Serving the Quantitative Finance Community

Search found 42 matches

by trendkiller
November 19th, 2010, 12:00 pm
Forum: Trading Forum
Topic: A tale of two betas
Replies: 1
Views: 23075

A tale of two betas

<t>Consider the following situation:-Beta is measured using weekly returns over a well diversified market index.-A long-only equity strategy selects a porfolio of stocks each week with a holding period of exactly one week (It may be the case that the strategy selects no stocks, so the strategy remai...
by trendkiller
November 19th, 2010, 11:23 am
Forum: Trading Forum
Topic: Dangers of Short Selling
Replies: 28
Views: 29205

Dangers of Short Selling

Daniel Drew put it nicely about 150 years ago:"He who sells what isn't his'n, Must buy it back or go to pris'n"
by trendkiller
September 13th, 2010, 9:40 am
Forum: Technical Forum
Topic: Vasicek Mean Reversion Parameter
Replies: 12
Views: 31613

Vasicek Mean Reversion Parameter

QuoteOriginally posted by: EzraOne typically uses 1/24 = 4.1667%. Don't ask me why -- if I find out the reasoning, I'll post here.Thanks, I'll be interested in knowing why.That quantity is also the equivalent of 1 karat, or 1 24th of pure gold
by trendkiller
September 2nd, 2010, 4:59 pm
Forum: Technical Forum
Topic: Vasicek Mean Reversion Parameter
Replies: 12
Views: 31613

Vasicek Mean Reversion Parameter

Thank you all for your help.
by trendkiller
September 2nd, 2010, 12:13 pm
Forum: Technical Forum
Topic: Vasicek Mean Reversion Parameter
Replies: 12
Views: 31613

Vasicek Mean Reversion Parameter

Actually in the model delta_rt = a*(long_run_avg - rt) + sigma*delta_z,I was asking about mean reversion the parameter "a," not the long run average interest rate. I don't think the sample average rate is a good estimation of that.
by trendkiller
September 2nd, 2010, 11:54 am
Forum: Technical Forum
Topic: Vasicek Mean Reversion Parameter
Replies: 12
Views: 31613

Vasicek Mean Reversion Parameter

Hello,Do you mean the sample average of interest rates? What would be the reasoning behind such an estimation?Thanks
by trendkiller
September 2nd, 2010, 11:10 am
Forum: Technical Forum
Topic: Vasicek Mean Reversion Parameter
Replies: 12
Views: 31613

Vasicek Mean Reversion Parameter

<t>Hello,I need to make interest rate simulations using the Vasicek model. My simulations will have a time step size of 1 year, and I lack sufficient yearly data to estimate model parameters properly. Specifically, I am having trouble assigning a value for the mean reversion parameter a. I was wonde...
by trendkiller
March 4th, 2010, 6:32 pm
Forum: Trading Forum
Topic: Constructing a pairs trade
Replies: 19
Views: 50180

Constructing a pairs trade

<t>Marine & eppob: Thank you for your comments.To Marine: I implemented my model on Matlab. I thought it was pretty standard practice to run the ADF test on the residuals of the (cointegration) regression to check for stationarity. I will look into spacial cointegration when I have the time.To e...
by trendkiller
February 28th, 2010, 11:23 am
Forum: Trading Forum
Topic: Constructing a pairs trade
Replies: 19
Views: 50180

Constructing a pairs trade

<t>I've been fiddling with my pair trading model, and I noticed the following curious phenomenon:Considering both the stock pair A/B and B/A, it often turns out that A/B seems to be cointegrated while B/A is not BUT trading signals generated from B/A outperform signals generated from A/B. Similarly,...
by trendkiller
February 28th, 2010, 5:46 am
Forum: General Forum
Topic: Optimization problem on two trading strategies
Replies: 9
Views: 33156

Optimization problem on two trading strategies

QuoteOriginally posted by: AnthisThus, unless outrun and trendkiller object, we expect our royalties corresponding to our fair share of profits. I second that! Especially nowadays when my boss keeps mentioning a monkey could trade better than I do
by trendkiller
February 19th, 2010, 6:21 pm
Forum: General Forum
Topic: Optimization problem on two trading strategies
Replies: 9
Views: 33156

Optimization problem on two trading strategies

<t>I would check for autocorrelation for each trading strategy as well. It is difficult to forecast future returns of DAX, but it may be easy to forecast future returns of your DAX trading strategy. If you detect strong autocorrelation in your trading returns, I'd go about fitting a simple linear mo...
by trendkiller
January 14th, 2010, 12:54 pm
Forum: Careers Forum
Topic: A question about risk management
Replies: 21
Views: 36002

A question about risk management

<t>You should be fine with your standard curriculum material from your Master's degree...I'm pretty sure that your interviewers would like to hear the acronym "garch," no matter how funny it sounds The important thing for you to realize is that the chances of applying the stuff you learn at school a...
by trendkiller
January 14th, 2010, 5:57 am
Forum: Careers Forum
Topic: A question about risk management
Replies: 21
Views: 36002

A question about risk management

<t>The skills that will impress your interviewers are different from those skills that you will use as a junior in the risk management department. You will most likely be dealing with day to day reporting for a (possibly long) while. So, the core skills would be MS Office skills for such a position ...
by trendkiller
December 30th, 2009, 5:54 pm
Forum: General Forum
Topic: bonuses: why?
Replies: 10
Views: 34214

bonuses: why?

<t>Incentive should be a factor. If the bonus is a meaningful amount, it may keep the employee at the company until the year's end. A suspecting employer will have some extra time to start looking for replacements. This can be important especially if multiple employee's at a single division decide t...
by trendkiller
December 23rd, 2009, 8:30 pm
Forum: Trading Forum
Topic: Constructing a pairs trade
Replies: 19
Views: 50180

Constructing a pairs trade

<t>QuoteOriginally posted by: acastaldoQuoteWill a simple linear regression provide the correct weightings? Here is a riddle for you: which regression are you going to do, x on y or y on x. The results will be different...People generally advise choosing the independent variable such that the standa...