Simple Mean Reversion | Page 3 | 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

  • #41
  • Mar 6, 2018 12:00pm Mar 6, 2018 12:00pm
Quoting alphaomega
Here you can see example for correct trading using the standard rules with EURUSD price action from this morning: {image}
Sorry mate, but I didn't understand why the position when the final rally occured shouldn't be opened, following the system rules. It's a new high, I should place a sell order. Am I missing something? Thanks again
  • #42
  • Mar 6, 2018 12:19pm Mar 6, 2018 12:19pm
Quoting GavynSykes
{quote} Sorry mate, but I didn't understand why the position when the final rally occured shouldn't be opened, following the system rules. It's a new high, I should place a sell order. Am I missing something? Thanks again
You can take that trade if you want, but according to system rules, it is more risky since it would be taken after the mid cycle (12:00)...therefofe it is not advised.

Btw, great thread Alphaomega.
The biggest risk is not taking a risk.
  • #43
  • Mar 6, 2018 12:26pm Mar 6, 2018 12:26pm
Quoting leppozdrav
{quote} You can take that trade if you want, but acording to system rules, it is more risky since it would be taken after the mid cycle (12:00)...therefofe it is not advised. Btw, great thread Alphaomega.
Ok thanks. This evening I will perform a complete backtest with historical tick data and I will later on share the results to have an idea...
  • #44
  • Mar 6, 2018 4:41pm Mar 6, 2018 4:41pm
@alphaomega

Why close at the 50% line?

Looking back to Jan 8 of this current year on the EUR/USD M15, the price has gone back to the open line (the point where it was at exactly 00:00 on GMT+2) on a quick count about 39 times out of 41. Presenting an opportunity to make more profit most of the time.

Thank you for sharing your system. I too am following with interest.
Don't look down when learning to fly.
  • #45
  • Mar 6, 2018 8:56pm Mar 6, 2018 8:56pm
Quoting SuperPip
Hi! In threads "Quantum London Trading" by mer071898 https://www.forexfactory.com/showthread.php?t=551382 and "Q A M Trading Strategy" by Masterrmind https://www.forexfactory.com/showthread.php?t=580721 guys did a lot research regarding to such trading style. Masterrmind says in his first post: Initially only 2 currency pairs EURCHF and EURGBP will be traded due to their common 'turtle' like behavior. EURCHF WILL BE THE ONLY PAIR traded at the start. EURGBP will be traded at a much later stage. TRADING WINDOW Asian Session...
Yes there has been plenty of research in the mean reversion space applying Quantum trading concepts and building up cycles of trades either in a constant lot or progressive lot sequence however the DD has always been a key issue.

Mean reversion tendencies on turtle pairs among others are common in the Asian trading session where price will move away from the daily open price in a certain direction over a number of pips only to later return to the daily open price.

The EURCHF pair is a great example of this type of PA.

Having a good idea of the expected trading ranges within a given trading session or preferred part of a trading session assists in knowing when its time to think about exiting a position since extended trends in a given direction accelerate DD and lead to huge losses based on the trade setup margin.

One idea is to apply a hard stop just outside the typical average trading session ranges.

Waiting for price to revert to the mean or to a preferred price level is a dangerous practise after average ranges have been exceeded based on experience and historical testing of plenty of data.

Huge account balance with small total position size can weather many storms in the market in a mean reversion strategy but there will always come a time when price simply continues to trend heavily against your position(s) and puts a huge dint in your account balance.

An effective exit strategy is the key to success but it is not easily found.

I am keen to see where AlphaOmega takes this thread since the concept interests me.

Masterrmind.........
Master your Mind then Master your Trades
  • #46
  • Mar 7, 2018 3:07am Mar 7, 2018 3:07am
I did some stats on this. There is no timing analysis in this. 10 years stats.
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.)
*the trade gets 25% retracement in 2 hours 94% of time.
*Problem: max range before 25% retracement is 8 times larger.(It means if you enter trade at 30p range, the 25% retracement will be after 240p range.) Probably it was some news.

If you take trades from 12:00-23:59:
*93% that the range won't be 2 times larger before 25% retracement occur.
*the trade gets 25% retracement in 2 hours 76% of time.
*Problem: max range before 25% retracement is 8.5 times larger.

So you can see here, everything is pretty much the same, but if you take trades in the first half of cycle, you get your TP quicker.
  • #47
  • Mar 7, 2018 4:28am Mar 7, 2018 4:28am
Thanks for this trading strategy.
  • #48
  • Mar 7, 2018 4:31am Mar 7, 2018 4:31am
Initial testing from Jan 1, 2018, standard rules, without optimization:

Attached Image (click to enlarge)
Click to Enlarge Name: test 1.png Size: 51 KB


Attached File(s)
File Type: ex4 Simple Mean Reversion (Test 1).ex4   12 KB | 1,894 downloads
  • #49
  • Mar 7, 2018 5:11am Mar 7, 2018 5:11am
Quoting IenDzi
I did some stats on this. There is no timing analysis in this. 10 years stats. 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.)
Hi!

Thanks for the stats!
Is it correct? If pair is moved away 30pips from Daily Open,
then you open the reversal trade SL30 and TP7,5 pips and winrate is 95%?
Kelly criterion will be here 0,75 ( Holy Grail for sure).
May I ask what pairs you tested?

SP
  • #50
  • Mar 7, 2018 6:13am Mar 7, 2018 6:13am
Quoting SuperPip
{quote} If pair is moved away 30pips from Daily Open, then you open the reversal trade SL30 and TP7,5 pips and winrate is 95%?
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.
  • #51
  • Mar 7, 2018 6:24am Mar 7, 2018 6:24am
thank you for the file. I did some tests as well.

Below is Eurusd test from January 2015 until now on a 10.000$ account. 0,01 lots were used.
It looks promising, but like it was already mentioned, diversification amongst more pairs and bigger account is a must, or sooner or later it will go downhill fast.
DD for this test was 57%, which is way too much. So maybe limiting trade units for opening new trades could help...or even adjust mid cycle to the left. (10:00) for example.

On the other hand, I did the same test for Eurusd, but this time with 1000$ account....and it failed.

I am looking forward of the second part of the strategy and square root logic, seeing the whole picture and possible adjustments.

Keep up the good work.
Attached Image(s) (click to enlarge)
Click to Enlarge Name: TesterGraph.gif Size: 25 KB
Click to Enlarge Name: TesterGraph2.gif Size: 22 KB
The biggest risk is not taking a risk.
  • #52
  • Mar 7, 2018 7:07am Mar 7, 2018 7:07am
Quoting leppozdrav
thank you for the file. I did some tests as well. Below is Eurusd test from January 2015 until now on a 10.000$ account. 0,01 lots were used. It looks promising, but like it was already mentioned, diversification amongst more pairs and bigger account is a must, or sooner or later it will go downhill fast. DD for this test was 57%, which is way too much. So maybe limiting trade units for opening new trades could help...or even adjust mid cycle to the left. (10:00) for example. On the other hand, I did the same test for Eurusd, but this time with...
Yes your points are correct.
What is interesting is that if you analyze the results in details, you will find that the system fails only a few days each year. And almost all of the other days are profitable.
For example on EURUSD from Jan 1, 2017 - up to now there are only 4 big losing days. The last big losing day is Nov 14, 2017.
Without these black swans we have smooth sailing.
  • #53
  • Mar 7, 2018 7:30am Mar 7, 2018 7:30am
Excellent Indicator and Strategy! I'm using and getting a good hit rate! Could you send the indicator buffers scheme? I would like to develop an EA with this indicator. Thank you!
  • #54
  • Mar 7, 2018 7:31am Mar 7, 2018 7:31am
So as you say - if you cast a wide net the baskets should take care of those few bad apples which are not likely to happen across the board. Sweet!
  • #55
  • Mar 7, 2018 8:09am Mar 7, 2018 8:09am
Just pay an attention the max DD almost equals the annual return. It is slightly over, actually. This is caused by the equilibrium, I suppose.

To lower the DD, you can start with 100k account, instead of 10k, trading the same values (0.01 lot), but then the annual return of ~4.5% will be big enough, if trading is just for a hobby?

I think CP should have said a word or two about the Hs/Ls in general, but surprisingly or not surprisigly maybe, CP mentioned nothing?!? Well, in this situation, I can advise you, driven by the collective brain power to start digging into that direction?

Quoting leppozdrav
Eurusd test from January 2015 until now on a 10.000$ account. 0,01 lots were used. DD for this test was 57%, which is way too much.
Maadik Hugiis. IQ 69.
  • #56
  • Edited Mar 8, 2018 7:31am Mar 7, 2018 8:11am | Edited Mar 8, 2018 7:31am
Quoting traderefx
Excellent Indicator and Strategy! I'm using and getting a good hit rate! Could you send the indicator buffers scheme? I would like to develop an EA with this indicator. Thank you!
The indicator is not necessary. I use it only for visual back testing. (For confirmation)

For the EA you only need:

current price
current time
daily high
daily low
daily ATR.

