Avalanche - page 103

 
khorosh писал(а) >>

QF on the interval 01.01.08 - 01.04.10 is 3.43. max drawdown 6930.87 (30.51%). In the interval from 01.01.09 to 01.04.10 maximal drawdown is 1738.26. Parameters are certainly not brilliant, it should be expected due to such a risky strategy. But for those who like risky trading, I think, that will do.

Now think about it: in the tester you have found (in fact, adjusted) parameters that could have had such results in the past. And this is the best possible result out of the fitted ones.
In fact, the annual TP is slightly greater than 1, which means (for the uninitiated) that the risk at any given moment is equal to the profit you plan to make in a year. If we consider that in real trading, things will go worse than in tests (the market won't repeat exactly the model during testing, brokerage company will nibble from deposit on every opening/closing ...), the annual EF will really be lower than 1, maybe even several times.
.
I wrote an Expert Advisor on my knee to test the very idea of Avalanche purely for the tester using my author's algorithm, but with conversion of lots to net-trading. The essence of the net trading system does not change, but only creates favorable conditions for Avalanche (margin and spread savings on the counterclosing). So even after the optimization for the same period (early 2008 to date) the results are more than modest. I will not describe in detail - everything can be seen on the header of the report. What deposit doubling the author was speaking about? Where did he/she get it from?

Lavina
InvestTechFx-Server (Build 225)


Symbol EURUSD (Euro vs US Dollar)
Period 5 Minutes (M5) 2008.01.02 09:00 - 2010.03.31 23:55 (2008.01.01 - 2010.04.01)
Model All ticks (most accurate method based on all smallest available timeframes)
Parameters Channel=110; Koeff=2;
Bars in history 165721 Modelled ticks 17669226 Modeling quality 90.00%
Chart mismatch errors 0
Initial deposit 1000.00
Net profit 1639.78 Total profit 6848.48 Total loss -5208.70
Profitability 1.31 Expectation of winning 1.77
Absolute drawdown 14.20 Maximum drawdown 705.30 (31.61%) Relative drawdown 31.61% (705.30)
Total trades 929 Short positions (% win) 414 (76.57%) Long positions (% win) 515 (59.81%)
Profitable trades (% of all) 625 (67.28%) Loss trades (% of all) 304 (32.72%)
Largest profitable trade 348.80 losing deal -177.60
Average profitable deal 10.96 losing trade -17.13
Maximum number continuous wins (profit) 13 (141.70) Continuous losses (loss) 1 (-177.60)
Maximum Continuous Profit (number of wins) 370.60 (4) Continuous loss (number of losses) -177.60 (1)
Average continuous winnings 2 Continuous loss 1



Once again, I would like to emphasise that this is the best result, which was achieved by optimising the width of the channel.
In essence, it is an adjustment.

Files:
lavina.rar  36 kb
 
JonKatana писал(а) >>

1. What a "crumbs" to open an order with a volume of 4 lots! If you do not have enough deposit to open an order of the required volume, it is foolish to open "any" order. Under your conditions you need to close ALL the orders on the border of the channel (in your example +20). Then there will be no losses from the top direction (more precisely, they are equal to the spread), the loss from the bottom orders (4 lots) will be equal to the whole width of the channel (40 points in your example), but the margin from both orders (for volumes of 2 + 4 = 6 lots) will remain in your deposit.

2. Your loss will be 118430 - 117266 = 1164 RUB, i.e. 1% of your initial capital.


Compare the introductory part (1.) in highlighted font and the summary part (2.) also in highlighted font. Can you see the difference? In the first part, you closed a losing trade with 4 lots, which equals 46 672 rubles, buy without loss (spreads are not taken into account for convenience), and in the second part of your post, what nonsense, what 1164 rubles? Do you understand that I have fixed my loss in 4 lots, long 40 points? In money, it would be 1164rub. ? Why are you twisting and building a pyramid of numbers? Don't complicate things.
That is what I wanted to hear. What's next? you lock in a loss, release the margin, wipe the sweat off your forehead.... then what?

 
goldtrader писал(а) >>


The profits could be greater, I think by orders of magnitude. Profits are easier.

 
sever29 >>:

Сравните вводну часть (1.) выделенным шрифтом и резюлятивную (2.) также выделенный шрифт. Видете разницу? В первой части Вы закрыли убыточный сел объемом 4 лота, что равно 46 672 руб., бай в безубытке (спреды в расчет для удобства не беру), а во второй части Вашего поста что за ерунда, какие 1164 руб.? Вы понимаете, что я зафиксировал убыток в 4 лота, длинной в 40п.? В деньгах это будет по-Вашему 1164руб. ? Что передергивать и городить пирамиду из цифр, не усложняйте, все просто.
Именно это я и хотел услышать. Что дальше? зафиксировали убыток, высвободили маржу, вытерли пот со лба.... дальше что?

