Monday, February 14, 2011

Day One - Knowledge is Power

So today I begin my quest to amass all the knowledge I can relating to Quantitative Trading. Unfortunately for me most of the material is written for markets outside of Australia. However, on first impressions it would appear the principles to Quantitative Trading - and more importantly - building a Quantitative Trading Platform, can be applied to the Australian market as well, albeit with more work.

Here is a list of material I am currently reading:

Other than reading books and papers I have also commenced setting up infrastructure that will run the platform. This is currently made up of:

  • A dual core Linux server w/ 4 GB RAM with solid state disk for data (lives under my house)
  • A dual core Windows desktop w/ 6 GB RAM (when I am home) with 3 x 24" monitors
  • A lightweight laptop (when I am not home)
  • An Internet router with VPN software (so I can access my server remotely)
  • A high speed cable Internet connection that links to my Internet router
  • Gigabit network connections between all systems

To execute trades I signed up to an Australian online broker which will remain unnamed for the time being.

For charting and tracking my portfolio I use Insight Trader. This package was selected for general purpose stock trading. At this point I am unsure whether it will form part of my quantitative trading platform. As part of the package I purchased the database manager subscription. This costs me $154 p.a. and gives me access to adjusted stock prices. This however does not provide me ongoing end of day prices and must be sourced elsewhere. For this I rely on the online broker who offers end of day prices.

I realise now I may be using terms without defining what they mean, like end of day prices. I encourage you to search the web in context to stocks, if you are unsure of a particular term.

To save me from the daily burden of downloading prices from the web I have written a program that wakes up my UNIX server every weekday at market close and downloads the end of day file from the online broker. If for some reason the system skips a beat and fails to download prices it automatically detects this exception and reconciles the directory that stores all the end of day files - by date. In fact I have used this same process to automate the download of several years of end of day prices. The remaining years I purchased from the guys at Speculative Trading Ideas for a very reasonable price. They send it to you in the mail on CD. At this point I should offer a word of caution. If you are going to base your entire trading platform on someone else's data it would be wise to check it first. I have yet to go through the data supplied by STI. When I get around to it I will post my experience.

I hope this gives you some insight as to where I am at in this journey. It will no doubt be a long and eventful experience. I hope you enjoy reading about it. Here are a couple of things on my todo list:

  • Finish reading Ernie Chan's book
  • Select a database management system that will store stock prices that I will use for back testing (I am considering NoSQL)
  • Research numerical computing environments for UNIX
  • Develop and thoroughly test a program that can automate trades using my online brokers system. Unfortunately, they do not offer an API to the general public so I will need to develop a system that interfaces with the brokers web site (like an individual would). I have a couple of tricks up my sleeve which should make this easier

1 comment:

  1. Looking forward to reading your blog. I am seriously considering reinventing myself as an algo trader. (I was an equity quant researcher for almost two decades, but my models were for long-term performance, not intra-day movements.) I think my old job has been commoditized and I'm thinking I can trade my own account (with leverage provided by the broker) well enough to put food on the table.

    ReplyDelete