In all seriousness, in modern automated trading, the execution performance is worth more than the strategy. The platform facilitates models, the models don't facilitate the platform. The business types try to maintain the mystique, but this is changing. My suspicion is that the real reason the industry is fighting so hard against the automated guys is because they know that once the government steps in, it goes away from being a meritocracy back to the "good ol' boys" club, where profits are a function of deal-making more so than raw quantitative and computational elegance. Automated trading was the best thing that ever happened to the markets, and as you know, government has a habit of ruining things in this country (the US.) In terms of market insight, it's really a question of whether the programmer has the drive or not. The job of "trader" is done for, unless that trader has some special quantitative ability or programming ability that can work magic with simple ideas.
i agree that traders in the automated trading space are useless. and in chicago they are usually guys right out of school watching systems. however, i disagree that the platform is worth more than the strategy. for the most part, the platforms are the same and 99% of firms are NOT microsecond focused; they are competitive. what good is the fastest platform if you have nothing to run on it.
Sorry, people coming out of universities cannot replace trained programmers. If you are talking about the bleeding edge, just any programmer will not do. It really takes some serious code-nazis to come up with the right implementation, and the experience to manage performance. And, the hard part is, it's not just a degree from MIT or CMU that makes you a good programmer. If you're talking about the average system that just works, where slippage is ok, then sure, I agree. Why even bother with programmers then, just go use EasyLanguage or some other descriptive, high-level language.
The less risk there is involved in a trade (e.g, some well-known arb), the more the platform matters. If your trajectory is to take your business and put it on a colo and run some fancy algorithm, you still need to focus on developing the code right so you don't get screwed in the future when your VB/EasyLanguage/stuff-that-could-be-coded-by-anyone needs a costly, expensive re-architecting. If you have some sophisticated hedged trade and a long-term forecast, you can hire anybody and write the code in visual basic for all it matters. Why even bother with customization? If you are getting away with 10-second executions and can use java to talk to Interactive Brokers, you can hire anyone has a pulse and 100-level computer science class training.
If you are involved in some 'well-known arb' there is nothing proprietary about this trade and then of course one good programmer might be worth more than this entire firm.
Exactly. I think we ultimately see eye to eye, here. Although, let's say hypothetically that programmer realized he was worth more than the firm. He may have issues raising the capital and getting situated into a position where his skill is applicable, so ultimately he ends up trading off his value for a much lower salary. What do you think it would take a programmer like that to get started? Probably 3 to 10 mil?
But even though index being a trading strategy it does not mean that no one else can trade it, does it?