Simple Mean Reversion | Page 4 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

A Simple Mean Reversion Strategy 392 replies

Pairs Trading: Reversion to the Mean 22 replies

Forward test of my new mean reversion strategy 11 replies

Synthetic hedges, cointegration, mean reversion and similar stuff 419 replies

Cointegration, Synthetic hedges, mean reversion in R, Tech Thread 0 replies

Attachments: Simple Mean Reversion
Exit Attachments

Simple Mean Reversion

  • #61
  • Mar 7, 2018 11:54am Mar 7, 2018 11:54am
[quote = alphaomega; 10848503] {quote} O indicador não é necessário. Eu uso isso apenas para testes de retorno visual. (Para confirmação) Para a EA, você só precisa: preço atual atual horário diário alto diário baixo diário ATR. Tudo o resto é simplesmente derivado das coisas acima. Se você codificar com mql4: double price = Bid; int time = Hour (); duplo alto = iHigh (símbolo, PERIOD_D1,0); duplo baixo = iLow (Symbol, PERIOD_D1,0); intervalo duplo = alto-baixo; duplo atr = (_ Símbolo, PERIOD_D1,20,0); Ret0 duplo = (alto + baixo) / 2; double retr25sell = high- (range * 0.25); double retr25buy = low + (range * 0.25); double addstep = atr * 0.1; Double MinRange = atr * 0.1; Isso é tudo que você precisa. E daqui você pode modificar os parâmetros e você pode adicionar outras variáveis. [/ Quote]

Tanks!
  • #62
  • Mar 7, 2018 3:45pm Mar 7, 2018 3:45pm
Read the thread only superficially!
Wonder if the black swan can be mitigated, with a hedge, locking the position in a "specific" span.
Then, wait for the next opportunity, and use both (long an short ) as part of a new campaign, as opportunity arrives, with the same system.
I use Oanda, live in Europe, you can hedge as much as you want, you can see both lots open on the same chart.
It´s a way of giving your failed system, on that day, a (hopefully) better chance tomorrow.
I understand those ocurrences are seldom.
Just my humble 2 cents.
  • #63
  • Mar 7, 2018 4:03pm Mar 7, 2018 4:03pm
Quoting cuchuflito
Wonder if the black swan can be mitigated, with a hedge, locking the position in a "specific" span. Then, wait for the next opportunity, and use both (long an short ) as part of a new campaign, as opportunity arrives, with the same system.
This sounds like a recipe for a disaster... For so many years here, no one ever described a proper hedging technique. You must double or so the proper direction lots and when the rate turns around into the direction of the black swan lots, which are the trouble ones, close the hedging lots and open them AGAIN into the new direction (the black swan lots one).

That is more easy said than done, cause one needs an exit plan and very very good signaling system for a direction change.

The other option would be to just close the losers and open net lots and harvest the trend and then the new trend... Most often I do this, but it is far less profitable than the first option.

Hedging is not complicated when you have a nice signaling/exit system, but most traders do not have such, cause thinking is hard and why make life complicated?
Maadik Hugiis. IQ 69.
  • #64
  • Mar 7, 2018 8:42pm Mar 7, 2018 8:42pm
This system indeed is a very good system if combined with technical analysis. You just need to have a mechanism to exit and technical analysis sound enough to retrace the price to 50% at least. It can be applied on weekly also.

There are systems with 98% win ( specially position building systems) but can drain your account in a single failure. This is the problem with most of the learners that they jump in and declare a system Holy-Grail once they hit a couple win trades.

Forex trading cannot be a profession until you have experience of at least 10 years with good knowledge of all technical and fundamental effects. But a very good time pass if experimenting bizarre systems.

just my experience.

