientation > (i.e., try to pick outperforming stocks and > sectors) so I think this job would be helpful in > rounding out my perspective. > > The company focuses on building “structured > portfolios” that minimize cost and risk under the > following assumptions: > > Equity markets are reasonably > efficient—out-performance is difficult and > hard-won. > Taxes matter—they can be the biggest cost a > private investor faces. > Private clients have unique needs—and may benefit > from a customized portfolio. > > Duties and Responsibilities: > Construct optimized equity index portfolios for > separate accounts > Rebalance and transition portfolios to maximize > tax-efficiency > Construct custom benchmark data > Generate after-tax return data for both portfolios > and benchmarks > Performance attribution > > Qualifications and Requirements: > Strong mathematical and analytical skills > Familiarity with academic financial theory and > applications > Working knowledge of equity risk models and > portfolio optimizer > Analytical writing ability > Bachelor’s (4yr) degree w/ concentration in > finance, accounting, engineering or related > quantitative focus > > > Thanks for all helpful answers.
Later today…
Check out the MSCIBarra multi-factor risk models. We are specialized in factor risk modeling. It is widely used in the industry.
You could probably also could download CPLEX opl studio. It allows you to solve small models on a trial basis. You probably won’t get asked this but one of the difficulties is that if your constraints on items such as minimum buys turns your makes your model have a mixed integer element to it.