When I had decided on the MSc dissertation topic, my supervisor recommended to do it on Value at Risk (VaR). A risk management tool developed by JPMorgan based on confidence interval concept of statistics. She lent me few books so that I could read and familiarize myself.
It did not buy me, despite the fact that confidence interval subject is one of my favorites in statistics.
Arbitrage Pricing Theory, looked more challenging and interesting. I had to find a way to test the theory. Undefined beta coefficients found by cross sectional regression and then put into time series. It worked with many many regressions.
Value at risk is one of the concepts, highly being criticized during the current economic crisis for encouraging excess risk taking.