> Reacting before “individual investors” and trying to read the news 5 minutes faster than them is not a sufficient edge to consistently beat the market. You have to react faster than the market moves, which means you have to move faster than hedge funds, banks, prop shops, mutual funds, etc. The market moves very fast, and it is dominated by very large sophisticated participants.
5 minutes is plenty of time, well shorter than most successful day traders who are trying to predict on 15 minute or hourly horizons, or swing traders trying to predict on daily horizons. There are plenty of trends that occur on the horizons of minutes, hours, days, even weeks.
> First of all, 5 minutes is a very long period of time in the world of high frequency trading, where machines are executing millions of trades every millisecond. Secondly, the same machines are equipped with AI algorithms for reading real-time news feeds, earnings releases, etc, and reacting very quickly on the news. Thirdly, the amount of money that the machines (and their owners) are trading with dwarfs any impact that individual investors would have on market prices.
This comment reads like a science fiction fan's made up ideas about how NASA works. You've taken snippets of vague ideas about the truth and mishmashed them into a frankenmonster that would never exist in reality.
* Millions of trades every millisecond is an exaggeration several orders of magnitude off from the real world volumes of HFT firms.
* HFT algorithms don't use AI. They wouldn't be able to react fast enough if they did. HFT algorithms in practice are rarely more complicated than linear regression.
* Those who use AI in their algorithms are doing so with incredibly varied levels of success. They are typically trading on horizons much longer than HFT firms. They are trading volumes orders of magnitude lower than HFT firms, even if those volumes are orders of magnitude higher than individual investors. From what I can tell, outside of text analysis of the potential sentiment generated by news, the finance industry is pretty pessimistic about AI.
* There are plenty of traders operating successfully across several time horizons. It is not necessary to be faster than HFT firms to be successful. Most day traders like me actually depend on the liquidity provided by HFT firms in order to be profitable; we couldn't do it without the deep books that they provide.
* Likewise, a lot of the predictability of how the market moves comes from the market effects of large actors: when you trade $100M in a day in a single direction, you move the market with you, and that creates a trend pattern that smaller traders can pick up on and take for a ride.
5 minutes is plenty of time, well shorter than most successful day traders who are trying to predict on 15 minute or hourly horizons, or swing traders trying to predict on daily horizons. There are plenty of trends that occur on the horizons of minutes, hours, days, even weeks.
> First of all, 5 minutes is a very long period of time in the world of high frequency trading, where machines are executing millions of trades every millisecond. Secondly, the same machines are equipped with AI algorithms for reading real-time news feeds, earnings releases, etc, and reacting very quickly on the news. Thirdly, the amount of money that the machines (and their owners) are trading with dwarfs any impact that individual investors would have on market prices.
This comment reads like a science fiction fan's made up ideas about how NASA works. You've taken snippets of vague ideas about the truth and mishmashed them into a frankenmonster that would never exist in reality.
* Millions of trades every millisecond is an exaggeration several orders of magnitude off from the real world volumes of HFT firms.
* HFT algorithms don't use AI. They wouldn't be able to react fast enough if they did. HFT algorithms in practice are rarely more complicated than linear regression.
* Those who use AI in their algorithms are doing so with incredibly varied levels of success. They are typically trading on horizons much longer than HFT firms. They are trading volumes orders of magnitude lower than HFT firms, even if those volumes are orders of magnitude higher than individual investors. From what I can tell, outside of text analysis of the potential sentiment generated by news, the finance industry is pretty pessimistic about AI.
* There are plenty of traders operating successfully across several time horizons. It is not necessary to be faster than HFT firms to be successful. Most day traders like me actually depend on the liquidity provided by HFT firms in order to be profitable; we couldn't do it without the deep books that they provide.
* Likewise, a lot of the predictability of how the market moves comes from the market effects of large actors: when you trade $100M in a day in a single direction, you move the market with you, and that creates a trend pattern that smaller traders can pick up on and take for a ride.