[Archive] Learn how to make money villagers! - page 770

 
GEFEL:


... it will only be painless for a few to lose that money.

...
I think KimIV is unlikely to face this, because the money is probably spread over a thick portfolio of multi-currency EAs, with many loosely correlated strategies and the diversification effect gives the possibility to work steadily even with increased risk for individual EAs.
 
khorosh:
I don't think KimIV is in any danger, because the money is probably allocated to a thick portfolio of multi-currency EAs, with many poorly correlated strategies and the divergence effect allows to work steadily even with increased risk for individual EAs.


I agree, I also preach diversification of trading different markets, instruments, and strategies. Then we have the resultant stable system.

But here some are arguing for one single strategy giving 100% per week with exorbitant risks...

 
GEFEL:

But here some people are arguing for one single strategy that gives 100% a week with exorbitant risks...

Don't overreact. If you still don't understand the risk and return of the EA in question, then I'm sorry and can't help you.

And everyone here also knows that portfolios are highly desirable. I think so, too.

 
OnGoing:

For the villagers to experiment with.

Found a curious hedge of 12 pairs. A couple of days already working in a narrow corridor of no more than +5 c.u. - 5 c.u. (with given lot 0.01).

Total spread including commissions -2.5 c.u.

In the trailer script in one direction, respectively, to speculate in the other, you need to reverse all positions to the opposite (I think most have no problem doing the second half).

The problem with this kind of design is that you have to close them all almost instantly, and the more orders there are, the worse it will be. Especially if the broker will hold back closing.
 
4x-online:
The problem with this kind of construction is that they all have to be closed almost immediately, and the more orders there are, the worse it will be possible to do it. Especially if the broker will hold back the closing.

The fact is that each of the positions included in a hedge has a small weight compared to the aggregate, and even if there is a move against us at closing time, it is very small. And sometimes it happens even in our favor.

Besides, if we correctly arrange the closing, correctly handle busy flow and other errors, we can reduce the time of closing to a minimum.

Plus it makes sense to leave a small profit margin (above the norm) just for the execution costs.

 
OnGoing:

The point is that each of the positions included in a hedge has a small weight compared to the aggregate, and even if there is a move against us at closing time, it is very small. And sometimes it happens even in our favor.

Besides, if we correctly arrange the closing, correctly handle busy flow and other errors, we can reduce the time of closing to a minimum.

Plus it makes sense to leave a small profit margin (above the norm) just for the execution costs.

This is all understandable. But I'm a pessimist about the real world and I prefer to, ahem...ahem...overdo it. :)
It's not very clear what "excess profit" means. The range specified is +5 -5. What do the tests say for at least 3 years? Is it a stable structure? Or is it possible to fly up sometimes?
 
4x-online:
It's all understandable. But I'm a pessimist in respect of real account and I prefer to, ahem...ahem...overdo it. :)
It's not very clear what the "excess profit" means. The range specified is +5 -5. What do the tests say for at least 3 years? Is it a stable structure? Or is it possible to fly up sometimes?

Haven't tested it on history. I've only been watching it for the last three days. But I can already see that the potential is there. Those who understood it, I think will do the best possible execution to keep the probability of collisions to a minimum.

I see the target in this case in the range of 2-3 c.u. A margin of about 0.5 c.u. For quicker execution of the target it is better to enter from the limits of the corridor. The latter ones can be set by monitoring of a constantly open hedge on a demo.

Of course +5 and -5 was only for estimation, and real limits of the range are very tentative.

I personally see manual trading as the most promising with this strategy. It is enough to make 1-2 profitable entries per day and have 3-5% of deposit per day from it. Risks may well allow to use the appropriate mm for this.

 
OnGoing:

Haven't tested it on history. I've only been watching it for the last three days. But I can already see that the potential is there. Those who understood it, I think will do the best possible execution to keep the probability of collisions to a minimum.

I see the target in this case in the range of 2-3 c.u. A margin of about 0.5 c.u. For quicker execution of the target it is better to enter from the limits of the corridor. The latter ones can be set by monitoring of a constantly open hedge on a demo.

Of course +5 and -5 was only for estimation, and real limits of the range are very tentative.

I personally see manual trading as the most promising with this strategy. It is enough to make 1-2 profitable entries per day and have 3-5% of deposit per day from it. Risks are good enough to use the appropriate mm for this.

It is better to run tests to understand the behaviour of the "portfolio" in fast markets. No one has ever cancelled the picks (including those drawn by the broker). Plus one can try to find the "corridor" timeframes in tests. If it exists. I.e. is there a certain corridor boundary forming at one or another terminal time, since the entries are supposed to be made manually/script as far as I understood. In short, tests are zero-sightedness even in this case and useful statistics can give.
 

Alternatively, instead of a permanently open hedge, a simple indicator can be made. I.e. display the same number on the chart - the total profit of the "virtual" hedge, taking as a reference point the moment of the indicator initialisation and remembering the price values for all symbols.

Plus you can monitor the range boundaries by displaying them as well. I.e. maximum and minimum values of the total profit for all time, or for a period of time.

 
4x-online:
It is better to run tests to understand the behaviour of the "portfolio" in fast markets. No one has ever cancelled the picks (including those drawn by the broker). Plus one can try to find the "corridor" timeframes in tests. If it exists. I.e. is there a certain corridor boundary forming at one or another terminal time, since the entries are supposed to be made manually/script as far as I understood. In short, tests are zero-sightedness even in this case and useful statistics can give.
I was thinking about the best way to perform such analysis. To run a simple opening/closing test on history is one thing. But identifying the pattern of behavior and tracing the time of movement within a range is another thing.