Archive for December, 2010

As a sidenote, we mention that it is probably our last post in 2010. We started this blog on 28th January 2010. In these 11 months, we wrote 44 different articles in this blog. This one is the 45th. We didn’t aim to find the most profitable ANN (Artificial Neural Network) approach in the first […]


We were confident that the dayOfTheWeek ANN works. We had successful backtests earlier. Those showed that using only the dayOfTheWeek as input, we can achieve about 12% GAGR and 150-200% TR. For example, see the 10 members homogeneous ensemble in this posthere. The D_stat/CAGR/TR numbers were 51.43%/11.97%/126%TR. Therefore we were surprised when backtesting it for […]


This is a test with only the 1 dimensional DayOfTheWeek input. For the first day prediction, we run 50 different ANN trainings. They gave 50 slightly different forecasts for the first day of course. As we increased the maxEpoch, we decreased the randomness, the std. The same in chart form: Notes: – The dayOfTheWeek 1 […]


subject: FXE (Euro) input, ensemble of 11 This is the continuation of a previous post (see here), in which we studied the discretization of the FXE input. Our previous backtest was performed without ensembles, so our backtests were very volatile. We have only 1 input now, the FXE current day return. We try to discretize […]


This is the continuation of a previous post (see here), in which we studied the discretization of the RUT index current day change input. Our previous backtest was performed without ensembles, so our backtest were very volatile. We repeat most of the text of that post now again. We have only 1 input now, the […]


So far, we could achieve about 52% D_stat. That means we could correctly predict the direction only 52% of the time. That is not too much you may reckon. It only means that during a 250 trading days of a year, our edge is only 2%. That is 5 days. So, we may speculate that […]


Our research was motivated by Varadi post here 1. Varadi achieved 49% CAGR gain and 56% D_stat in 2010!!. Quite remarkable. But with fixed rules and in hindsight, it is easy. 2. Varadi is right with his strategy in 2010, but we prove that in the last 5 years, the relationship is exactly the opposite. […]