DislikedI can tell you the basic principle of my method and maybe you can come up with something better.
My initial problem was to find a way to obtain an average value of the market. And I don't mean a regular moving average (simple, exponential, weighted etc) but one that is not that susceptible to variations. I found a formula that uses some kind of regression to average AND even forecast the price (it is widely used in physics and biology statistics and it is very reliable, I know because I work in research)...
After that, the rest is just textbook statistics: 2 stadard deviations from this average will give you a 95% rate of successful trades; if you combine this with the direction of the trend, you can filter the unsuccessful trades and always be a winner.
A simpler way to explain my system is this: I use a channel that tells me what are the most probable upper/lower limits for the price. I sell when price hits the upper limit and I buy when price hits the lower limit.
For the rest of the adjustments I use indicators for daily fib pivots, daily average and weekly average. Also vegas tunnel for H1 timeframe is very useful for my entry/exit points.
That's about it.Ignored
congrats for this excellent concept. when reading the above description i came to some thoughts that could help to further develop this concept. I suppose your trades will continue to do excellent as long as the trend models you rely on continue to go the same direction. But as soon as the trends start to change the risk is high to suffer some larger losses though the low risk entries. Therefore my question here and at the same time my suggestion is the following:
how about feeding the regression& forecasting model with trend slope data (maybe regression slope data from lower time frames) instead of the price action and thus truying to forecast the trend turning points ? Would this additional concept maybe help to avoid those potential losses ?
Many thanks
nick