But seriously -- the author's comments seem to indicate that he is not totally comfortable with the technique and that he is aware it could lose, even though it didn't in this case.

I still dabble with variations of it (made like 250 bucks last friday), but I'm now doing way more research and backtesting before I drop things into the market for real.

and on having the "secret sauce" to beat the market. Read about Edward Thorpe. He's a badass. He wrote the book "beat the dealer" which was the bible on counting cards and beating vegas at blackjack.

Then he figured out how to price options (convertibles, really) and made a killing in his own hedge fund (Princeton Newport Associates, I believe).

He had the secret sauce, used it to make a killing instead of publishing it. YEARS later, black and scholes published the essentially SAME formula for price options and eventually won a nobel price for their work.

But Thorpe stuck to his guns and made his fortune. Merits to both sides, I suppose.

In this particular case: a Nobel prize.

I did it for fun, to learn, to gamble, and hopefully get lucky. God knows I've spent money in worse ways before.

Regarding the theory that other players would jump in the market and cause the opportunity to go away: you're right, but they'd really have to parametrize the same strategy in the same way as you. thousands of people use moving averages already, but all parametrized differently so they produce buy and sell signals at different times.

There's certainly no ideal parametrization that is alawys profitable, as most research shows (and I tend to believe)

There's other quant strategies that HF's use... like taking a bid out of a stock near support to trigger all the stop loss orders and then covering their shorts quickly.

FWIW I think open source software is a great thing but in general a horrible way for any sizable corporation to try to make money.