Reputation
8,587
Next tag badge:
103/100 score
16/20 answers
Badges
23 64
Impact
~288k people reached

Aug
7
answered Will price levels fall even though money supply increases?
Aug
5
comment What is the difference between the methods for calculating VaR?
Expected Shortfall also has the benefit of being a "coherent" risk measure, unlike VaR. Most of the recent literature in portfolio construction and optimization tends to use Expected Shortfall (or equivalently Conditional Value at Risk)
Aug
4
comment Portfolio optimization with monte carlo sampling from predictive distribution
It is purely empirical and cannot be modeled analytically.
Aug
3
revised Portfolio optimization with monte carlo sampling from predictive distribution
added 103 characters in body
Aug
3
revised Portfolio optimization with monte carlo sampling from predictive distribution
added 20 characters in body
Aug
3
asked Portfolio optimization with monte carlo sampling from predictive distribution
Aug
3
comment robust portfolio optimization re-balancing with transaction costs
Brian, 1) Can the PortfolioAnalytics function handle custom constraints (example: sector-neutral weights)? The constraint class looks pre-specified. 2) Also, I see that the optimize.portfolio function can accept a matrix of returns. I have a posterior distribution of expected market returns (i.e. K possible realizations of market returns for N assets resulting from MCMC). Can this function identify the optimal weight vector with respect to the sampling from the posterior? Of course, I would continue to use covariance of historical returns (de-noised) for calculation of ES. Thanks!
Aug
2
comment How do I graphically represent the evolution of a covariance matrix over time?
Those charts look great - thanks for posting! This seems like an interesting tool for identifying regime change points.
Aug
2
answered Efficiency vs. Robustness - To use a constant or not in single factor time-series regression?
Aug
2
comment How do I graphically represent the evolution of a covariance matrix over time?
Cool! I updated the answer. I'm not sure if you are able to share/upload that output but it would be very intellectually interesting to look at!
Aug
2
revised How do I graphically represent the evolution of a covariance matrix over time?
added 540 characters in body
Aug
2
comment How do I graphically represent the evolution of a covariance matrix over time?
This may be a bit abstract but here goes. Create a table: rows represent dates, and columns represent eigenvectors. The entries of the table represent changes in the angle of the eigenvector from the previous row. This will show how stable your covariance structure is. You can also create a second table this time with eigenvalues as the columns sorted from high to low (and the corresponding values below for each date).
Aug
1
comment What methods do you use to improve expected return estimates when constructing a portfolio in a mean-variance framework?
Michaud's re-sampling approach has perverse behavior (namely assigns higher weights to assets that high volatility in some circumstances). See Bernd Scherer's "Robust Portfolio Optimization", or Martin/Scherer's text for the details. There is also no theoretical basis for re-sampling. For this reason I would avoid re-sampling.
Aug
1
answered How do I graphically represent the evolution of a covariance matrix over time?
Jul
29
answered What research is available on the performance of convertible bond arbitrage models?
Jul
29
revised What methods do I need to learn in order forecast asset price movements?
added 1 characters in body
Jul
28
revised What methods do I need to learn in order forecast asset price movements?
added 256 characters in body
Jul
28
answered What methods do I need to learn in order forecast asset price movements?
Jul
27
comment robust portfolio optimization re-balancing with transaction costs
The links and example in powerful is great - thanks! Your choice of optimizer seems general enough to solve the optimization problem while also referencing current weights. I believe I can pass an additional penalty to the objective function that measures the distance from the current weight vector to the weight vector the optimizer is solving for time a constant (e.g. transactions costs). Since your optimizer can take expected return as an input I expect it could take a current weight vector as well. This would enable me to optimize in the presence of transaction costs.
Jul
25
answered Most successful investors using academic-based framework?