Our next Faculty Spotlight of the semester is on Dr. Alan Chernoff, Visiting Instructor, Economics. Dr. Chernoff joined TCNJ’s School of Business in the Fall of 2022. He has his BA in Mathematics, and MA in Economics, and, in 2024, earned his PhD in Economics, all from Rutgers University. In 2024, he won two awards: “Best Paper” at the 2024 NABET Conference, and “Best Paper in the Crypto/AI Category” at the ABR Fall 2024 Conference.
Dr. Chernoff is currently researching, “Estimating Integrated Volatility via Combination.” Recently, we caught up with him to learn more about this research.
Can you share with us a little about this topic?
Volatility as a variable is really important for stock and derivative pricing and can be really difficult to measure. In recent years there have been a lot of new ways to measure volatility, which initially sounds like a good thing. The problem is, that there are now so many ways to measure it, that it can be difficult to decide which one to choose. So in my research, I decided to take a look at if you can combine some of the many volatility measures to get a more useful measurement method. The combination could be as simple as the mean of different measures, or using machine learning to find a more optimal combination.
What drew you to this topic?
Prior to graduate school, I worked as an economist and submitted forecasts to Bloomberg. One of the things I always noticed was that the mean forecast always seemed to outperform any forecaster’s individual submission. When I learned there were a large number of ways of measuring volatility, I wondered if the same principle could apply here as well, ie, that the mean could outperform any individual measure.
Have you encountered any challenges in conducting this research?
The hardest part for me was the coding. Most of the papers that introduce them are very mathematically rigorous, and it was a lot of work translating the formulas into code. I also had to simulate my own data which was so computationally intensive that my own laptop couldn’t handle it. Fortunately, I was able to use the computer in my office at TCNJ.
What have you found the most interesting to come out of this research so far?
The implications of using machine learning in contexts like this are really interesting. Since we have so many ways to measure volatility, we’re essentially generating our own data sets, and then reducing the dimensionality of what we just generated to get the optimal measurement. I think that removing some of the subjectivity behind the selection criteria in these kinds of measurements is really neat!
Any other information you would like to share about this research?
Currently, I’m looking to implement a trading strategy based on the relation between stock price and volatility measures. I have some simple ones, but if anyone has a clever idea for a trading strategy that they’d like to test in that area, I’d love to hear it!
Read more about Dr. Chernoff and see what other things he has been working on.