Attached Image (click to enlarge)
Click to Enlarge Name: mme.png Size: 23 KB
  • #65
  • Mar 7, 2018 9:28pm Mar 7, 2018 9:28pm
Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot2.png Size: 39 KB
  • #66
  • Mar 7, 2018 9:37pm Mar 7, 2018 9:37pm
Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot1.png Size: 78 KB
  • #67
  • Mar 7, 2018 9:48pm Mar 7, 2018 9:48pm
Quoting traderefx
{image}
hi traderefx whats your win % following this rules
  • #68
  • Mar 7, 2018 9:59pm Mar 7, 2018 9:59pm
We are discussing here several ways to use this system, I hope that you even make a visual backtest and draw your own conclusions!
  • #69
  • Mar 7, 2018 10:02pm Mar 7, 2018 10:02pm
It is important to check the price inclination in this system, so I suggest, buy above 33.3% and sell below 66.6%, to ensure that we will have the established trend!
  • #70
  • Mar 7, 2018 10:08pm Mar 7, 2018 10:08pm
Quoting traderefx
We are discussing here several ways to use this system, I hope that you even make a visual backtest and draw your own conclusions!
I have visually back tested and still dd is to high thats why I ask. OK thank you
  • #71
  • Edited 3:17am Mar 8, 2018 2:03am | Edited 3:17am
Quoting IenDzi
{quote} The pair is EURUSD. The stats are correct, but you must think deeper. The SL will be the same, so yes 30p. But TP target will change, when the range changes. So TP will be from 7.5(30p range)-15p(60p range). So if range will be 59p and you get 25% retracement, your TP will be negative.
I see.
There is possible to go with grid
EU DO 1,2400, price goes up 30 pips.

1,2430 sell TP 7,5 SL 10
1,2440 sell TP 10 SL 14
1,2454 sell TP 14 SL 6
1,2460 sell TP 15 SL 2

Maximum DD is 36 pips with trading costs.
With 95% win rate is Kelly criterion still 0,5 even if your average win is only 4 pips.

Open trade is always closed, when there happens 25% retrace between 1,2430 - 1,2460.
For example, it can come down from 1,2437 to 1,24277 (25% from 37 is 9,25)
Profit will be 1,2430 - 1,24277 = 1,3 pips trading costs extracted.

Can it work?

In 95% cases trade result is starting from BE(minus trading cost) and going to 25% of full retrace TP.
Grid is not ideally composed, because it does not cover trading costs when going to BE.

With regards,
SP
  • #72
  • Mar 8, 2018 4:06am Mar 8, 2018 4:06am
Quoting SuperPip
{quote} I see. There is possible to go with grid EU DO 1,2400, price goes up 30 pips. 1,2430 sell TP 7,5 SL 10 1,2440 sell TP 10 SL 14 1,2454 sell TP 14 SL 6 1,2460 sell TP 15 SL 2 Maximum DD is 36 pips with trading costs.
Range = 100p; High = 1.21000; Low = 1.20000; Sell trades
1 Trade: Entry = 1.21000; SL at 33% of range; TP at 25% retracement; SL trades: 5180;(-33p) TP trades: 11990 (avg high before retracement =112; avg pips = 16p(112/4 -(112-100))
2 trade: Entry = 1.21330; SL at 80% of range; TP at 25% retracement; SL trades: 1503;(-47) TP trades: 3677 (avg high before retracement =151; avg pips = 21p;
3 trade: Entry = 1.21800; SL at 100% of range; TP at 25% retracement; SL trades: 990;(-20) TP trades: 513 (avg high before retracement =189; avg pips = 38p;
This stats without spread;
When I add 1p spread to all trades: pips are 3317; It's 10 years data, and trade hours = 0-24;

I looked at my code. There can be some missing trades. These missing trades are the ones, which doesn't get retrace until the end of the year. It shouldn't be more than 5% of these results. But these trades would have range extended more then 2 times, in other words SL trades.
I will do the same stats with continuous flow of data, that wouldn't be those missing trades.
  • #73
  • Mar 8, 2018 5:25am Mar 8, 2018 5:25am
Quoting IenDzi
Minimum range to open trade = 30pips. If you take trades from 0:00-11:59: *95% that the range won't be 2 times larger before 25% retracement occur.(It means if you enter trade at 30p range, the 25% retracement will be before 60p range 95% of time.
Dear IenDzi,

Could you describe please the way how do you got the result above?
In range between 30-60 pips return 25% happens with 95% percent of days.
Or did I understood something wrong?

