Friday, December 18, 2009

Public service announcements for quants

   1.  Conference on 'Computational Topics in Finance', February 19/20, 2010, National University of Singapore. The topics will include using R/Rmetrics in finance, but the conference is by no means confined to R. See http://www.rmetrics.org/.

   2.  Consulting position (6-month renewable contract) available at a major Canadian bank in Toronto:  research in  various mathematical algorithms used for pricing of interest rate derivative instruments like swaps, caps, swaptions, FRAs. Please contact their recruiter at http://www.linkedin.com/pub/kevin-p-w-wang/6/899/29a.

   3.  A free copy of Chapter 8 of "High Probability ETF Trading" which I mentioned here is now available for download.

11 comments:

Jez Liberty said...

Thanks for the links!

Singapore and Toronto are a bit too far for me for a conference and/or a job... But the free chapter will surely be interesting and am planning to use R for trading and automated system development so I am sure I'll find some useful material on the conference website.

Anonymous said...

High Probability ETF Trading is a deeply flawed book IMO. Win ratio is no way to evaluate a trading strategy and having independently backtested every strategy presented, most result in unbearable drawdowns at one time or another. Sure, you can delever to avoid wipeout but the returns become pitiful even without costs/slippage.

Ernie Chan said...

Anonymous,
Thanks for sharing your backtest results.
As I pointed out elsewhere, no one should trade a published strategy without their own improvements or revisions. Those strategies serve only as inspirations.
Ernie

Anonymous said...

hi,

I was wondering if there is a statistical test that exists for a pair of time series which says how many times a zero line is crossed.
I know cointegration tests find whether the pair don't stray to far from one another, but it doesn't say if the zero line is crossed few or many times.

Ernie Chan said...

Anonymous,
If you are just concerned with a finite time series, why not just count the number of times it crosses zero?

If you are concerned with a theoretical infinite mean-reverting time series, google "ornstein uhlenbeck first passage time", and the references therein will tell you how to estimate the time it takes to cross zero.

Ernie

Anonymous said...

Do you know how it would be possible to include the bid ask data when backtesting a system. I am not aware of any such software. Most are all based on price charts of last price.

Ernie Chan said...

Anonymous,
I believe tickdata.com will sell you historical bid-ask quotes for many markets.
Ernie

Anonymous said...

Hey Ernie if you are trading many stocks, like over a few hundred. Do you know of any news services that allow you to enter a portfolio of >500 stocks and have the ability to find your stocks that have news on them? I mean other than the usual google and yahoo stuff. I find they are always missing some news items on theportfolio page and they have a 200 symbol limit which makes it very annoying.

Ernie Chan said...

Hi Anonymous,
If you use Dow Jones news (available through Rediplus, RealTick, and other trading platforms), you can enter your portfolio of stocks into their watchlist. Otherwise, you can run a program to continuously scan and parse the news of each stock on yahoo or google one by one.
Ernie

habteabogbu said...

Hi Ernie,

I was inpired to get into an algo trading after I've read your book and thanks for sharing.

I am a quantitative programmer at an investment bank and planning to become a full time quantitative trader using my own capital.

I'm taking a stochastic calculus course and wondering if I need any further math cources?

Thanks

Ernie Chan said...

Hi habteabogbu,
Actually, you don't need stochastic calculus to trade algorithmically. In fact, no math beyond basic college statistics is necessary.
Ernie