THE IDEA EXCHANGE - page 4

New comment
 
Mathemat:

P.S. The good thing about swing trading is that the result of a trade (modulo) is roughly proportional to its holding time (with a small spread, or rather variance). This implies that the deposit changes roughly in proportion to the time. And you have trades with a result of a few pips and a few hours of holding time. Why do that when you can ruthlessly cut off a trade with a move against you and open the opposite one?


I'm just trying to build on vinvin's thought. Let's see where the profit curve takes it. :)

If you implement stop-losses intelligently, there should be a tidy sum by the end of the championship...

 
NYROBA:

Mathematician, by 5 do you mean stop loss or take profit?

Nope. PF (profit factor) - see 'What do the numbers in the Expert Advisor test report mean'. It is the ratio of the total profit to the total loss during the testing interval.

But it is the simplest criterion, there are more serious ones. They are not available in the tester.

 
Mathemat:
NYROBA:

Mathematician, by 5 do you mean stop loss or take profit?

Nope. PF (profit factor) - see 'What do the numbers in the Expert Advisor test report mean'. It is the ratio of the total profit to the total loss during the testing interval.

But it is the simplest criterion, there are more serious ones. They are not available in the tester.


How much more serious could it be!?
 
Dear colleagues, has anyone used tic-tac-toe in trading? I am a lousy programmer, I can hardly write indicators. It would be interesting to create an EA.
 
Alex, I'll find Pardo's and post it in my optimisation thread. And there's no need to litter this thread.
 
Mathemat:
Alex, I'll find it at Pardo's and post it in my thread about optimisation. And it's not worth littering this thread.


You can email it to me at niroba@bk.ru/

thanks in advance ;)

 
Figar0:

Another at first, second, etc. - a delusional idea, a fried beard, meet "A no-loss system withmartingale elements for the poor"!

1. Expose the initial position ---- open 4 orders - BAY, BAY, sell, sell at the same time.
2. Suppose the price goes up. When Buy + Buy > Sell, we close three orders with a profit. 3.
There is one losing Sell order left. Since the profit has been taken as a result of the price going upwards, we open two orders again, buy, sell, and on the level of the unprofitable price set a pending Sell Stop order.
4.
a) The price goes upwards. In this case we repeat point 2 (BAY+BAY) and set the SELL-STOP. 2 - ( Buy + Buy > Sell, we close it with profit). Only one Pending Sell Stop Order remains from the entire position. The position is fully closed and we proceed with point 1 again.
b) The price goes down and activates a SELL STOP which is similar to step 1 (4 orders BAY, BAY, SELL, BOY) but the only difference is that BAY, BAY have been moved to the position to take profit and spread from SELL, BOY. Then again with point 2.
The idea is given as I first heard it from pidchybii on the Alpari forum. Bullshit? - Delusional... And if I tell you that based on its motives I finally succeeded in writing an Expert Advisor that successfully works in my real account for over a year - would you believe me? And this is exactly so...
I don't believe it... Or don't understand... Can someone translate this into clearer language?

1. We opened 4 buy, buy, sell, sell orders. Ok, let's do it that way, although we could also have two.
2. The price goes up. The formula buy + buy > sell is a bit weird, because two buys are always more than one sit, once we have passed the spread. Let's say 10 pips.
2,5. To close two bars in order to open them immediately is to pay additional spread, that is why we simply close one losing sell - we fixed loss in the same 10 points.
3. We already have two points, now we just set a Sell Stop at the level of the losing point.
4.
a) price goes up by 10 points, we close/delete everything. Once we have 20 points up, once we lose 20 points, and once we have a loss of 10 points from step 2, i.e. grand total - we have gained 10 points.
b) the price goes down - the paper loss decreases, but the paper profit also decreases twice as fast (if we had followed the initial procedure and closed/opened, the loss would have increased rapidly), when the price reaches the initial level, the pendulum will trigger and we will have a point 1 situation - two buys and two settlements with zeros and a 10 point loss on the balance.

The probability of going up is equal to the probability of going down, as a result we have a stable loss equal to the spread.
 
timbo: I don't believe it... Or don't understand... Can someone translate this into a clearer language?


Who says it's so simple and cloudless?) It's just an idea. ... Here's the EA's state where this idea is implemented to the letter head-on. I haven't used it that way, it took me about a month or two to get it right(real account)
Files:
massacre.zip  18 kb
 
Figar0:

Leonid's post that suggested mixing a lot of seldom but aptly, probably trading, into one EA has slipped and disappeared... (Already jammed, or I've become blind:)) The idea has been floating around for a long time and I want to try it, but I have only 2 Expert Advisors, so it would not be enough ...

I have one more subset of this idea: the market is conditionally divided into 3 phases: 1.UPTREND 2.DOWNTREND 3.FLAT. For each of these phases (for the first and second, one EA will do), we select an EA whose task is to chop a cabbage in "your own zone" and not to severely lose money in someone else's zone (that's all). And then there are 2 options: either parallelize them all, each of which works with its own orders, or transfer the order management from one to another estimating the current situation. Of course it may come down to "Who knows what is around the bend?", but taking into account that the change of trend (especially on higher TFs) occurs less frequently than its continuation, and all we have to do is to earn more than to lose, it may work.

Z.I. I will try it now.

No, Figar0, - did not squeeze! Just suddenly it seemed that the idea is of little interest to those present and unpromising ... So I deleted my post.

But to avoid reproaching me that I "squeezed it" - I will describe the simplest methods of practical realization:

There is an Envelopes indicator and we know the classical tactics of using it. But due to its structure it is too "sensitive" or it is too late with signals when the period is large. However, if we smooth out this indicator, the situation will change immediately! We select the deviation of the borders so that the borders cover only the tips of the candles and we enter by these crossings strictly following the trend. - set it (the trend) programmatically by the slope angle (for example) of these borders.

One version works in buy. The other version works for sell. At the same time we surprisingly miss losing trades at trend reversals! - No irony! And besides during a flat - no deals! (because the trend is set by the angle of slope!)

Here's a smoothed indicator chart - entry points are shown with arrows.

One more trick. You can use it as a filter or as a separate version. Stochastic. It should not be used according to classical rules but a little bit non-standard! Take a long period and enter at crossing not from outside to inside of overbought/oversold zones, but vice versa! - I have shown the entries with arrows in the stochastic window.

I have already made primitive Expert Advisors using both described methods. The results are satisfactory so far...

 

There are plenty of ideas. It's just a matter of getting it right.

1234567891011...36
New comment
Reason: