Menu
Sister Sites

Options

Bookmark Thread

First Page First Unread Last Page Last Post

Print Thread

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

Quoting zkchyo
Because higher volatility leads to fatter tail. The hit rate of higher volatility entry levels increases and brings more profits when the market reverts to its fair value.
Essentially yes. Larger volatility means larger expectancy in terms of pips, not necessarily in terms of % of risk.

Something I want to make sure is clear though, volatility itself does not lead to fatter tails. You can have low and high vol in a normal distribution. It essentially just describes the size of the distribution's standard deviation.

Kurtosis is what drives tails to be fat or thin.
ri · skuht
Quoting Riskcuit
To help, here is a visual aid of how this kind of kurtosis changes the distribution relative to a normal one (red line):
Another way to look at it, is to consider each M1 bar in the last years and measure how far price has drifted away after 1 hour, 2 hours, etc. Here are the stats for AUDUSD over the last 4 years showing the 1 sigma range (p16-p84) and the 2 sigma range (p5-p95):

Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 22 KB


So in 68% of the cases (=1 sigma), the price has not drifted further than +/- 40 pips away from the current price after 24 hours! In almost all case (95% of the cases) price has not drifted more than +/- 75 pips.

The MT5 script below measures exactly this. Store the .ex5 file in the Scripts/ folder and just drop it on a M1 chart and grab a cup of coffee (might take a while...) while it calculates the statistics. It writes a .csv file in the data folder that you can open in Excel to make a plot.
Attached File(s)
File Type: ex5 yzPriceDistribution.ex5   22 KB | 155 downloads
Quoting yoriz
{quote} Another way to look at it, is to consider each M1 bar in the last years and measure how far price has drifted away after 1 hour, 2 hours, etc. Here are the stats for AUDUSD over the last 4 years showing the 1 sigma range (p16-p84) and the 2 sigma range (p5-p95): {image} So in 68% of the cases (=1 sigma), the price has not drifted further than +/- 40 pips away from the current price after 24 hours! In almost all case (95% of the cases) price has not drifted more than +/- 75 pips. The MT5 script below measures exactly this. Store the .ex5...
So when price AUDUSD drifts at +/-75 pps, we can almost mortgage the house?
Quoting Zookeeper85
{quote} I see your point! Even though the thread isn't directly about correlation, I thought it would be prudent to mention it. Buying the EURUSD and selling the GBPUSD is essentially equivalent to buying the EURGBP vice versa. For this reason, it's essentially a matter of measuring the distance from the EURGBP to its ma(x) value. Here you can see the aforementioned indicator alongside an indicator that measures the distance to the moving average for current symbol (in this case EURGBP). They are pretty much identical. {image}
hi, would you be happy to share the "distance_to_ma" indicator at the bottom of your screen? How do you find this compares to using bollinger bands etc?
ZTIRHNV All Time Return: -29.7%
Quoting Sc11
{quote} hi, would you be happy to share the "distance_to_ma" indicator at the bottom of your screen? How do you find this compares to using bollinger bands etc?
hi test it out
Attached File(s)
File Type: ex4 distance_to_MA.ex4   7 KB | 195 downloads
Quoting Zookeeper85
{quote} hi test it out {file}
ok thanks will do
ZTIRHNV All Time Return: -29.7%
Quoting driven18
{quote} So when price AUDUSD drifts at +/-75 pps, we can almost mortgage the house?
It's a piece of info for risk management.
Quantitative Trader
Quoting yoriz
{quote} Another way to look at it, is to consider each M1 bar in the last years and measure how far price has drifted away after 1 hour, 2 hours, etc. Here are the stats for AUDUSD over the last 4 years showing the 1 sigma range (p16-p84) and the 2 sigma range (p5-p95): {image} So in 68% of the cases (=1 sigma), the price has not drifted further than +/- 40 pips away from the current price after 24 hours! In almost all case (95% of the cases) price has not drifted more than +/- 75 pips. The MT5 script below measures exactly this. Store the .ex5...
Hi Yoriz,
This is a very nice work, congrats.....
However personally I would never do calculations based on pips, only percentages. For example 50 pips on EURUSD when the rate is 1.5000 is totally different than when the rate is 1.0800...
I often see that everybody somehow stuck on pips. If one is doing a calculation based on a day or a week than pips are fine. But when you are doing the calculations based on multiple years, pips are futile...
My 2 cents...Good luck to all....
Quoting GoldTheHun
I would never do calculations based on pips, only percentages. [...] when you are doing the calculations based on multiple years, pips are futile
Excellent point! Indeed, I generated the stats based on the last 4 years which is a long time.

Below is the corrected version using percentage return starting at each M1 bar of the last 4 years going forward up to 24 hours. Note that I added the 3 sigma lines also to add to Riskcuit's point about Kurtosis and fat tails.
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 22 KB


Data was generated by the MT5 script below:
Attached File(s)
File Type: ex5 yzPriceDistribution.ex5   22 KB | 143 downloads
Quoting yoriz
{quote} Excellent point! Indeed, I generated the stats based on the last 4 years which is a long time. Below is the corrected version using percentage return starting at each M1 bar of the last 4 years going forward up to 24 hours. Note that I added the 3 sigma lines also to add to Riskcuit's point about Kurtosis and fat tails. {image} Data was generated by the MT5 script below: {file}
Wow Yoriz, excellent work
My hats off to you.....
This looks to me like a Random Walk....Bravo...
Incredible symmetry...
Quoting simon.says
{quote} I would like to. If you're not selling something, having such an extensive Terms of Service on your website wouldn't make any sense. It already covers Subscription, Renewal, and of course Money Back Guarantee. Here it is for everybody else to read: https://riskcuit.com/terms-of-service -------------------------------------------------------------------- Terms of Service Published on 10-10-22 This Agreement is made between Riskcuit and you, the user and/or member of the Services (“you”). Riskcuit (“we,” “us” or “Riskcuit”) provides...
Glad you find it interesting!

