2,649 reputation
514
bio website
location
age
visits member for 2 years, 5 months
seen 13 hours ago

Mar
25
comment Backtesting with fundamentals
Python is a full programming language so it has a lot of potential, but you'll probably have to roll up your sleeves and program some of the stuff you want. There is dedicated backtesting software out there, but you'd have to pay for it.
Mar
24
comment Backtesting with fundamentals
What analysis do you need to do that you can't do with numpy and pandas? Why don't you try going through the documentation and playing around with it a bit more?
Mar
21
comment Log returns vs Relativizing to Portfolio size of $1
It makes it difficult to recommend something without understanding the context. For instance, if I am performing mean-variance optimization including transaction costs I might think about them differently than if I were trying to evaluate how much transaction costs have impacted my portfolio in the past. In one case, I might think about transaction costs as a function of a change in weights or holdings, while in the other I might think in terms of dollars as a % of AUM. I'm not sure why you would analyze them in terms of log returns.
Mar
17
comment Log returns vs Relativizing to Portfolio size of $1
It might be useful if you add more details about what you're trying to do and what the various calculations are.
Mar
14
comment Use of geometric mean for average return of several indices
@ChrisDegnen What you're trying to do doesn't make much sense to me. You don't need a geometric average to get an average of returns cross-sectionally. Arithmetic average is fine. Weighted averages (based on market-cap or something) are also common.
Mar
13
comment Critique against consumption-based asset pricing theory?
Also, this theory isn't all that popular among practitioners. You'd get more information talking to some finance professors or something.
Mar
13
comment Critique against consumption-based asset pricing theory?
@Investor To your first point, advocates of this approach would likely say that it is an abstraction that wouldn't substantially impact the analysis. Others (like yourself) may disagree with that assessment, but oh well. To your second point, this is not agent based modelling. Representative agents are be more-or-less alike. Same utility functions, same preferences, same expectations. Maybe different ages in overlappying generation models. For the most part, all the agents are exposed to the same consumption goods in the same way.
Mar
12
comment Optimal lag length selection criterion in GARCH(p,q) model using MATLAB
Matlab's GARCH outputs the log-likelihood, which is the primary input to AIC/BIC. Just write a function that loops over the relevant parameters, calculate AIC/BIC, then selects the one with the best.
Mar
11
comment Econometrics - Granger Causality
davegiles.blogspot.com/2011/04/…
Mar
11
comment Optimal lag length selection criterion in GARCH(p,q) model using MATLAB
mathworks.com/help/econ/conduct-a-likelihood-ratio-test.html
Mar
5
comment Copula Value At Risk
I have no idea what those charts are supposed to be saying.
Mar
4
comment Copula Value At Risk
What do you mean you don't want a single value? You mean you want a distribution for the portfolio Value at Risk? Or, do you mean you want the quantile for each security? Also, I think Copula Value at Risk might be a misleading term. Sort of implies that the Value at Risk calculation is different, when it's really just the modelling that's different. Before setting what you want up with Copulas, I might first set it up with multivariate normal (as that is equivalent to what you're doing).
Mar
3
comment Beta and the Assumption of IID Returns
This answer would be improved by an explanation of what iid means and how it relates to each part.
Mar
3
comment Fitting a sigmoid function to incomplete, structured, data
I'm not entirely sure that sigmoid is optimal in this case, but I'd have to know about what you're trying to do to be sure.
Mar
1
comment How to calculate modeled asset volatility by industry factor?
Try using a factor model?
Feb
28
comment What is the necessary level of Econometrics-Know-How for a quant
Census X12-ARIMA is also pretty popular for de-seasonalizing also.
Feb
27
comment What is the necessary level of Econometrics-Know-How for a quant
@Probilitator I added it as an answer and tried to expand some of the reasoning a little.
Feb
27
comment What is the necessary level of Econometrics-Know-How for a quant
I think that's a good list (though I never really used wavelets myself). I would add missing, mixed frequency, and irregular data as some issues that I'm constantly either dealing with or begrudgingly ignoring. Seasonal adjustment is important too for some types of analysis (like electricity futures), but I might combine that with the ARMA stuff. I would say that the reason not to focus too much on Panel Modelling is that you'd probably get stuck trying to remember random or fixed effects when instead you should just ignore those and read Gelman's Bayesian Data Analysis.
Feb
27
comment Sharpe Ratio, annualized monthly returns vs annual returns vs annual rolling returns?
The best advice I can give is to just do whatever your boss wants. Annualizing monthly returns might be more common, but there's nothing wrong with using yearly returns to calculate Sharpe ratios. Just don't do the first method.
Feb
26
comment Sharpe Ratio, annualized monthly returns vs annual returns vs annual rolling returns?
What you refer to a yearly return, others might refer to as a rolling 1-year return or a year over year (YoY) return. I can't recall anyone ever doing this to calculate Sharpe ratio over a full sample. It is more common to annualize monthly returns. I'm also a little confused on what you want the output to be. Do you want the Sharpe ratio for each year, a rolling Sharpe ratio, or over the whole sample?