Monthly Archives: September 2010

Scraping 10 years of historical data per security from NASDAQ

This entry is part of 15 in the series Grails Finance

Grails Finance 1.1
You can like request 10 years of historical OHLC data from the NASDAQ website. That seems more than enough for all intents and purposes. Also you can download index composition details of the Dow Jones Industrial Average Index [...]

Share
Posted in programming | Tagged , , | Leave a comment

Overfitting to polynomials

This entry is part of 15 in the series Grails Finance

Grails Finance 1.0
Overfitting occurs when a fit to historical data only works for the dataset used to fit the data. I am using polynomial fits up to the 4th degree, but in principle you could use any other sort of [...]

Share
Posted in programming | Tagged , , | Leave a comment

Are stock markets efficient?

This entry is part of 15 in the series Grails Finance

Grails finance 0.9
According to theory a market is efficient, if it has random prices. So tests for market efficiency come down to testing for randomness. Obviously the same goes, if stock prices follow some distribution that is not random.
Services
So [...]

Share
Posted in programming | Tagged , , | Leave a comment

Can we predict stock prices with Grails Finance?

This entry is part of 15 in the series Grails Finance

Grails Finance 0.8
Yes, we can. Not always correctly, of course. One way to predict things, is to try to find the probability distribution for a phenomenon. Usually a distribution is charted as a histogram, so that is what I did. [...]

Share
Posted in programming | Tagged , , | 3 Comments