Topic: Portfolios testing?

Hi

I am aware of Multi Markets function.

Do you plan to add function of strategy generating, optimizing directly for chosen basket of markets? Now software just work with one market and then one have to go to Multi Markets to see if it is any good.

Re: Portfolios testing?

Once we start to process several instruments simultaneously, the program will require vast memory and will be very slow.

I can understand your idea, not everyone has the computer power to do this.

Myself I work on one currency pair at a time and I have had some great results over a few years. I use simple strategies with few indicators and higher time frames. So far I have not needed anything to powerful to get these results.

One thing to keep in mind about baskets is, you miss a huge amount of entries for each instrument. I have a number of ea's that I have created for MT5 for multiple instruments, the results are less than satisfactory.

My 'secret' goal is to push EA Studio until I can net 3000 pips per day....

3 (edited by archeolog108 2017-12-03 18:41:21)

Re: Portfolios testing?

Blaiserboy wrote:

Once we start to process several instruments simultaneously, the program will require vast memory and will be very slow.

I can understand your idea, not everyone has the computer power to do this.

Myself I work on one currency pair at a time and I have had some great results over a few years. I use simple strategies with few indicators and higher time frames. So far I have not needed anything to powerful to get these results.

One thing to keep in mind about baskets is, you miss a huge amount of entries for each instrument. I have a number of ea's that I have created for MT5 for multiple instruments, the results are less than satisfactory.

Diversification is the only free lunch in finance. Why not use it?
If strategy works in one market and not in other then it is not robust and it is just curve fitting.

Trading a trend following system on a single market or only a few different markets is suicidal. There may be long periods, even years, where there simply are no trends in any given market or asset class. The key idea is to trade many markets covering all asset classes at the same time. If you fail to do so, this strategy will simply not work. Andreas Clenow

So I suppose you are talking about convergent system.

Re: Portfolios testing?

Well, from what I have seen, the big day traders in Wall Street specialize in one or two instruments such as S&P 500 and use size to make their money. People are trading as much as 1000 contracts per order, they only need a liquid market.

Diversification is great for those that wish to be at their compuiter 24/7, doing research and managing, for me the simple approach is far easier....

I use very few indicators, I do not worry about curve fitting, I use only strategies that have a high win rate.

Well, regarding Mr Cleanow, perhaps he will look at EURUSD and he will notice that there is always a bit of money of the table.

Personally, I like simple, I like to make money every day and I have done my time sitting at the screen.

I am continually generating new strategies that have high acceptance requirtements, I limit the number of pairs to one. For a time i was trying to diversify and finally realized that I could do quite well with only one pair, and one time frame.

My 'secret' goal is to push EA Studio until I can net 3000 pips per day....

5 (edited by sleytus 2017-12-04 03:19:00)

Re: Portfolios testing?

Do you plan to add function of strategy generating, optimizing directly for chosen basket of markets?

Well -- you can sort of already do that.  I stitch together 6 months of EURUSD and 6 months of USDCHF in a single *.csv file and use that for generating and optimization.  It's a "hybrid" data set -- I've posted about this before.  When I follow-up with a Multi-Market test the statistics are usually excellent for *both* EURUSD and USDCHF.

Diversification is the only free lunch in finance. Why not use it?

With regards to diversification -- I think about it, but if I think too hard then I get confused about what it is I'm trying to diversify.  I mean, it's not like how you are supposed to eat a diversified diet of  fruit, meat, dairy, vegetables, etc.  When you refer to different instruments it's not like drums, guitars, saxophones -- forex "instruments" are just sequences of numbers with different patterns.  And though the current pattern of EURUSD numbers may differ from the current pattern of USDCHF numbers, I suspect the same patterns probably exist in both at some time in their histories.

When the topic of "diversification" comes up then I'm interested to understand more about what exactly it is we are trying to diversify.

If strategy works in one market and not in other then it is not robust and it is just curve fitting

It's my opinion there is nothing wrong with "curve fitting" == optimizing == training.  Curve fitting is your friend -- it's gotten a bad rap.  When you train a strategy you are simply solving for the best indicator parameters that give the best stats for the input data you are feeding into your formula (strategy).  The alternative is to close your eyes and pick numbers from a hat.  If you want your formula to fit well with multiple "instruments" then the data set you use to train your formula should include patterns from the different instruments.  I mean, how else could a formula be expected to work well with data patterns it had never seen before?

Re: Portfolios testing?

The authority on diversifiication, Mr Clenow, is a manager of large funds. He is looking for secure returns for his investor clients.

His  business is quite a bit different from being a forx trader, I assumne that his office has vast computer power to generate their portofolios.

I would not expect that FSBPro will ever be able to handle huge numbers of instruments.

I suggest that you develop a group of single strategies, perfect that and then when competent with that stage, there are softwares on the net that will allow constructing baskets. I have looked at a couple of those, I am not really interested.

One thing that you can do is to have a look at the ea's for sale and choose an MT5 ea that will trade several instruments and experiment with it. You will see that there will be trouble with the number of digits, the stop losses and the missed deals. Believe me, I have had a long look at baskets for forex and it is not worth the bother.

My 'secret' goal is to push EA Studio until I can net 3000 pips per day....

7 (edited by archeolog108 2017-12-08 18:46:10)

Re: Portfolios testing?

Blaiserboy wrote:

The authority on diversifiication, Mr Clenow, is a manager of large funds. He is looking for secure returns for his investor clients.

His  business is quite a bit different from being a forx trader, I assumne that his office has vast computer power to generate their portofolios.

I would not expect that FSBPro will ever be able to handle huge numbers of instruments.

I suggest that you develop a group of single strategies, perfect that and then when competent with that stage, there are softwares on the net that will allow constructing baskets. I have looked at a couple of those, I am not really interested.

One thing that you can do is to have a look at the ea's for sale and choose an MT5 ea that will trade several instruments and experiment with it. You will see that there will be trouble with the number of digits, the stop losses and the missed deals. Believe me, I have had a long look at baskets for forex and it is not worth the bother.

If you want to stay retail trader, you are welcome. But there is a reason why institutions make money and are rich and not retails traders: http://www.followingthetrend.com/2016/04/you-cant-beat-all-the-chimps/

Re: Portfolios testing?

Well, for now, I will stay a simple old retail trader doing my simple retail trading routines....

My 'secret' goal is to push EA Studio until I can net 3000 pips per day....