PrincessTenko’s delayed Stock transactions

Been getting questions about her transactions (pls post them on my blog as comments) Thought I’d clarify that the transaction section is delayed.

This means that there is a lag between the time the trade occurred  and when it is printed, i.e. a few periods ago PrincessTenko dabbled in that stock.

See her in action (and possibly here arch rival soon) here: quant.peterirojah.com


Tools And Then More Tools

Recently started developing a stock screening system. I say system because unlike my spreadsheet based process this is a fully automated System.

1. Create the Asset Universe: Using criteria that I believe identifies performing stocks (statistically good assets exhibit for example trend / PE etc). The infrastructure that powers this is a perl scrapping tool + SQL Server to store all the data.

2. Analyse the Universe separating the real winners from the losers. Again here more analysis is performed such as predictive econometrics etc. Matlab is used for number crunching with Yahoo as the auxiliary quote source (historical mainly). Data is stored in SQL Server.

3. What to do with the final list of stocks … not decided yet but I can say that the performance (YTD) thus far is 20% above the FTSE 100.

Quant, Random Inno

Black Scholes Merton and LTCM – Part 1

I have just watched BBC’s documentry about Black-Scholes (for the 10th time).

Black Scholes (BBC)

Every time I watch it i learn something new, mainly because each time my knowledge of the subject matter has significantly increased in between viewings.

The formula below is the fundamental BS equation (solution a la Merton)

C(S,t)=N(d_{1})~S-N(d_{2})~K e^{-r(T-t)}\,

My sister was asking me so why did they assume that the strategies the employed at LTCM will consistently return profits … I’m seriously pondering this. So far I believe that in the long run there startegies failed for several reasons but before we mention them it’s worth saying what these starts were. From general material like Wikipedia, one can deduce most pf the starts were based on mean reverting models (this is essentially the foundation for a pairs trading strategy which LTCM employed extensively.)

Reasons for the demise of LTCM:

  • Assumed non stochastic rates
  • Assumed constant volatility
  • Failure of rationality
  • Fundamental breakdown about the understanding of a pairs trading model.

In part 2 I would delve deeper into the fundamentals.