My name is Paul Borochin. I have taught graduate and undergraduate courses in portfolio and risk management, financial modeling, mathematics of financial derivatives, and investments. This channel is dedicated to exploring and teaching the techniques of quantitative finance.
My academic research draws on information from the public markets to provide quantitative insights on future equity and option performance, identify expected value effects of corporate events, and predict policy decisions. I also study the effects of institutional ownership and common ownership linkages between firms for different types of institutional owners defined by their investment styles. I am also interested in financial applications of machine learning. My research has been covered in posts by the CFA Institute's Enterprising Investor blog, the Harvard Law School Forum on Corporate Governance and Financial Regulation, and the Columbia Law School's Blue Sky blog on corporations and capital markets.
Hey, great video, was just wondering why prefer statsmodel over scikitlearn here? I use both. I mean, statsmodel might seem more intuitive coming from R, but i think scikitlearn doesn't exactly lack anything that the statsmodel can do. Anyways, great video!
Im carrying out an assignment on portfolio allocation and this video has been immensely helpful. It made simple the calculations necessary to estimate a risk aversion coefficient and risk premium!
The derivative dr_a / dsigma_a at a=0, reflects addition of a very small amount of asset i to the portfolio (and deducing a very small amount of M from the portfolio as well), instead of the addition of the total market M (the exposure to new small risk disgma_a, even at a=0, is done by increasing a alittle). Hence, the new portfolio is not on the CML. So why would this derivative be equal to the sharpe ratio of M? I believe one should add an equilibrium statement as follows: Assuming a person in the CAPM world holds the tangency portfolio M, and considers whether to add to his portfolio a very small amount of asset i, or just another copies of M. At this position, by equilibrium consideration, the investor would buy asset i, only if the risk-return of this small addition to his portfolio M, would lead to the same sharpe ratio he could have been obtained by simply buying more stocks of M. Otherwise, the investor won't have the incentive to buy a small amount of asset i, and prices of asset i would fall. Hence, the equation is correct. Am I right?
I m comfortable with the process and the coding, but struggle to understand the motif of the results. For example, line 1, APP_RF -18... what that means ?
Thank you for this series, really helped summarise a lot of work. Have you done or will you do a series/video on the consumption-based capital asset pricing model?
The idea is that all assets earn the risk-free rate plus the risk premium (the good part), but they also incur risk (the bad part) - the utility function attempts to relate these.
@@TechFin Thanks for your explanation. It's just that the utility formula in Bodie et al. does not include the rf and the addition of rf in the utility formula does not affect the weights allocated to each asset when using solver.
Great video! I have an odd question, what software do you use to record this video? I love that format! I really like how you have the camera with yourself cut out.
@@TechFin thanks! I use OBS too but had trouble removing my background. I got it to remove the background, however it was too laggy to stream. Perhaps if I recorded two separate videos and blended them together later.
Hello sir, thank you so much for theses explanations. Could you please make a video on the enstimation of the Liquidity adjusted CAPM using the Fama MacBeth regression. Thank you so much.
It’s interesting study and as a student I always wanted the Professor to show us step by step while following along instead of just reviewing their work.. that way students could go on their own and perform their own research
Hey, really important question. Im searching for this all over the internet but i dont get clear answers. Everyone is explaining markowitz with the two asset model. Its easier to calculate i get it but how exactly do i calculate the efficiency curve with n > 2 assets? Can you help me there. Would be amazing thx
Great video! My background is neither in CS or finance, but I have an interest in both. Despite that, I was able to follow and understand most parts of your walk-through. Is there somewhere I can read more about the risk-factors to understand what exactly we are measuring here? Also a place where I can read more about the output of the model, i.e. what the different outputs mean (for example Durbin-Watson, or R-squared that you talked a little bit about)? Not sure if my question makes any sense, like i said i'm not well versed in the world of financial models and metrics, so excuse my potential ignorance!