Structuring trading day
- Thread starter tradermike99
- Start date
Why can't you automate it? Even with automation tools working outside and apart from your strategy code?
https://www.logigear.com/blog/test-automation/12-best-automation-tools-for-desktop-apps-in-2020/
A handful of thoughts here:
1) Develop morning routine before you even do anything with the markets: devotionals, meditation, yoga, workout, hydration.
2) Review daily charts / indicators from the day before: It's very easy to get caught up in what the market is doing now. For instance, I might have thought the charts looked bullish based on prior daily close but the futures are down to start the day. I shouldn't assume it's supposed to be a down day -- maybe it's just an opportunity to get long.
3) Determine what time frames you are trading and what times you need to be really locked in. I can't stare at my screens all day. Hourly charts are my shortest primary time frame for trading, although I will drop down to 15-minute for more precise entry.
4) Build in breaks. Get up. Move. Walk the Dog. Exercise. Play guitar.
1) Develop morning routine before you even do anything with the markets: devotionals, meditation, yoga, workout, hydration.
2) Review daily charts / indicators from the day before: It's very easy to get caught up in what the market is doing now. For instance, I might have thought the charts looked bullish based on prior daily close but the futures are down to start the day. I shouldn't assume it's supposed to be a down day -- maybe it's just an opportunity to get long.
3) Determine what time frames you are trading and what times you need to be really locked in. I can't stare at my screens all day. Hourly charts are my shortest primary time frame for trading, although I will drop down to 15-minute for more precise entry.
4) Build in breaks. Get up. Move. Walk the Dog. Exercise. Play guitar.
Great thread. My routine:
7am-9:30 premkt scanning and developing watchlist.. may do a few premkt trades. Also set up buy-stop orders
9:30-10:30 daytrading entries and exits, sequences of trades
Midday: check charts occasionally and tighten trailing stops for open positions
3pm-4pm closing hour trades including eod short squeeze plays
Aftermarket review of lessons learned
7am-9:30 premkt scanning and developing watchlist.. may do a few premkt trades. Also set up buy-stop orders
9:30-10:30 daytrading entries and exits, sequences of trades
Midday: check charts occasionally and tighten trailing stops for open positions
3pm-4pm closing hour trades including eod short squeeze plays
Aftermarket review of lessons learned
Part of the strategy that cannot be translated into code.
Ok, thanks. Are you a skilled programmer? I mean, are you saying it can't be translated; or did someone else tell you that?
Just curious, as I've never fathomed a strategy that can't be reduced to code--of course, it may be time consuming and require that the model be trained in order to derive the 'rules' the human trader can't properly articulate.
I don't know if I am a skilled programmer but my background is computer science and been writing all codes related to trading by myself. Saying it is impossible is exaggerating. I would not say it's impossible to write the code but I have yet successfully translate the element (overlapping price structure) into code. Quite possibly I am just not a good programmer. Who knows.
You sound like a skilled programmer. It's just that, imo, certain things may require a previously, unconsidered perspective.
Maybe instead of trying to hard code the rules, have an algo store the examples along with your determinations for each example, and then try matching the new examples with prior ones, and using your determination from the prior one that matches best...if that makes sense.
Sounds like AI which is not my turf. I specialize in telecommunication/network engineering.You sound like a skilled programmer. It's just that, imo, certain things may require a previously, unconsidered perspective.
Maybe instead of trying to hard code the rules, have an algo store the examples along with your determinations for each example, and then try matching the new examples with prior ones, and using your determination from the prior one that matches best...if that makes sense.