Help with backtesting¡¡¡¡

[Fran8]

Hi

Please I need help with back testing a cross over moving average strategy, the problem is that I use freestockcharts which has not got backtesting capabilities and if it had it would not matter since I do not have a clue on how to program.
In the book “all about market timing” it says that the best moving average for intermediate term is the 66 ma and for the long term is the 132 ma.

I would like if possible if someone could back test a crossover system of the 66 and 132 ma for simple moving average, exponential moving average and if possible front weight moving average to see with which kind of average works best and if possible do the same with the 50 and 200 crossover ma to compare it with more typical time frames.

I thought on back testing it with ETF that represent stocks (SPY), Commodities (DJP), Bonds (TLT), Forex (UUP) and gold (GLD).

Thanks any help would be appreciated

 

[tony1979]

Do it by hand.

 

[the hare]

Better still, dont do it at all !

I suppose its a stage everyone goes through, presuming that you could make money from a moving average cross, if only you could work out the right period lengths to use !

If you really must take this approach, use walk forward optimisation over the last year, find the optimum period lengths, and then fade it.

The equity curve will be up and down like a tarts bloomers, but it should be profitable over the long run.

 

[tony1979]

Better still, dont do it at all !

I suppose its a stage everyone goes through, presuming that you could make money from a moving average cross, if only you could work out the right period lengths to use !

If you really must take this approach, use walk forward optimisation over the last year, find the optimum period lengths, and then fade it.

The equity curve will be up and down like a tarts bloomers, but it should be profitable over the long run.

Not strictly true if traded correctly...daily/weekly period moving average crosses work, and will continue to work, if they're practiced with proper money management. The problem with them, is that they tie up capital, and there are large and prolonged drawdowns involved, which have to be overcome - as is standard for any trend following method.

 

[Fran8]

Thanks for your recomendations but does anyone have a charting software were the strategy can be backtested?

 

[Dominator4fx]

Not strictly true if traded correctly...daily/weekly period moving average crosses work, and will continue to work, if they're practiced with proper money management. The problem with them, is that they tie up capital, and there are large and prolonged drawdowns involved, which have to be overcome - as is standard for any trend following method.

I have seen many traders who take unwarranted risks into the markets and this is the main reason why they get so much draw-downs into their trades.

 

[tony1979]

I have seen many traders who take unwarranted risks into the markets and this is the main reason why they get so much draw-downs into their trades.

That can be one reason, although if you trade a statistically based method, drawdowns are unavoidable. In fact, if you try to avoid them, they'll be bigger.