The other day I got Joel Greenblatt’s The Little Book That Beats the Market out from the public library. (What money-minded person would buy books that are in the library?)

It’s a light, cute read, but it offers suggestions that accord very well with what I understand of value investing. At the end you are presented with a “Magic Formula” for ranking stocks that is supposed to be likely to do better than the market, and again I should think it would, since it basically selects companies that get the best return on their capital that are currently priced cheapest in the market.

I thought it would be an interesting exercise to write a stock screener for New Zealand and Australia that uses this formula. This would be of some value, because no one seems to be offering such a service here. So I’m going to use Python to do it.

The formula requires you to know various figures about each company. Some are published by Yahoo and other websites, and the others can be derived from published data.

So what I need to do is:

• get a list of stocks
• generate URLs for those stock’s pages on Yahoo etc
• for each stock
  • spider the relevant page from Yahoo or wherever
  • record the data I want
  • calculate the results of the two Magic Formula metrics
• generate a ranking of stocks according to the Magic Formula results

This is clearly going to involve some sort of HTML parsing, so I think my first major technical decision is going to be what to use.