I remember some years ago we had a discussion (with FXEZ I think), in a thread I couldn't find back, about trading the equity curve like MP suggests. The conclusion was that you need your equity curve to be positively autocorrelated. Many of the ...

Diversified Trend Trading Approach

Would it be possible for you to post these trades in some excel/csv format?

Diversified Trend Trading Approach

Theil-Sen is just another way to perform a linear regression. It is like SDC just bit more immune to the non-white noise. you still need to decide the width of the channel (1, 2, 3 SD). The idea of the indicator was indeed to replace the parabolic ...

Diversified Trend Trading Approach

Don't confuse fat tail and black swan. A 5 ATR sudden move isn't a black swan. A black swan is an event not even considered by you model. A black swan is more like you're trading "Dutch guilder vs Deutch Mark" and suddenly you trade "EUR vs EUR".

Diversified Trend Trading Approach

I love it when there is no parameters in an indi. When everything is data driven. - Mann-Kendall Test tells you the probability that some monotonic trend, non necessarily linear, exists in a dataset. - Theil-Sen estimator tells you what is the slope ...

Diversified Trend Trading Approach

When you filter a sin wave with a SMA of exactly the same period the positive values of the sin cancel out with the negative ones. The output is flat. If the signal is a trend plus a sin wave you recover the trend. Of course if there are more than ...

Diversified Trend Trading Approach

This is more or less what KAMA does. For the MA in this post url I do this: The Goertzel algorithm measures the power at a specific frequency in a discrete signal. I build 74 such filters each tuned to a period between 2 and 75. At the same time I ...

Diversified Trend Trading Approach

In a trending market the price gets far from a MA. If the lag of your MA is constant at all frequencies (linear phase) and if the trend is constant (straight line) you can show that the distance is proportional to the trend. Even if it is not the ...

Diversified Trend Trading Approach

I don't have a definitive preference for the volatility. Thus far the standard deviation of the returns is the best I have. The issue is that to estimate the variance (SD=sqrt(variance)) you need quite a lot of samples (300-500 to have 98-99% ...

Diversified Trend Trading Approach

I see you're playing around with Kaufman's KAMA. Do you use the filter he describes in his book "Smarter Trading Improving Performance in Changing Markets"? I can't figure out how it can be used... The description says that he computes the standard ...

Diversified Trend Trading Approach

Hey Mr Copernicus! Sorry to disturb the flow of your thread with a completely different topic but you can't be PM'd. You wrote recently again that you're willing to find a way to backtest your strategy in an automated way. Obviously the construction ...

Diversified Trend Trading Approach

Thanks. I undestand the maths but I don't understand what are X and Y. Are they the returns of the two time series? How to apply it here? Your second link is broken.

Choppy market index: any good ideas?

That was only to try and understand the difference PlanAndTrade was having in his calculation. This is not to put in the formula. I guess he used the mid-price instead of the bid and the ask separately. I asked him to check if this could explain the ...

How to calculate average price for open orders

The spread needn't be considered because if you're net long (LotSum>LotSum1) the BE price to compute is on the bid side. And of course if you're short the BE price is on the ask side. The solution depends on your broker. With some broker you pay the ...

How to calculate average price for open orders

I don't understand what you mean by "trading a timeframe". Kullback-Leibler divergence is about comparing two probability distributions. Which ones do you suggest to compare? I found two threads by aleksey which may be related: url url Are those ...

Choppy market index: any good ideas?

What level of beginners? At the very least you need to know about probabilities and statistics: everything in machine learning boils down to estimating a probability distribution. A basic understanding of linear algebra is needed too: almost ...

Machine Learning with algoTraderJo

I tried to add an estimation of the trend too (I'm not claiming I succeeded here!). Originally I wanted to see if these "volatility cones" can help recognizing quickly a bad day. Look what happens when the price escapes the cone in term of ...

Simple Mean Reversion

In post#7 alphaomega shows a chart with a visual representation of the diffusion process (the sqrt thingy). To scale this parabola I wanted to use an estimation (forecast) of the volatility. I use the realized variance of the previous days computed ...

Simple Mean Reversion

alphaomega can you explain the rationale behind the 25%? Yesterday I took a pen an a paper and wrote down what happens if the price goes straight in one direction in the first part of the cycle. At 25% retracement the grid is in loss. Shouldn't the ...

Simple Mean Reversion

70% with 1:1 is next to HG! If I had such a system I wouldn't bother 2 minutes with money management: Kelly is 40% per trade with such a system!! Fiddling with the trades can only reduce your edge. The risk reduction would come from diversification ...

position-sizing/pyramiding strategies to decrease risk?