Anytime Dan.
For me I do something similar but instead of scanning a bunch of symbols looking for certain things I've written that in to the main buy/sell strategy and just run it on ~500 symbols everyday. I gathered my symbol list from filters on barchart.com's Stock Screener for volatility, avg price range, and etc. Then once per year I'll rescan and remake that active symbol list. I created my strategy to only invest X dollars per buy with a min account value left for any account fees and etc. This is also what I believe is the better way to diversify as A) the symbol list is spread across all industries and B) my investments aren't heavily focused on just a small handful of stocks spreading things out even more as in general at any given time 50% of stocks are going down and 50% are going up as people shift money from one place to another.
This approach, using smaller investments across many stocks, allows you to better play with and be safe with crazy ideas such as buying at a 52 week low and selling at a 52 week high. From an optimists, mean reversionists, approach either the stock will return as that's always the goal of the CEO or the company goes out of business. If it goes bankrupt then no big deal because you only had say 200-500 dollars invested anyways and statistically the others that rebounded far outweighs the companies that went under. I'm not saying to do that but a highly diverse playing field allows for trying all the things the 'big guys' say to never do but could return huge profits.
One last note on testing unless you've written something specific for a specific symbol is to always test your strategy across hundreds of stocks using the Strategy Analyzer. It is easy to write code that curve fits to a single stock but testing it across say 500 stocks and it still does well says that the code is robust and will mostly likely work well as general market changes slowly occur. Also testing across 500 symbols means your code has pretty much seen all that is possible to be seen with fast dips, gaps up/down, choppy terrain, and long slides up/down. Then just remember that it's a numbers game and that it's impossible to win them all but if you're only using smaller amounts per investment and something tanks it's not so bad to just let it sit there for a year or so and not realize those unrealized losses.
Anyways have fun.
For me I do something similar but instead of scanning a bunch of symbols looking for certain things I've written that in to the main buy/sell strategy and just run it on ~500 symbols everyday. I gathered my symbol list from filters on barchart.com's Stock Screener for volatility, avg price range, and etc. Then once per year I'll rescan and remake that active symbol list. I created my strategy to only invest X dollars per buy with a min account value left for any account fees and etc. This is also what I believe is the better way to diversify as A) the symbol list is spread across all industries and B) my investments aren't heavily focused on just a small handful of stocks spreading things out even more as in general at any given time 50% of stocks are going down and 50% are going up as people shift money from one place to another.
This approach, using smaller investments across many stocks, allows you to better play with and be safe with crazy ideas such as buying at a 52 week low and selling at a 52 week high. From an optimists, mean reversionists, approach either the stock will return as that's always the goal of the CEO or the company goes out of business. If it goes bankrupt then no big deal because you only had say 200-500 dollars invested anyways and statistically the others that rebounded far outweighs the companies that went under. I'm not saying to do that but a highly diverse playing field allows for trying all the things the 'big guys' say to never do but could return huge profits.
One last note on testing unless you've written something specific for a specific symbol is to always test your strategy across hundreds of stocks using the Strategy Analyzer. It is easy to write code that curve fits to a single stock but testing it across say 500 stocks and it still does well says that the code is robust and will mostly likely work well as general market changes slowly occur. Also testing across 500 symbols means your code has pretty much seen all that is possible to be seen with fast dips, gaps up/down, choppy terrain, and long slides up/down. Then just remember that it's a numbers game and that it's impossible to win them all but if you're only using smaller amounts per investment and something tanks it's not so bad to just let it sit there for a year or so and not realize those unrealized losses.
Anyways have fun.
Comment