Everything else is simply derivative from the things above.

If you code with mql4:

Inserted Code
double price=Bid;
int time=Hour();
double high=iHigh(Symbol,PERIOD_D1,0);
double low=iLow(Symbol,PERIOD_D1,0);
double range=high-low;
double atr=iATR(_Symbol,PERIOD_D1,20,0);
double retr50=(high+low)/2;
double retr25sell=high-(range*0.25);
double retr25buy=low+(range*0.25);
double addstep=atr*0.1;
double minRange=atr*0.1;

That's all you need. And from here you can modify the parameters and you can add other variables.
  • #57
  • Mar 7, 2018 8:41am Mar 7, 2018 8:41am
Quoting MoneyZilla
Just pay an attention the max DD almost equals the annual return. It is slightly over, actually. This is caused by the equilibrium, I suppose. To lower the DD, you can start with 100k account, instead of 10k, trading the same values (0.01 lot), but then the annual return of ~4.5% will be big enough, if trading is just for a hobby? I think CP should have said a word or two about the Hs/Ls in general, but surprisingly or not surprisigly maybe, CP mentioned nothing?!? Well, in this situation, I can advise you, driven by the collective brain power to...

You are right. DD almost equals the final return. Increasing the account size is ok and would lower the DD, but it's just a bandage for the strategy or "escaping the problem" in this case. I would focus more on filtering those bad spikes and close them sooner - limit the opening of new orders.

So, Alphaomega suggested to play out this strategy on uncorrelated Fx pairs. I don't say, we should just put it on a net of baskets and let it run.
Choosing the right Fx pairs is one of the keys here for this to go better. So I did that.

Below is a quick test of three, uncorrelated Fx pairs; Eurusd, Usdchf and Usdjpy.
The same period was selected: 2015 - Present
Account size: 10.000k
Lot size: 0.01

As you can see, we corrected the DD. (Relative DD, for all three pairs is now 27.47%)....so, all I am saying is playing this out on uncorrelated fx pairs makes more sense. But for further optimisation, we would need to hear further logic.
Attached Image (click to enlarge)
Click to Enlarge Name: multi currency test.png Size: 72 KB
The biggest risk is not taking a risk.
  • #58
  • Mar 7, 2018 8:57am Mar 7, 2018 8:57am
Thank you so much, This is what I am looking for.
Answering question investment and where to take profit of smart money.
Smile, Breath, and Go slowly
  • #59
  • Mar 7, 2018 9:27am Mar 7, 2018 9:27am
Quoting leppozdrav
{quote} Below is a quick test of three, uncorrelated Fx pairs; Eurusd, Usdchf and Usdjpy.
u are worng, this three are in fact correlated

even eur/usd can be correlated with gbp/jpy (at macroeconomic level)
plenty of smart people are doing this mistakes, just make sure that u bett small.

and another thing that everybody should be aware: using mt4 for backtesting systems its a joke.
Price itself it's an indicator
  • #60
  • Mar 7, 2018 10:33am Mar 7, 2018 10:33am
Quoting m3rlin
{quote} u are worng, this three are in fact correlated even eur/usd can be correlated with gbp/jpy (at macroeconomic level) plenty of smart people are doing this mistakes, just make sure that u bett small. and another thing that everybody should be aware: using mt4 for backtesting systems its a joke.

I understand what you are trying to say. They are, but they aren't as well. If you take for example Eurusd and Usdchf, the correlation between those two might even be positive if you look at short enough time scale. But further out we go, more stable are the correlation numbers. For example weekly correlation numbers between those 2 pairs might be around -80%. More you will zoom in, all the way down to the tick charts...more chaotic will those numbers look like.

Regarding the strategy, I don't plan on trading it. As it is now, it is far too dangerous for my taste. And I am sure, the strategy wasn't even fully disclosed.

regarding mt4 backtest; I know, I know they are a joke.....like I said those were just some quick tests, of testing multi currency idea and not much more.

But isn't it irrelevant?
You can do all those backtests on a super quantum computer if you want....they will still be, just some backtests.
The biggest risk is not taking a risk.
0 traders viewing now