Avalanche - page 66

 
JonKatana писал(а) >>

The idea is not mine, but you are right - in this way the Avalanche is always entered. But the Avalanche is harmonic and automatically reacts to the market in proportion to its impact. The stronger the opposition of the market (or AC), the greater will be its loss.

If the first (initial volume) order triggers and the price obediently moves in the same direction when trading in a simple mode (entering the corridor), you calmly close your profit and move your orders to new positions. The profit is not large due to the small volume of the initial order.

If the price turns around, then another one, and then another one, trying to resist you - then the market gets itself into an even worse situation. The "avalanche" increases the volume of orders exponentially with each reversal and when.......

stop... no further, you have stopped at the main thing, next is sweet fantasy. Agree that the "and when" may not come and reversals may continue. It is the amount of them (evil) that needs to be dealt with.

 

Don't kick me, I may be silly, but it is only for "brainstorming" purposes... If we have a standard corridor, order placement, volume increase algorithm, to reach b/y level, which is at a distance from "corridor" borders, equal to its width... At opening of the first and following positions, trailing pending order of increased volume should be used. Under this condition, the farther price passes in the direction we want, the closer an opposite pending order comes, so when it triggers, we obtain a reversal, but with proportionally smaller volume of the opposite position and with smaller locked loss... Something like this .... I need to gather mush in my head :)
I.e., the idea is that every pip in our direction, takes away a pip from possible locked loss from price movement not in our direction...

 
sever29 >>:

стоп... дальше не надо, Вы остановились на главном, дальше сладкие фантазии. Согласитесь, что "и когда" может и не наступить, а развороты могут продолжится. Именно с их количеством (зло) надо бороться.

Turns cannot continue indefinitely, otherwise anyone would make continuous profits by placing inverted (Buy Limit and Sell Limit) orders on the channel boundaries with Take Profit on the opposite channel boundary. A large number of reversals are the exception. You can minimize them by aggressively increasing the opposite volume - then profit will begin after the price passes just a few points from the border of the channel. For example, given the distance between the orders of 40 points and the initial volume of 0.01, you need to set an opposite order with the volume 41 times larger than the initial order (i.e. 0.41), in order to make a profit after the price has passed ONE point from the border of the channel. And the profit will be a huge volume, 40 times greater than the initial one! So, what are the chances, by accidentally placing an order, to get to the ideal position the price will not move a single point further but will instead reverse and go 40 points in the opposite direction?

 
sever29 >>:

Сильно не пинайте, может глупость горожу, но это исключительно в целях "мозгового штурма"... А если при стандартном корридоре, расстановке ордеров, алгоритма увелечения их объема, для достижения уровня б/у, который находится на расстоянии, от границ "корридора", равном его ширине... при открытии первой и последующих поз, использовать трейлинг отложенного ордера увеличенного объема. При таком условии, чем дальше пройдет цена в нужном нам направлении, тем ближе подтянется противоположно направленный отложенный ордер, соответственно при его срабатывании мы получаем разворот, только уже с пропорционально уменьшенным объемом противоположной позы и уменьшенным локированным убытком... Как то так.... Надо собрать кашу в голове:)

There will be an imbalance of orders placed at different distances and different price positions. You can only pull up the first order that failed after the initial order was activated - until a channel is formed, it can be narrowed down. This has already been covered in the thread.

 
sever29 >>:

Сильно не пинайте, может глупость горожу, но это исключительно в целях "мозгового штурма"... А если при стандартном корридоре, расстановке ордеров, алгоритма увелечения их объема, для достижения уровня б/у, который находится на расстоянии, от границ "корридора", равном его ширине... при открытии первой и последующих поз, использовать трейлинг отложенного ордера увеличенного объема. При таком условии, чем дальше пройдет цена в нужном нам направлении, тем ближе подтянется противоположно направленный отложенный ордер, соответственно при его срабатывании мы получаем разворот, только уже с пропорционально уменьшенным объемом противоположной позы и уменьшенным локированным убытком... Как то так.... Надо собрать кашу в голове:)
т.е. мысль такова, чтоб каждый пипс в нашу сторону, отнимал пипс от возможного локированного убытка от движения цены не в нашу...


There is already a similar option... One of the forum members has written to me in person... chatted... almost finished exactly this version. No worse and no better...
The difference is that it's much quicker to get into the loki... ...and the multi-branched ones... If the price has gone where it needs to go, it jumps out faster... And if it doesn't go where it's supposed to go, it flips faster.
Same eggs but hairier...
 
sever29 писал(а) >>


The flexibility of thinking is good. Only, unfortunately, it's the same eggs only from the side. You didn't make a profit. This is 0. You have covered one position with Takei, and the other one, equal in number of pips, only loss-making, has not. Moreover, on the contrary, artificially, you have added to your troubles, like a "second knee of martin" on nothing. Compare the same situation, only with one buy or sell order. For example, you have guessed, and there is no need to roll over, and calmly close at take. Suppose not, then the knee. What is the difference between the first and the second variant? When in the second one you take a profit of 50%, but in your ( first variant ) you always get into martin?

I absolutely agree with you. I think the better variant is the one I described above. The first market order is opened on the basis of the analysis and then it is covered by locking orders according to the avalanche algorithm. There is no initial idle run like in a classic avalanche. I gave a picture of the run with the period of 01.01.09 above.
 
lexandros писал(а) >>

There already is a similar variant. One of the forum members wrote to me in a private message... chatted... almost finished exactly this variant. No worse and no better...
The difference is that it's a lot quicker to get into the loki... ...and the multi-branched ones... If the price has gone where it needs to go, it jumps out faster... And if it doesn't go where it's supposed to go, it flips faster.
Same eggs but hairier...

on a trailing pause, did the b/o level change? It shouldn't. Another point... Corridor width should be constant, i.e. when trailing stop-loss order triggers (with proportionally decreased lot), the opposite one should be set at the distance equal to the width of initial corridor.

 
khorosh писал(а) >>

I absolutely agree with you. I think the alternative described above is better. The first market order is opened using the market analysis and then it is compensated by locking orders according to the avalanche algorithm. There is no initial idle run like in a classic avalanche. I gave a picture of the run with the period of 01.01.09 above.

If memory serves, you suggested a trend on ZZ, you created a separate thread... In my opinion, your method was objectively criticised. But this third one is important to me personally. I am interested in your Expert Advisor, it is a 1-2-3 pattern. Do you know it? I have an opinion that it goes well with "avalanche" or vice versa.

 
No... the b/w level has not been changed... there is no level as such... I.e., closing is not at any particular price, but when a given profit is reached...
Clinging to a specific price level is the worst evil, IMHO... The price may reach this level in a couple of years, or even never.
The width of the corridor is of course constant... There is one more trick though... I can't explain the TS in full, for obvious reasons - I wasn't the author... If abstractly - there are fills by trend... That is, it almost always goes out... It is difficult to go into a margin call... But the drawdowns so far are very large - this is not good.
 
sever29 >>:

если память не подводит Вы предлагали по ЗЗ тренд, ветку отдельную создавали... На мой взгляд Ваш метод объективно подвергся критике. Но эт третье, лично для меня важна мысль, с Вашей подачи присматриваюсь к индикатору- патерн 1-2-3. Знакомы? Есть мнение, что он органично подходит "лавине" или наоборот она к нему.


No... It's not him about the zigzag... it's Kharko... It's just the nicknames are very similar :) and I confused them too at first :)