After testing the script extensively, I come to the conclusion that the backtesting looks far better than the actual results that are created by following the signals when they actually occur. This is because the entire signal curve gets repainted after considering the price development of later bars.
In my testing I would for instance get a short signal at the end of a 15 minute bar and a long signal at a much higher price say after an hour. I would see both signals in the chart right in front of me confirming the loss incurred by these two transactions. BUT after refreshing the tradingview page, both signals were gone and the indicator curve showed as if I had always been long - never incurring that loss.
Another example of repainting: The price was on a down trend with some recoveries in between. The script created a series of short and long signals on its path down (most of them led to a loss). After refreshing the tradingview chart, there was only one short signal left right at the top of the downtrend.
So, clearly the results of the backtesting (see chart above) don't represent the actual flow of signals and in my opinion, the net profit percentage seems vastly overstated.
I tried to confront the developer with these questions but only got this:
"there is a repainting issue caused due to choppy markets, it is a sacrifice I had to take in order to make this work like a scalper."
Carbon-Investments
⋅
@stabilus, you are correct. This doesn't even consider trading fees which at smaller timeframes destroy 99% of strategies. As a professional portfolio manager I would not advice at all to buy this.
In my testing I would for instance get a short signal at the end of a 15 minute bar and a long signal at a much higher price say after an hour. I would see both signals in the chart right in front of me confirming the loss incurred by these two transactions. BUT after refreshing the tradingview page, both signals were gone and the indicator curve showed as if I had always been long - never incurring that loss.
Another example of repainting: The price was on a down trend with some recoveries in between. The script created a series of short and long signals on its path down (most of them led to a loss). After refreshing the tradingview chart, there was only one short signal left right at the top of the downtrend.
So, clearly the results of the backtesting (see chart above) don't represent the actual flow of signals and in my opinion, the net profit percentage seems vastly overstated.
I tried to confront the developer with these questions but only got this:
"there is a repainting issue caused due to choppy markets, it is a sacrifice I had to take in order to make this work like a scalper."