Quant Finance Research

[pokerguy]

Hi, previously I mentioned that I am final year of an MBA, but was lamenting that I did not enroll in a Quant Finance masters instead of the MBA. The Dean of my school put me in touch with a operations management professor with a PhD in Operations Research and they have agreed to allow me to do two research projects that would allow me to explore quant finance. I need some help from the QN community as I I would like some suggestions, not what to research, but what Quant Finance methods to include in the projects. The first project is meant to provide a base to get much more technical in the 2nd one. I know this is no substitute for a MFE, but the faculty at my school has been pretty cool in letting me run with this so any help will be greatly appreciated.

 

[Ken Abbott]

How about quantitative techniques utilized by regulators as part of the Basle rules? These would include:

• VaR Models (HS, V/CV, Monte Carlo)
• The specific risk component of VaR models
• Gaussian copula-based techniques for counterparty credit risk measurement
• the Incremental Risk Calculation that is part of Basle 2.5 (single-factor Vasicek + idiosyncratic component)
• Internal Ratings-Based credit risk calculations (IRB)
• The Comprehensive Risk measure (tail risk measurement of credit derivative portfolios)
• The Advanced Measurement Approach to Operational Risk (AMA)

This would both a) provide an interesting project and b) make you employable.

 

[Ian Kaplan]

Ken has much more experience than I do, so I hesitate to suggest an addition, but I will do it anyway.

To the VaR models, I would add Conditional Value at Risk (CVaR - also known as Expected Tail Loss or ETL). Although Basil doesn't specify CVaR, it may be that they should.

Say you have a bank with a number of areas that incur risk (or, these days, a fund). You would like to get a picture of the overall risk of the organization. Each area has their own VaR estimate. The natural inclination is to try to sum up the VaR for each group to arrive at the overall VaR of the organization. However, VaR is not additive. However, CVaR can be summed. Also, CVaR can give a more accurate picture of risk.

 

[Ken Abbott]

Ken has much more experience than I do, so I hesitate to suggest an addition, but I will do it anyway.

To the VaR models, I would add Conditional Value at Risk (CVaR - also known as Expected Tail Loss or ETL). Although Basil doesn't specify CVaR, it may be that they should.

Say you have a bank with a number of areas that incur risk (or, these days, a fund). You would like to get a picture of the overall risk of the organization. Each area has their own VaR estimate. The natural inclination is to try to sum up the VaR for each group to arrive at the overall VaR of the organization. However, VaR is not additive. However, CVaR can be summed. Also, CVaR can give a more accurate picture of risk.

Good point. Expected Shortfall is part of Basle's Fundamental Review of the Trading Book. It's worth a look. (I had some of my stdents explore its properties as part of their final project last semester..)

 

[pokerguy]

Sorry for the delay in responding, thanks for the input Ken and Ian but after a few meetings with my professor, we have settled on developing pricing models for weather derivatives as the focus on the 1st research project. Does anyone know of any good books regarding weather derivatives? There are a few on Amazon, but if anyone is aware of "the best" or "most definitive" book on the subject, your opinion would be greatly appreciated.