Just as it has been this far, nothing in this thread or system is secret, propertiary or sold by me. There is nothing to buy.

Everything is and will always be clearly explained and given for free in the thread. And any questions will be answered!
ri · skuht
Quoting yoriz
{quote} Excellent point! Indeed, I generated the stats based on the last 4 years which is a long time. Below is the corrected version using percentage return starting at each M1 bar of the last 4 years going forward up to 24 hours. Note that I added the 3 sigma lines also to add to Riskcuit's point about Kurtosis and fat tails. {image} Data was generated by the MT5 script below: {file}
Here is my version of the same data but going to 4th sigma (red). Solid lines are reality and dotted lines assume a normal distribution. I think fat tail is clearly visible ;-)
Also visible is the covid-19 effect on the negative 4th sigma
Attached Image (click to enlarge)
Click to Enlarge Name: distribution.png Size: 35 KB


Note: I made the data using pips not %-age because it was way easier for the code.
No greed. No fear. Just maths.
Quoting PipMeUp
{quote} Here is my version of the same data but going to 4th sigma (red). Solid lines are reality and dotted lines assume a normal distribution. I think fat tail is clearly visible ;-) Also visible is the covid-19 effect on the negative 4th sigma {image} Note: I made the data using pips not %-age because it was way easier for the code.
Hi Pip, incredible work as usual
Same stats but for 2014-2018 to not include the covid period. Now the negative 4-th sigma is due to the downtrend of A/U over this period.

Attached Image (click to enlarge)
Click to Enlarge Name: distribution-2014-2018.png Size: 36 KB
No greed. No fear. Just maths.
Quoting PipMeUp
Solid lines are reality and dotted lines assume a normal distribution. I think fat tail is clearly visible ;-)
Great that you added the normal distribution for comparison! It is clear to see that price action is not a normal distribution.

But now the main question: how can we benefit from this knowledge?
Quoting PipMeUp
{quote} Here is my version of the same data but going to 4th sigma (red). Solid lines are reality and dotted lines assume a normal distribution. I think fat tail is clearly visible ;-) Also visible is the covid-19 effect on the negative 4th sigma {image} Note: I made the data using pips not %-age because it was way easier for the code.
beautiful chart PipMeUp!

The tails are very easy to see here!

The higher the timeframe the “smoother” the distribution will be. The lower the timeframe, the larger the tails will be.

As kurtosis grows, most of the distribution will cluster more around 0 with larger tails further away from zero, The closer it is to 3 the more the distribution will snap into the normal bell curve.

Kurtosis helps trading deviations from “fair value” because price is more likely to get much further away from fair value faster than it is in a normal distribution. And because there is no skew, it is also more likely to return faster than it normally would (chop up, and chop down) from these far off places.

If you try to take advantage of the difference in the distribution from what is normal, one move at a time, it is very hard, it’s more about what effect these changes have at a cumulative level when trading. This is why you need to stay small and trade often to let CLT take effect.
ri · skuht
To continue the look at price as folks like PipMeUp and others have been doing, I want to share the concept of "fair value" candles.

I have written about it in the past in my other thread regarding indicators and how when you use them, you are essentially just re-slicing the data to form a new type of candle, that is hopefully easier to trade than traditional candlestick.

Let's take the opportunities in this system as an example. What we are doing by trading around the MA is we are essentially removing time from the equation, redrawing the candles using only the price itself and a theoretical "fair value".

Even though time is removed, the candles are still similar to traditional ones in that we know the candle will always start and end. It starts when it moves away from fair value, and ends when it returns. What happens in between is anybody's guess...

Below are some good visuals to show what I mean.

First, we have the normal candle stick chart for AUD/USD 60min timeframe, looking at the last 15,000 candles.
Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot 2023-08-25 at 8.59.09 AM.png Size: 106 KB

Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot 2023-08-25 at 8.59.26 AM.png Size: 68 KB



Next, is the exact same price charted with "fair value" candles".


Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot 2023-08-25 at 8.52.20 AM.png Size: 73 KB

Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot 2023-08-25 at 8.52.35 AM.png Size: 50 KB


Notice how different the candles are. Especially the wicks...

Which candle do you think is easier to trade?
ri · skuht
Quoting Riskcuit
Next, is the exact same price charted with "fair value" candles". Notice how different the candles are. Especially the wicks... Which candle do you think is easier to trade?
"Fair Value Candles"! Truly inspiring and a novel way of plotting charts! Interesting to play with to see how to trade them.
I am going to create an indicator for this...
Quoting yoriz
{quote} "Fair Value Candles"! Truly inspiring and a novel way of plotting charts! Interesting to play with to see how to trade them. I am going to create an indicator for this...
what is the fair value candle ??
0 traders viewing now