Great! Are you running into any issues? Lord knows there are a bunch when backtesting in EasySignal. I toast you globally with a Kirin Ichiban "Special Premium Reserve" brewed in Los Angeles.
You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an alternative browser.
You should upgrade or use an alternative browser.
eSignal back-testing limitations
- Thread starter Fractal
- Start date
Great! Are you running into any issues? Lord knows there are a bunch when backtesting in EasySignal. I toast you globally with a Kirin Ichiban "Special Premium Reserve" brewed in Los Angeles.
Quote from Arthur Deco:
Great! Are you running into any issues? Lord knows there are a bunch when backtesting in EasySignal. I toast you globally with a Kirin Ichiban "Special Premium Reserve" brewed in Los Angeles.
No issues so far, happy to say. Aside from the 6 month limitation and the account graphs not drawing on the html output, but the latter's likely due to some anti-script add-ons I have for Firefox. Remains to be seen if the tweaked method stands up to live trading, but I am very pleased regardless.
Quote from Fractal:
No issues so far, happy to say. Aside from the 6 month limitation and the account graphs not drawing on the html output, but the latter's likely due to some anti-script add-ons I have for Firefox. Remains to be seen if the tweaked method stands up to live trading, but I am very pleased regardless.
Glad to hear it. My account graphs won't draw either, but I don't care so I never investigated it. Re the six months of data, as we say here in Texas, "If you can't fix it, feature it!" I have found for my systems that six months roughly represents sea changes in the market, so it seems to be just enough to keep my optimizations current. I have no idea what your system is like, but kindly permit me to share something with you I lernt over the past few months. I almost always intuitively beat my systems' performance on profit taking, so I now think of the optimizations as rough guides. The stop losses I honor like wearing a rubber when fucking a whore, but I am more flexible about when to "come."
Quote from Arthur Deco:
Glad to hear it. My account graphs won't draw either, but I don't care so I never investigated it. Re the six months of data, as we say here in Texas, "If you can't fix it, feature it!" I have found for my systems that six months roughly represents sea changes in the market, so it seems to be just enough to keep my optimizations current. I have no idea what your system is like, but kindly permit me to share something with you I lernt over the past few months. I almost always intuitively beat my systems' performance on profit taking, so I now think of the optimizations as rough guides. The stop losses I honor like wearing a rubber when fucking a whore, but I am more flexible about when to "come."
Fantastic metaphor and sound advice -- I'll try to keep an eye out for that dynamic when trading this systematically. I see some slight nuances where I'd trade differently than what's on the table. Worth keeping separate stats on the differences.
Agreed on the stats. When I first started coding and testing intraday price patterns I ignored the nuances. But like you, I began to wonder if subtle refinements to the basic crude patterns might be in order. Consider the worst case I would ever remotely consider trading. I can't stomach a system which wins only once out of four trades, regardless of how great the expectation is. In six months of data let's say the pattern pops once a week, roughly 24 times, for 6 wins and 18 losses. The 18 losses are IMO sufficient to generate a stop loss which is statistically significant. But the 6 wins are utterly inadequate to refine an accurate profit target. So I don't swing for the stands. I bunt with modest expectation systems which trade more often and which I think are statistically reliable. Such systems also give you a shot at refinements to the crude pattern to incorporate with some confidence the intuitions you have.
I have 2 questions concerning esignal:
1. Is it possible to backtest based on a daily and intraday indicator at the same time such as buy if the the 14 day RSI is above 80 and if the intra-day 15-minute RSI is above 80.
2. To auto-trade a strategy developed in esignal, we need to use one of their "Integrated Trading Service Providers" such as Patsystems or Rithmic.
Thanks for any input.
1. Is it possible to backtest based on a daily and intraday indicator at the same time such as buy if the the 14 day RSI is above 80 and if the intra-day 15-minute RSI is above 80.
2. To auto-trade a strategy developed in esignal, we need to use one of their "Integrated Trading Service Providers" such as Patsystems or Rithmic.
Thanks for any input.
1. You'll have to calculate the daily indicator on the 15 minutes chart. You'll get 6 months of data.
2. ESignal does not support fully automated trading with any broker.
2. ESignal does not support fully automated trading with any broker.