that paper is available at: http://www.creditflux.com/public/public ... xGuide.pdf

<t>Coming from a lower tier university is a hinderance, but overall won't affect you getting into the field if you are willing to do a bit more work. The main advantage from graduating from a top school is that more companies come to campus to recruit so you're able to get a lot of interviews with v...

I would recommend reading the manual on the particular package you want to use. They're very straight forward and contain alot of examples you can follow along with. Additionally, you can call mathworks to set up a free demonstration over the web and they'll walk you through what you want

I may not be understanding you entirely, but why not just use a log returns series and forget about inflation all together?

If you want to to your graduate work in math finance, econometrics isn't really a requirement. I would focus more on finishing your linear algerbra, pde's, ode's, probablitlity and Calculus (all levels available at you college) or skip the calc and take courses in Analysis.

there's no immediate answer to a question like this. it depends on your risk preference (so-called lambda). try looking at the sharpe ratio to get a more standardized reference measure

<t>I don't agree with taking the GRE subject test. This will certainly prove that you have the necessary math background. However, there are things on that exam that you will never need to know in qf and also won't get exposure to unless you take higher level math courses which will make for a full ...

I sent you a PM

It is a test for cointegration... if you find that the residuals from the cointegration regression exhibit unit root behaviour (from DF), that means the price series are cointegrated. You can also look at the partial autocorr. function to get an idea of the behaviour.

<t>I'm not sure why you're referring to generating the DF stat as a step preceeding the test for cointegration. DF is a test for cointegration... not something preceeding it.I've got stacks and stacks of info on this since I'm implementing something very similar. If you want any particular papers, P...

<t>dynamically forecast means change the forecast with time.So doing as you said ([4]+[3]=next term) seems correct to me. If you have 300 observations and you're expected to forecast to the end and compare, then I would certainly suggest writing a small program to do this. Shouldn't be too time cons...

For sure. Work a couple years and then apply. Average age of a typical MBA class is usually around 28. In two years, you'll only be a bit over the average.. far from a huge deal.

<t>I think it depends on the kind of trading you want to do. Most traders are supported by quants, the people who create the models. In my experience these are the people that graduate with MFE's and often end up supporting senior quants who have PHd's. Typically they fall into a more programming ty...

I need to poll real time data every x seconds, but I'm not sure how to time in splus. Is there a way to do something like:current.time<-get.current.time()and from there I can figure out whether or not enough time has elapsed?

<t>I'm not sure about the columbia program specifically, but even if it's optional, I would write it. Recieving an application without supporting GRE scores would probably ring some alarm bells in people on the admission comittee. I would imagine the vast majority of people submit them. Also, if you...