A Simple Mean Reversion Strategy | Page 20 | Forex Factory

Options

Go to
Permalinks
Print Thread

Related Instruments

EUR/GBP

Similar Threads

Simple Mean Reversion 621 replies

M5 Mean Reversion Strategy 65 replies

mean reversion strategy 8 replies

Forward test of my new mean reversion strategy 11 replies

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

Attachments: A Simple Mean Reversion Strategy
Exit Attachments

A Simple Mean Reversion Strategy

  • #381
  • Dec 2, 2025 3:08pm Dec 2, 2025 3:08pm
Quoting Riskcuit
They almost surely are just measuring the max range (high - low) for each period and comparing them.
You are right. I was overthinking it. I now also reproduced the same graph by just ignoring time-of-day and looking at the absolute maximum rather than percentiles.

Quoting Riskcuit
Take a second look at the SD probabilities I posted earlier. These are applying to MA candles, which have much larger kurtosis than normal OHLC candles but the idea is the same. If anything, its more important because their is less kurtosis to take advantage of in OHLC candles. But the main point is, when there is high kurtosis, more occurrences fall within 2SD than normal BUT less occurrences fall within 3SD, like a lot less.
Thanks. Trying to recalculate that myself to check if I understood correctly, and to see how to exploit that in trading.
  • #382
  • Dec 2, 2025 11:35pm Dec 2, 2025 11:35pm
Quoting Riskcuit
Today I would like to share what I consider, the most fun strategy to trade that I personally use. It’s simple, it works, and anyone can do it. Not only that, you can use it on any timeframe, depending on how much time you’d like to spend in front of the charts. Some people really enjoy spending lots of time on the charts, while others only want to check in only a few times a day to adjust their positions. Regardless of where you are on that spectrum, this strategy can work for you. Although, on the surface, the strategy is very simple and easy...

hi Riskcuit,

this is awesome, are you producing EA on this?
  • #383
  • Dec 3, 2025 7:45am Dec 3, 2025 7:45am
Quoting yoriz
{quote} You are right. I was overthinking it. I now also reproduced the same graph by just ignoring time-of-day and looking at the absolute maximum rather than percentiles. {quote} Thanks. Trying to recalculate that myself to check if I understood correctly, and to see how to exploit that in trading.
My 2 cents @yoriz @Riskcuit

I'm not expert like you guys but by experience I found that mean reversion systems soon or later lead to accounts blow.
Why?
1) Because there is no way to know in advance if the market is going to range/trend.
2) soon or later you will face adverse conditions. If you don't stop trading and cut losses for sure account blow.

The only way to survive is trading manually or semi auto because it's easier to stop trading when things go wrong or having big accounts and trade very very small.
Alternative, trade small, have decent account size and withdrawn frequently. The hope is to withdraw more frequently than blowing the account.

BTW, I do not know any professional trader that trades using the method described in this thread.
Maybe because it's too dangerous etc...
  • #384
  • Dec 3, 2025 8:26am Dec 3, 2025 8:26am
Quoting Merka
{quote} My 2 cents @yoriz @Riskcuit I'm not expert like you guys but by experience I found that mean reversion systems soon or later lead to accounts blow. Why? 1) Because there is no way to know in advance if the market is going to range/trend. 2) soon or later you will face adverse conditions. If you don't stop trading and cut losses for sure account blow. The only way to survive is trading manually or semi auto because it's easier to stop trading when things go wrong or having big accounts and trade very very small. Alternative, trade small,...

Hey Merka, thanks for sharing!

I totally agree with your points, but I don't think it means something shouldn't be traded if it fits your trading style.

You are correct that you don't know in advance if the market is going to range or trend, and this is true of all trading systems at all time and it's true that eventually bad times do come for all systems. Thankfully you don't need to know in advance, you simply need to know in the now what is happening. Consider the chart yoriz posted showing the maximum ranges and the SD probabilities I posted.

These ranges are the blow up swings/moves that will kill mean reversion. In a mean reversion strategy, all of your profits are essentially going to come from moves within 1SD. All of risk is outside 2-3SD. You don't need to know if it will be 1SD or 3SD in advance, you just need to know what's happening now and adjust your risk as you mentioned.

With the numbers I posted, we know the tail moves beyond 3SD happen almost 10x more than they should, so as you move closer to 3SD and beyond, you should never be thinking "this is going to revert super hard from here", you should be thinking, "I should flip/reduce my exposure because we are significantly more likely to stay above 3SD than normal, and our maximum risk of 3SD moves are significantly larger than normal". Most people think the first though, and take on more reversion risk when in reality, you need to be reducing risk, just from a probability stand point.

It's no different than managing delta, gamma or theta risk in an option portfolio, or interest rate risk in a bond portfolio. As conditions change, you need to change or else you blow up.
ri · skuht
  • #385
  • Dec 3, 2025 8:42am Dec 3, 2025 8:42am
Hi all,
I've coded a little indicator that show +-1SD (blue), +-2SD (orangered), +-3SD (red) related to SMA200.
For example in this scenario on GU M15 best option is to remain flat waiting for price to touch +1SD before entering a short with TP = SMA200, is that right ?
Btw this indicator seems to be somehow similar to Bollinger Bands.


Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 14 KB
  • #386
  • Dec 3, 2025 9:17am Dec 3, 2025 9:17am
Quoting tomo22
{quote} hi Riskcuit, this is awesome, are you producing EA on this?
No, and as a someone who has been a software engineer professionally for over 10+ years it may be surprising that I don't like programmatic trading. I love building models, back tests, etc and using software in that way, but in terms of actual execution in the market and how I trade, everything I do is manual and "discretionary". Its essentially probability-based/data-driven discretionary trading. I am not saying programmatic trading doesn't work or that it won't work for this system, I think others have posted that it does in fact work for them, just that it doesn't work for me and how I like to trade.

Quoting skyline
Hi all, I've coded a little indicator that show +-1SD (blue), +-2SD (orangered), +-3SD (red) related to SMA200. For example in this scenario on GU M15 best option is to remain flat waiting for price to touch +1SD before entering a short with TP = SMA200, is that right ? Btw this indicator seems to be somehow similar to Bollinger Bands. {image}
Yes Bollinger bands is essentially this indicator, the rolling nature of the indicator will be misleading with the numbers I posted though and how they are being traded/used to make decisions.

The numbers I posted are for MA candles, so the lines wouldn't be rolling but would be fixed for each candle, meaning when an ma candle starts, the levels would be drawn and would remain static till the ma candle closes, so it'd look more like pivot points than a Bollinger band. Then when the next candle is opened, new lines would be drawn and they'd remain static till that candle closed.
ri · skuht
  • #387
  • Dec 3, 2025 9:38am Dec 3, 2025 9:38am
Quoting Riskcuit
{quote} Hey Merka, thanks for sharing! I totally agree with your points, but I don't think it means something shouldn't be traded if it fits your trading style. You are correct that you don't know in advance if the market is going to range or trend, and this is true of all trading systems at all time and it's true that eventually bad times do come for all systems. Thankfully you don't need to know in advance, you simply need to know in the now what is happening. Consider the chart yoriz posted showing the maximum ranges and the SD probabilities...
Thanks @Riskcuit
I totally agree with you but let's talk with concrete trading examples as several readers here do not have university degree to understand the mathematical reasons behind your concepts.

Let's pick whatever hystorical chart you experienced +-3SD move and let's analyse the moment you flipped direction.
Will be a good visual example for everyone here.

Few noob questions for you.
1) are you actually implementing these concepts in your personal trading (I assume you are a professional full-time trader)?
2) how many years of backtesting data you require to reasonable determine the +-3SD buffer? Asking this because I had EA performing well over 3 to 5 years but over 20 years or more of backrtest failed. Why? Because the was the 1 time black swan event that crashed everything.
  • #388
  • Dec 3, 2025 11:09am Dec 3, 2025 11:09am
Quoting Riskcuit
{quote} No, and as a someone who has been a software engineer professionally for over 10+ years it may be surprising that I don't like programmatic trading. I love building models, back tests, etc and using software in that way, but in terms of actual execution in the market and how I trade, everything I do is manual and "discretionary". Its essentially probability-based/data-driven discretionary trading. I am not saying programmatic trading doesn't work or that it won't work for this system, I think others have posted that it does in fact work...
Ok basically you're saying that the only bands value to be taken into account are the previous ones which are closed and don't move up/down during candle formation.
So the indicator should be something like this , where lines are calculated only once a candle is closed and a new candle is forming :
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 14 KB
  • #389
  • Dec 3, 2025 1:10pm Dec 3, 2025 1:10pm
Quoting skyline
{quote} Ok basically you're saying that the only bands value to be taken into account are the previous ones which are closed and don't move up/down during candle formation. So the indicator should be something like this , where lines are calculated only once a candle is closed and a new candle is forming : {image}
Yes that looks more correct. So now that price is up against/near the 3SD, its very likely that will be a large swing that is not going to end or revert anytime soon like the previous one that was hanging around the 1SD line, so getting really short up here can easily trap you in a bad move.
ri · skuht
  • #390
  • Dec 4, 2025 2:50am Dec 4, 2025 2:50am
Quoting Riskcuit
{quote} Yes that looks more correct. So now that price is up against/near the 3SD, its very likely that will be a large swing that is not going to end or revert anytime soon like the previous one that was hanging around the 1SD line, so getting really short up here can easily trap you in a bad move.
Ok thx Riskcuit!

So i've just placed on EURGBP a buystop @ 0.8749 which is -1SD , waiting to hit sma 200 :
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 14 KB
  • #391
  • Dec 7, 2025 7:54am Dec 7, 2025 7:54am
Quoting skyline
{quote} Ok thx Riskcuit! So i've just placed on EURGBP a buystop @ 0.8749 which is -1SD , waiting to hit sma 200 : {image}
Did end up well this trade?
  • #392
  • Dec 15, 2025 10:21am Dec 15, 2025 10:21am
Quoting Merka
{quote} Did end up well this trade?

no
  • #393
  • Last Post: Dec 18, 2025 2:15am Dec 18, 2025 2:15am
Quoting alphaomega
I made this tool some time ago for research purposes to see how exactly the prices (high-Low) are distributed during the 24 hour cycle. You can place it in indicators folder and see for yourself. It's a bit CPU intensive because of the way I code it.....but it works and the data is accurate. You can point at any level and it shows you the distance and the exact time and date. So you can go back in history to check how this day looks and what is the shape of the curve. You have to click on the button "Calculate" to draw the data. {image} {file}
Good morning, i`d like to ask about this indicator.
What`s the logic behind it? I`ve been using it along with Goldbach Dealing Ranges and i`ve found it very profitable.

Would you share .mq4 code or at least core logic for calculation ?
I`m willing to share my thoughts

Thanks in advance, Mat
3 traders viewing now