There is no error in the calculations. Under your conditions, there is 70,000 roubles in the account before the release of the collaterals. Therefore, after the losses are fixed and the collateral is returned, the deposit does not change much. You set these conditions - blame yourself. Next time count it BEFORE you write anything.


Once the margin is released the Avalanche starts all over again. Did you understand the words "close ALL orders" correctly? If not, read it carefully. All means all - both Buy and Sell. That is why you do not have any orders, neither open, nor pending, after the return of the collateral. The chart is clear, you have 117266 roubles in your account - you can do whatever you want.


Generally, when EURUSD is at current quotes and the leverage of 1:500, the loss at order closing begins to exceed the returned margin when the width of the channel exceeds 50 points. Up to this value the returned margin exceeds the fixed loss - you can safely close orders of any size.


Because of this you have suggested me to calculate an optimal channel width for each symbol: you may take 70-80 percents of this figure (50 pips) (i.e., 35-40 pips), which will be an optimal and secure channel width for EURUSD. Under most unfavourable conditions (many reversals, lack of deposit for placing an order, the price moves in an unnecessary direction), you will be able to exit the "Avalanche" saving a decent part of your deposit.


Thank you for stimulating the idea generator.

 
JonKatana писал(а) >>

There is no error in the calculations. Under your conditions, there is 70,000 roubles in the account before the release of the collaterals. Therefore, after the losses are fixed and the collateral is returned, the deposit does not change much. You set these conditions - blame yourself. Next time count it BEFORE you write anything.

Once the margin is released the Avalanche starts all over again. I hope you have noticed that ALL orders are closed - both Buy and Sell. That is why you have no orders - neither open nor pending - after the return of the margin. The chart is clear, you have 117266 roubles in your account - do what you like.

Generally, when EURUSD is at current quotes and the leverage of 1:500, the loss at order closing begins to exceed the returned margin when the width of the channel exceeds 50 points. Up to this value, the refundable deposit is higher than the fixed loss, and you can safely close orders of any volume.


Buy closed at breakeven, sat at 4 lots closed at a loss after 40p. I gave 46,672 roubles to the market. What does the deposit have to do with it, it is my money I paid it back, forget about it, it (the deposit) is not a profit and not a loss, and fixing a losing position in 4 lots through 40p. is a real reduction of my money with which I have entered the market, by 46 672 rubles. Do you understand it? How is it - "After fixing the losses and returning the collateral the deposit does not change much"? Are you crazy? Are you talking nonsense? Do you understand that with a lot of 4, each passed pips in the wrong direction, your funds are reduced by 1,166.8 rubles? Pledge is constant, this is your money, they are just "frozen", and here the funds melt with every point and when you fix the position, you will no longer see them and your deposit will be 46,672 rubles less (with the returned pledge)!

 
"He was making fun of us. Crazy - what can you take..." // V.S.Vysotsky "Letter to the Editor from Kanatchikova Dacha".
 
++++++++
Patient to orderly: Can I live a little longer?
Orderly: The doctor said morgue, so morgue!
 

Svinozavrlexandros
If you have a head like mine, programmers, traders, you have a lot of experience, tell me how much will the deposit decrease when you lock a losing position with a volume of 4 lots in 40 pips (not including the spread). How much will we lose in roubles, for example on the euro/buck?

 

JonKatana писал(а) >> Вообще, для EURUSD при текущих котировках и плече 1:500 убыток при закрытии ордера начинает превышать возвращаемый залог при ширине канала более 50 пунктов.

JonKatana, at what point did you get the brilliant idea to compare the deposit with the loss?

Up to this value, the refundable deposit is greater than the fixed loss - you can safely close orders of any volume.

That's it, you can relax: you have entered a new financial reality, in which you don't care about losses - as long as they are less than the deposit.

 
I recommend the topicstarter to open a trader's calculator on the website of any DC... And buy a 2nd grade maths book...
At lot 4 the deposit is $1079 (at 1:500 leverage). And the price of 1 pip is $40
It's simple - when we lock in a loss of -40 pips with lot 4. We get back our $1079. And we lose forever 40*40=1600 dollars. (Spread is not taken into account).

Calculations from school arithmetic...
But apparently the head starter had some other training ... This is not the first time I notice that he has serious problems with elementary arithmetic.