That works fine. The problem I am experiencing is minute data converted to 1 hour data is causing problems with the futures markets closing at 4:15 eastern where the last hour of the day is not included in the data set. I am using regular trading hours (9:30 to 4:15) so I expect the data conversion is expecting a full 60 minutes of data but there are only 45 minutes to end the day so the bar of data is not included. The behavior should be 3:30 to 4:15 should be the last hour bar of the day. Hopefully this can resolved.
Removing the golden handcuffs
- Thread starter fan27
- Start date
A brief testing of StockSharp looks like it might work as it supports testing/executing a portfolio of strategies. They now offer an app (S# Shell) which uses their APIs and all that is required is to write strategy code. I think I am going to roll the dice and buy the app ($490). If it does not work out will just view it as an R&D expense.
AlgoTerminal acknowledge the bug (which was recently introduced) and it should be fixed in a few weeks.
I am shelving the StockSharp idea. I asked to confirm that their training package would include samples utilizing multiple time frames in a strategy and they said no as multiple time frames in a strategy is not trivial for backtesting but works find for live trading. Uh...OK.
AlgoTerminal fixed the data bug and I am now able to generate strategy, risk and portfolio level code which can be dropped directly into AlgoTerminal. Very happy with this!
I will be starting with 48 unique strategies trading ES, NQ and YM targeting 15 min, 30 min and 1 hour time frames. Based on my account size there is a good chance I can cover my living expenses in the medium term from trading income alone which would be huge because it will allow me time to focus on getting my platform to an enterprise level offering. Some stats on the strategies I will be trading (slippage and commissions included).
2006 - 2018:
Average Yearly return: 120%
Max Yearly return: 297%
Min Yearly return: 22%
Average Yearly drawdown: -11.7%
Max Yearly drawdown: -22%
Min Yearly drawdown: -5.58%
Average trades per year: 107
Max trades per year: 144
Min trades per year: 38
Next step is run against sim for a week or so just to make sure AlgoTerminal is executing my strategies as expected and then start live trading.
Good times!
I will be starting with 48 unique strategies trading ES, NQ and YM targeting 15 min, 30 min and 1 hour time frames. Based on my account size there is a good chance I can cover my living expenses in the medium term from trading income alone which would be huge because it will allow me time to focus on getting my platform to an enterprise level offering. Some stats on the strategies I will be trading (slippage and commissions included).
2006 - 2018:
Average Yearly return: 120%
Max Yearly return: 297%
Min Yearly return: 22%
Average Yearly drawdown: -11.7%
Max Yearly drawdown: -22%
Min Yearly drawdown: -5.58%
Average trades per year: 107
Max trades per year: 144
Min trades per year: 38
Next step is run against sim for a week or so just to make sure AlgoTerminal is executing my strategies as expected and then start live trading.
Good times!
Excited for you! Do you think such low number requires you to take large size to make money? That would be a big concern for me.
I should have noted I am risking 2% of account equity per trade. I would prefer closer to 1.5% but that would mean numerous trades would be skipped given my current account size. Also, I do have risk controls in place so I will never have more than 6 positions opened at any given time across all strategies as some strategies are highly correlated.
How much is the subscription to AT?