Hacker News new | past | comments | ask | show | jobs | submit login

I recommend the book "Dark Pools" by Scott Patterson[1]. It is an account of how traders beat each other by microseconds to get an order in, and can find and exploit every possible way they can in order to get an advantage.

There is (perhaps) no way a human can compete in the stock market any more. This is the age of AI, where computers are pushing around billions of dollars in an automated fashion. Taking every arbitrage advantage in milliseconds, and are even able to find under- and over-valued stocks in the blink of an eye.


That book, and your comment, conflate two different concepts: trading and investing.

It's possible to invest in the stock market with very little trading. The GP's comment spells out the best strategy--use index funds to spread risk and reduce expense, and invest for long time horizons. This is still a great strategy.

It's also possible to try to turn a short-term profit by trading individual stocks--often called day-trading because you clear your positions within the time period of days. High-frequency trading has greatly reduced price spreads, which makes it harder for the average guy in his underwear to compete. This is not investing though--more like gambling.

For more on how high-frequency trading really works, I recommend:


There's a big difference between trading and investing. All laypeople should be investing in the market, but you are talking about trading taken to an extreme. A computer trader may get in and grab 1000 shares, then resell it right away for 10 cents more. He can do that 100,000 times a day and make a good living. But an average investor buys 1000 shares and holds them for 5 years. He's losing maybe a few dollars over his lifetime to these guys.

These high speed investors aren't drastically modifying the value of the market. If they were they would be creating arbitrage situations that would be closed right away.

You're right, it's trading vs. investing. Didn't take a genius to buy Google or Apple 5 years ago, no highspeed tricks or anything and you've made a tidy chunk of profit.

Whether or not they're drastically modifying the value of the market is really hard to say, I don't think we know that fully. Honestly, I don't think we understand it that well yet. You can make some fairly reasonable assumptions that it doesn't cost the typical investor that much over his investing life though. I think part of those assumptions are that the high speed traders are simply trying to move faster when they acquire knowledge though, to me, that seems kind of elementary for the guys that invented all these exotic synthetic derivatives and ways to mask risk and shift it around... You couldn't tell me that they aren't trying to think up other uses for those technologies if they exist; like maybe you can measure what counter parties are doing with high performance timing and get some insight.

Fundamentally, are they leaving that few dollars on the table or are they picking them up? It may only be a few dollars but it's a big difference between paying it and taking it.

Applications are open for YC Winter 2022

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact