Re: Portfolio Expert
You don't need any external services for the Portfolio Expert. It calculate all signals locally in MT.
Re: Portfolio Expert
Popov wrote:
You don't need any external services for the Portfolio Expert. It calculate all signals locally in MT.
ohhh, wow, that is new? Last year i have needed this url for running the portfolio ea 
- sidmcfarland
- Licensed Member
- Offline
Re: Portfolio Expert
Assume I have a portfolio expert containing three strategies. And let's assume that at the start of a new day the first strategy opens a long position of 0.1 lots. At the same time the second strategy opens a short position of 0.1 lots. Simultaneiously, the third strategy opens a long position of 0.1 lots. So now I have three positions open. Two are long and one is short. Overall I'm long 0.1 lots. However, I paid the spread/commission for three different positions when, in effect, what I have now is a long 0.1 lot position (since the other two positions cancel each other out). This seems wasteful to have strategies open positions that cancel each other out. Not only have I tied up margin on my account, but I've paid my broker more fees that I needed to. Instead of opening the three positions described above, I could have just opened one long positions and avoided the fees and margin resulting from the other two strategies.
Is there a way to modify the portfolio expert so that it calculates the net of all positions that each strategy wants to have open, and just performs one trade for that amount?
The only thing I can think of is to run the portfolio expert in a demo, then create a custom EA to monitor the open positions of that demo account, and trade the net sum of all those open positions in another demo account... and then provide a signal on that demo account that I can subscribe to in my live/real account. But this seems like a hack.
Any suggestions here? The portfolio experts seem highly inefficient in the sense that they open lots of trades in the opposite direction at the same time. How can we resolve this to stop wasting money on spreads and commissions?
- Popov
- Lead Developer
- Offline
Re: Portfolio Expert
Sid, this is something I'm thinking for 5 years.
Now the Portfolio Expert is a combination of individual strategies that are combined in a single Expert Advisor.
The system you are thinking for is different. I posted several times ideas of a new application that trades only the aggregate position only, but there was no huge enthusiasm in the community. Anyway, I plan to release such an application in the near future.
130 2020-02-20 08:32:10 (edited by hannahis 2020-02-20 12:09:33)
Re: Portfolio Expert
Popov wrote:
Sid, this is something I'm thinking for 5 years.
Now the Portfolio Expert is a combination of individual strategies that are combined in a single Expert Advisor.
The system you are thinking for is different. I posted several times ideas of a new application that trades only the aggregate position only, but there was no huge enthusiasm in the community. Anyway, I plan to release such an application in the near future.
Here are just some thoughts...
If we only look at trading positions in terms of Long (Buy) vs Short (Sell) then it is indeed wasteful to have "extra" trading positions that cancel each other out.
However, each strategy contains 2 Components 1) Open entries and 2) Close entries and Strategies have different trading objectives (scalper, ranger, breakouts, position trade etc), in other words, strategies can either be for short term or long term trading. Consequently, these strategies have different closing timing (purpose).
Imagine I would like to have a Long position with a trade duration of 1 day and a short position with a trade duration of 1 wk. Since these 2 positions (Long and Short) cancel each other out, I ended up missing out 2 opportunities to make money. One trade for ranging and another trade for position trading.
For example, I may have a Buy position with the aim to earn a couple of small pips and have another opposite position (Sell) that is position trading for long term gain. Thus while I may open a Long term trade (Sell, position trading), I may at the same time want to take advantage of the ranging markets to open Buy trades (for the short term). These short term trades duration are shorter than the long term trade, thus these positions will be closed earlier than the long term position, which may take weeks or months to close.
Therefore, combining 2 different strategies (Long and Short) and crossing out each other (with no trades) will defeat the purpose of developing different strategies as a means to diversify the portfolio (with different trading strategies)
- Blaiserboy
- Junior Part-Time Aspiring Space Cadet
- Offline
Re: Portfolio Expert
I think H has a valid point, this would have to be carefully constructed
132 2020-02-20 15:54:22 (edited by ats118765 2020-02-20 16:07:37)
Re: Portfolio Expert
hannahis wrote:
Here are just some thoughts...
Therefore, combining 2 different strategies (Long and Short) and crossing out each other (with no trades) will defeat the purpose of developing different strategies as a means to diversify the portfolio (with different trading strategies)
Great points Hannah
Here is a visual example.
https://atstradingsolutions.com/wp-cont … Trader.png
In this trend trading example we have been riding longs for say multiple years with a solid uptrend....but now the trend is turning and we are a long way from our long exits. Normally we would have to sacrifice the unrealised profits when the trend turns...but if we allow for short trades during this phase we 'hedge' the equity curve during trend turning points and prevent a significant deterioration in equity. This significantly smooths the equity curve for long term trend traders.
So please if this option is included....then make it optional. :-)
133 2020-02-20 17:50:42 (edited by hannahis 2020-02-20 18:25:25)
Re: Portfolio Expert
sidmcfarland wrote:
The only thing I can think of is to run the portfolio expert in a demo, then create a custom EA to monitor the open positions of that demo account, and trade the net sum of all those open positions in another demo account... and then provide a signal on that demo account that I can subscribe to in my live/real account. But this seems like a hack.
Any suggestions here? The portfolio experts seem highly inefficient in the sense that they open lots of trades in the opposite direction at the same time. How can we resolve this to stop wasting money on spreads and commissions?
Hi Sid,
Select the Best EA
You may want to consider examining your trades performance in greater details. If your portfolio opens lots of "redundant" trades that eventually cross each other out. Surely at the end of the day, there are some EA that shows better results than others. Simply select the best and reduce the number of EA in each Portfolio, you will then reduce redundancy.
Delete Bad EA from Portfolio Expert (PE)
When you have 2 EA with opposition directions, won't one of them emerged as the winner (eg if both are short term trades) or ideally both emerged as winners (one closed earlier, short term, while the other waited for the trend to turn around and take a kill shot, earn you pips).
In the case of similar trading strategies, those EA that opened the wrong opposite trades will have poorer performance statistics as compared to those opened in the correct direction. Hence over the time, you will eventually delete those poor performing EA out of the Portfolio, hence reduced the issue of having too many redundant trades.
Examine EA's Trade Performance Metrics
If you were to save your trading performance statistics in excel file and compare their weekly performance, you may also noticed a number of similar EA that are highly coorelated, which you may also want to consider trimming the number down. Or using the EAS tool, decrease the Correlation Analysis Threshold. However, do bear in mind, if you decrease the Correlation Analysis Threshold, you would also ended up with EA with different opening/closing rules which may ended up with more opening opposite direction trades.
Increase or Decrease Correlation Analysis Threshold
If you really don't want to have so many trades that cross each other out (as a means of diversification), you may want to consider increase your correlation analysis threshold to tolerate EA with similar rules.
Select Portfolio Expert based on Metrics
I would prefer to observe my trade performance and delete the EA according to their performance instead of "blindly" crossing them out using a net trading sum software/application and I would then be clueless which is a better EA to keep in the Portfolio.
At the end of the day, we still need to evaluate each EA's performance and decide which to delete and which to keep. By crossing the EA out of trades, I will never know the EA's (true) performance metrics and it would make my EA selection task very difficult. And won't knowing EA's true trading performance helps us make better selection decision?
Once you have made your (careful and detailed) selection of EA to be included in your Portfolio EXpert (based on real or demo performance metric), you would have eventually trimmed and removed all other unnecessary redundancy and you will not depend on any application/software to help you cross out the net difference.
Therefore, your experience of having so many unwanted redundancy in your Portfolio Expert only goes to show that you are still at the early stage of your Portfolio Expert (PE) selection and hence your PE is rather "raw and untested". Do give yourself some more time to observe their trade performance and based on their trading metrics (preferably demo account), you will find it easier to decide which EA to keep and which to delete and your early woes will be unfounded and you will then realised that having an application to cross out the trades would be unnecessary, if not, detrimental.
- sidmcfarland
- Licensed Member
- Offline
Re: Portfolio Expert
Popov wrote:
Sid, this is something I'm thinking for 5 years.
Now the Portfolio Expert is a combination of individual strategies that are combined in a single Expert Advisor.
The system you are thinking for is different. I posted several times ideas of a new application that trades only the aggregate position only, but there was no huge enthusiasm in the community. Anyway, I plan to release such an application in the near future.
Popov, thanks for considering this. Please let us know if any progress is made. In the meantime, I created an EA that copies trades from one chart (demo account running the portfolio advisor) to another chart (live account) and "flattens" the positions in the process by computing the net of all longs and shorts. This prevents opening wasteful opposing positions in the live account. I would prefer a solution that doesn't require two accounts and a custom EA.
135 2020-07-10 05:06:41 (edited by sidmcfarland 2020-07-10 16:52:20)
- sidmcfarland
- Licensed Member
- Offline
Re: Portfolio Expert
hannahis wrote:
For example, I may have a Buy position with the aim to earn a couple of small pips and have another opposite position (Sell) that is position trading for long term gain. Thus while I may open a Long term trade (Sell, position trading), I may at the same time want to take advantage of the ranging markets to open Buy trades (for the short term). These short term trades duration are shorter than the long term trade, thus these positions will be closed earlier than the long term position, which may take weeks or months to close.
Thanks for the example, but I think you misunderstand something about hedging. In your example you have a long term sell position and a short term buy position. If both positions are the same size then they cancel each other out for as long as that short term buy position remains open. If one profits the other looses. So why have them open at all. Instead of opening a short term buy position, get out of the market. Close your long term sell position until you would have closed that short term buy position (had you opened it). At that point you can re-open your long term sell position. This is effectively the same as what you described but doesn't tie up any margin. This is an often misunderstood point. Hedging is completely unnecessary. Instead of opening a new buy position just close your existing sell position. The result is the same. Think about it. For that short period of time when you have a long term sell position open at the same time as a short term buy position, how much money are you making while both positions are open? It's $0, because any gains from one position are offset by losses on the other. So, why be in the market at all? Use that money elsewhere.
And here's the main point and the reason I'm asking about this. If I have a portfolio EA, which of course contains several trading strategies, it's possible, even likely, that two of those strategies may decide at some point to hold a position at the same time but in opposite directions. Why would I want to waste money paying the spread and other fees opening two opposing positions created by a portfolio EA? It would be better wait until one of those strategies gets a close signal. Then and only then would it be wise to open the position for the other strategy. In doing so, I've only paid the spread and fees once instead of twice, yet effectively executed both strategies in terms of their exposure to the market. A lot of money is unnecessarily lost on the spreads/fees using portfolio EAs due to strategies within that EA that will sometimes work against each other. I am simply looking for a solution to avoid these unnecessary costs. In doing so, by the way, the portfolio EAs would also become compatible for U.S. accounts the prevent hedging. And if done well, even FIFO rules could be satisfied.