Pre setup....
Attached Image (click to enlarge)
2
M1 Countertrend Scalping Strategy
- #5,844
- May 27, 2023 1:29pm May 27, 2023 1:29pm
- Joined Mar 2023 | Status: Trader | 1,007 Posts
without patience how can you be a trader?
- #5,846
- May 27, 2023 1:43pm May 27, 2023 1:43pm
- Joined Mar 2023 | Status: Trader | 1,007 Posts
without patience how can you be a trader?
- #5,848
- May 27, 2023 1:52pm May 27, 2023 1:52pm
- Joined Mar 2023 | Status: Trader | 1,007 Posts
without patience how can you be a trader?
Algo trading experimentation:
When a lagging indicator says "buy" we SELL, when the same indicator says "sell" we BUY and we get a basic contrarian (counter-trend) strategy, we add a simple target to increase performance and a market exit triggered by the indicator itself. Image shows out-of-sample performance and trade examples.
Backtest shows promise, but we can find a better entry/exit logic and improve it:
The basic strategy does not use any kind of cost averaging.
Let me know what you think, does "out of the box" thinking pay off in the long-term?
When a lagging indicator says "buy" we SELL, when the same indicator says "sell" we BUY and we get a basic contrarian (counter-trend) strategy, we add a simple target to increase performance and a market exit triggered by the indicator itself. Image shows out-of-sample performance and trade examples.
Backtest shows promise, but we can find a better entry/exit logic and improve it:
The basic strategy does not use any kind of cost averaging.
Let me know what you think, does "out of the box" thinking pay off in the long-term?
- #5,852
- May 27, 2023 5:55pm May 27, 2023 5:55pm
- Joined Mar 2023 | Status: Trader | 1,007 Posts
without patience how can you be a trader?
- #5,853
- May 27, 2023 8:46pm May 27, 2023 8:46pm
- Joined Dec 2022 | Status: Trader | 3,262 Posts
every minute matters!
- #5,854
- May 27, 2023 9:19pm May 27, 2023 9:19pm
- Joined Sep 2013 | Status: Trader | 25,152 Posts
people leave, but this thread will go on for ever
- #5,855
- May 27, 2023 9:39pm May 27, 2023 9:39pm
- Joined Dec 2022 | Status: Trader | 3,262 Posts
every minute matters!
Different algorithm, same logic, but this time we are using the H4 timeframe:
A closer look into a good trading period:
Having no hard SL and a loose market exit allows the system to collect a lot of pips in most circumstances as price wiggles up and down for extended periods of time.
All mechanical systems have bad periods where they get out-of-sync with the market and begin to whipsaw (they need to be re-adapted periodically to reduce the risk of this happening):
It's also important to judge a group of trades over individual cases:
Note: More robustness tests must be done, a single stage out-of-sample is not enough to rule out "lucky" outliers.
The main weakness of this approach is that the losing trades tend to be bigger than the winning trades, so we are betting that the instrument retains its main traits in the future to be profitable, in this case = chopping a lot + weak trends.
Some configuration combinations show a seemingly perfect run in-sample and out-of-sample, however they don't have enough trades to be considered statistically significant, so we need to use more data and common sense to judge them:
A closer look into a good trading period:
Having no hard SL and a loose market exit allows the system to collect a lot of pips in most circumstances as price wiggles up and down for extended periods of time.
All mechanical systems have bad periods where they get out-of-sync with the market and begin to whipsaw (they need to be re-adapted periodically to reduce the risk of this happening):
It's also important to judge a group of trades over individual cases:
Note: More robustness tests must be done, a single stage out-of-sample is not enough to rule out "lucky" outliers.
The main weakness of this approach is that the losing trades tend to be bigger than the winning trades, so we are betting that the instrument retains its main traits in the future to be profitable, in this case = chopping a lot + weak trends.
Some configuration combinations show a seemingly perfect run in-sample and out-of-sample, however they don't have enough trades to be considered statistically significant, so we need to use more data and common sense to judge them:
- #5,858
- Edited May 28, 2023 12:00am May 27, 2023 10:11pm | Edited May 28, 2023 12:00am
{quote} those results look good, i'm guessing the TP has been optimised? i tried this concept many times and many different ways in ea's and often with worse results, when an indicator is unreliable in terms of consistency for its intended purpose, then it is more so in the opposite direction, but that's just my experience... in terms of "out of the box" thinking, it could work well as a manual approach, but within an ea it will suffer the same insurmountable challenge of remaining consistently profitable over the long term i believe...
It's actually a fixed target of 15 pips. I've found through testing that these work better than ATR (adaptative) targets over and over. I rather let the algorithm "fit" itself into a pre-defined TP than mixing everything together in a single optimization or using a slow indicator to define targets. A GOOD algorithm will work well out-of-sample with a 15, 20, 25 or 30 target.
I've made a volatility study using ATR averages over a period of 10+ years that I use for reference for my manual and algo trading (numbers refer to AVG pips per bar over those years):
However I don't use them strictly like that, I may use 15, 20 or 40 pip targets, but never include them in the optimization per se. Hope it makes sense
Mechanical systems need to be re-adapted to be in sync with the most recent market conditions to survive, otherwise they just fall apart! I haven't included that type of the analysis in this thread yet, but it's necessary. Algebra has no memory unlike AI and humans.
- #5,859
- May 27, 2023 11:35pm May 27, 2023 11:35pm
- Joined Sep 2013 | Status: Trader | 25,152 Posts
{quote} Agree 100% with you my friend! I am actually not bashing lagging indicators as I even use them myself for some of the systems, some of them are really good, modern and fast. The point that I am trying to make is that they may also be used in unconventional ways to make profits.
we both know they is many ways to make money on forex and this is just one of them , i do like your thread a worth while project for beginners 
people leave, but this thread will go on for ever
- #5,860
- May 28, 2023 3:00am May 28, 2023 3:00am
- Joined Dec 2022 | Status: Trader | 3,262 Posts
every minute matters!