4 ms·

So... anyone want to share some of these algorithms for predicting stock prices?

by jbob2000 11y ago

tdees40 11y ago
Search "trading algorithms" on Amazon. They're not secret.
TYPE_FASTER 11y ago
Register for Quantopian and see what people are working on.
aldanor 11y ago
Making money on daily trading != predicting stock prices. More often than not, you don't want to be trying to predict prices or use fair value models if you are a high frequency trader.
swah 11y ago
https://en.wikipedia.org/wiki/Trend_following https://en.wikipedia.org/wiki/Trend_following
gnaritas 11y ago
You don't have to predict the market to trade, you need only follow the market and ride the waves carefully. If the market is going up, buy, going down, sell, you're betting that it'll continue in that direction. More than a few people are rich off strategies as simple as buy when price breaks a 20 day high and sell when price breaks a 20 day low.
- dllthomas 11y ago
  "More than a few people are rich off strategies as simple as buy when price breaks a 20 day high and sell when price breaks a 20 day low." More than a few people are broke off such strategies, too. While momentum strategies manifestly can work out, there is no guarantee that the "momentum" is actually present. As my econ prof put it, "The market isn't going; it went."
  - gnaritas 11y ago
    > More than a few people are broke off such strategies, too. Of course they are, it takes more than an entry strategy. > While momentum strategies manifestly can work out, there is no guarantee that the "momentum" is actually present. There doesn't need to be. Price moves up, down, or sideways, if it doesn't go up/down as expected, get out and look for another opportunity or wait for the market to move, it will move.
    - dllthomas 11y ago
      "Of course they are, it takes more than an entry strategy." If all that is is their "entry strategy" then the strategy they are using is more complicated, which runs counter to your earlier assertion. "There doesn't need to be. Price moves up, down, or sideways, if it doesn't go up/down as expected, get out and look for another opportunity or wait for the market to move, it will move." It will move. But it needs to move in your favor more than against you, or you're losing money. There's no particular reason to expect that to be persistently the case.
      - gnaritas 11y ago
        > If all that is is their "entry strategy" then the strategy they are using is more complicated, which runs counter to your earlier assertion. Not really, you need to have a position size strategy as well as an exit strategy, all are fairly simple and only a few lines of code. Given a bankroll, how much do you bet on any one trade (2% is trader standard, or if you're really aggressive half kelly), and given an exit strategy, i.e. a stop if you're wrong and an exit if you're right, just compute how much to trade based on the amount you're willing to risk on the trade to the stop position. So with a 10k roll, you place trades that will lose $200 bucks if you're wrong and you let them ride until trend change hopefully netting $600 or more by the time it's over, perhaps lots more depending on how long the trend runs. Once that trade is safe, i.e. you're risk off, you start looking for another one and you build up your position size without ever risking more than your initial risk. > It will move. But it needs to move in your favor more than against you, or you're losing money. There's no particular reason to expect that to be persistently the case. Sure there is, markets aren't random, they trend persistently for longer than they should if they were actually random. The EURUSD was on a 6 month long downtrend this last year; that is not a random occurrence, that is a result of long term economic trends. If you're selling breaks of a running low, you'd have been shorting that all the way down; you need only adjust for volatility so you don't get knocked out on normal sized retracements and ride the trend down. You have to size your stops to adjust for normal volatility and allow the trend to carry you to a win and you have to tune your entry so you're already near an extreme against the trend so your odds are better, you can't enter blindly, but it doesn't have to be dead on accurate if your stops are wide enough. But wider stops require more capital to profit because you have to stay in the market longer to see your profit point hit if you're keeping a favorable risk to reward. When the market isn't trending, don't trade it, you'll lose your ass in random price movements.

So... anyone want to share some of these algorithms for predicting stock prices?

tdees40 11y ago

Search "trading algorithms" on Amazon. They're not secret.

TYPE_FASTER 11y ago

aldanor 11y ago

Making money on daily trading != predicting stock prices. More often than not, you don't want to be trying to predict prices or use fair value models if you are a high frequency trader.

swah 11y ago

https://en.wikipedia.org/wiki/Trend_following https://en.wikipedia.org/wiki/Trend_following

gnaritas 11y ago

You don't have to predict the market to trade, you need only follow the market and ride the waves carefully. If the market is going up, buy, going down, sell, you're betting that it'll continue in that direction. More than a few people are rich off strategies as simple as buy when price breaks a 20 day high and sell when price breaks a 20 day low.

dllthomas 11y ago
"More than a few people are rich off strategies as simple as buy when price breaks a 20 day high and sell when price breaks a 20 day low." More than a few people are broke off such strategies, too. While momentum strategies manifestly can work out, there is no guarantee that the "momentum" is actually present. As my econ prof put it, "The market isn't going; it went."
- gnaritas 11y ago
  > More than a few people are broke off such strategies, too. Of course they are, it takes more than an entry strategy. > While momentum strategies manifestly can work out, there is no guarantee that the "momentum" is actually present. There doesn't need to be. Price moves up, down, or sideways, if it doesn't go up/down as expected, get out and look for another opportunity or wait for the market to move, it will move.
  - dllthomas 11y ago
    "Of course they are, it takes more than an entry strategy." If all that is is their "entry strategy" then the strategy they are using is more complicated, which runs counter to your earlier assertion. "There doesn't need to be. Price moves up, down, or sideways, if it doesn't go up/down as expected, get out and look for another opportunity or wait for the market to move, it will move." It will move. But it needs to move in your favor more than against you, or you're losing money. There's no particular reason to expect that to be persistently the case.
    - gnaritas 11y ago
      > If all that is is their "entry strategy" then the strategy they are using is more complicated, which runs counter to your earlier assertion. Not really, you need to have a position size strategy as well as an exit strategy, all are fairly simple and only a few lines of code. Given a bankroll, how much do you bet on any one trade (2% is trader standard, or if you're really aggressive half kelly), and given an exit strategy, i.e. a stop if you're wrong and an exit if you're right, just compute how much to trade based on the amount you're willing to risk on the trade to the stop position. So with a 10k roll, you place trades that will lose $200 bucks if you're wrong and you let them ride until trend change hopefully netting $600 or more by the time it's over, perhaps lots more depending on how long the trend runs. Once that trade is safe, i.e. you're risk off, you start looking for another one and you build up your position size without ever risking more than your initial risk. > It will move. But it needs to move in your favor more than against you, or you're losing money. There's no particular reason to expect that to be persistently the case. Sure there is, markets aren't random, they trend persistently for longer than they should if they were actually random. The EURUSD was on a 6 month long downtrend this last year; that is not a random occurrence, that is a result of long term economic trends. If you're selling breaks of a running low, you'd have been shorting that all the way down; you need only adjust for volatility so you don't get knocked out on normal sized retracements and ride the trend down. You have to size your stops to adjust for normal volatility and allow the trend to carry you to a win and you have to tune your entry so you're already near an extreme against the trend so your odds are better, you can't enter blindly, but it doesn't have to be dead on accurate if your stops are wide enough. But wider stops require more capital to profit because you have to stay in the market longer to see your profit point hit if you're keeping a favorable risk to reward. When the market isn't trending, don't trade it, you'll lose your ass in random price movements.