With thanks,
SP
  • #74
  • Mar 8, 2018 5:30am Mar 8, 2018 5:30am
I coded a system on tick data.
Entry conditions:
Range >= 30pips. First entry is only from 0:00 - 11:59; Second entry is whenever the range extends.
1. Entry at new high/low; SL at 1.66 * range; TP = 25% retracement; 1p = 1usd
2. Entry at 1.33*rangeBefore1Entry; SL at 1.66 * range; TP = 25% retracement; 1p = 2usd;
1p spread; AccountBalance = 10000usd;
Date: [2004-2017]
Trades:15891;
Account balance from 10000 - 16703;

It is 14 years data stats. The money management is very simple, it's not even money management. But it's easier for me to see the stats better.
When you look at all 14 years stats, it seems that results are good.
But when you split in years, you can see that strategy works very well until 2011. The balance went from 10000 - 25100. And from 2011-2017.12 the balance went from 25100-16703. And it is more than a half of previous years profit.
It's not the first time that I see this. I backtested my created trend trading system and the results were the same. It's interesting for me, what happened at that year. What i see in the chart in 2011, it seems that ranging/change of trend starts at that time.

My previous stats were from 2004-2014 years.
So from 2014-2017 the stats should be worse.
  • #75
  • Edited 6:01am Mar 8, 2018 5:46am | Edited 6:01am
Quoting SuperPip
{quote} Dear IenDzi, Could you describe please the way how do you got the result above? In range between 30-60 pips return 25% happens with 95% percent of days. Or did I understood something wrong? With thanks, SP
First trade always be at 30p range. Then I measure the range until it gets 25% retracement. So the afterRange can get to infinity if there is no 25% retracement. So 95% of trades will get the 25% retracement before afterRange/range < 2.
When the retracement 25% occurs, the trade cycle starts over again.
Waiting for a new high/low. Now range can be 50p ,100p ... infinity. And it trades until the time limit. AlphaOmega said we should trade until midcycle (12:00).
The trades per day has no limit, it can go to infinity.

So I measure the range before trade, the range after trade and the duration to get 25% retracement hit;
The stats were 2004-2014 year, EURUSD.

We need timing, without timing there is no edge here. Or a very smart money/position manager: a hybrid which trades trend system with this mean reversion system.

I hope it is more clear now.
Best, IenDzi.
  • #76
  • Mar 8, 2018 5:54am Mar 8, 2018 5:54am
Quoting IenDzi
{quote} First trade always be at 30p range.
Ok, Thanks!
Can you check, what will be probability for the next situation:
1) We need only one 25% retrace for one day
2) Retrace can start from 30-60p range.

Rgrds,
SP
  • #77
  • Mar 8, 2018 6:18am Mar 8, 2018 6:18am
Quoting SuperPip
{quote} Ok, Thanks! Can you check, what will be probability for the next situation: 1) We need only one 25% retrace for one day 2) Retrace can start from 30-60p range. Rgrds, SP
Ok, I'll do it. But first I recode my program, that wouldn't be missing data.
  • #78
  • Edited 7:32am Mar 8, 2018 7:18am | Edited 7:32am
Quoting SuperPip
{quote} Ok, Thanks! Can you check, what will be probability for the next situation: 1) We need only one 25% retrace for one day 2) Retrace can start from 30-60p range. Rgrds, SP
One trade per day.
EURUSD.
2004.01.01-2017.12.31.
Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot2.png Size: 70 KB

Stats are the same if you take 1 trade per day or until the time limit.

And this is previous stats from 2004.01.01 - 2014.12.01. Trades from 0:00 - 23:59(All the time).
Attached Image


Stats from 2004.01.01 - 2014.12.01. Trades from 0:00 - 11:59.
Attached Image
  • #79
  • Mar 8, 2018 7:51am Mar 8, 2018 7:51am
Quoting IenDzi
{quote} One trade per day. EURUSD. 2004.01.01-2017.12.31. {image} Stats are the same if you take 1 trade per day or until the time limit. And this is previous stats from 2004.01.01 - 2014.12.01. Trades from 0:00 - 23:59(All the time). {image} Stats from 2004.01.01 - 2014.12.01. Trades from 0:00 - 11:59. {image}
Hi!

Thanks! Can you collect the same statistics for last 3 and 5 years?

The first row <1,1 includes trades only over 30p range?

With regards,
SP
  • #80
  • Mar 8, 2018 7:57am Mar 8, 2018 7:57am
Nice little forward test today on 19 pairs M1 charts default settings
Attached Image (click to enlarge)
Click to Enlarge Name: 10.JPG Size: 81 KB
1 trader viewing now