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

Update: I realized the calculation of the blue line in the indicator above was not fair. The indicator stopped searching for "red dots" (intercepts as they occurred in the past) when it reached the end of the red/gray lines (the MA forecast). However, there can obviously be intercepts that have taken longer than that. By omitting these red dots, the average was skewed towards the intercepts that were fast.

I have updated post #236 with version 1.1 that keeps searching for the intercept of each occurance beyond the right side of the screen until it has found it. I believe the calculation is now correct, but unfortunately I noticed that the blue line is now also much closer to the current price. Regularly the blue line is even at the "wrong" side of the current price suggesting not a mean reversion but a mean "aversion". Check the updated video for an example.
Does anyone have a similar indicator that has a variable to disable the audio notification?
Attached File(s)
File Type: ex4 Pips2MA.ex4   6 KB | 170 downloads
Please excuse the bad English via Google Translate.
Quoting yoriz
Update: I realized the calculation of the blue line in the indicator above was not fair. The indicator stopped searching for "red dots" (intercepts as they occurred in the past) when it reached the end of the red/gray lines (the MA forecast). However, there can obviously be intercepts that have taken longer than that. By omitting these red dots, the average was skewed towards the intercepts that were fast. I have updated post #236 with version 1.1 that keeps searching for...
you can modify the indicator so that it displays the number of points between the green line and the blue line on the chart ?
Quoting tarzon06
you can modify the indicator so that it displays the number of points between the green line and the blue line on the chart ?
Interesting request! What information are you trying to get from that?

Practically: the number of points varies each time the indicator updates itself (~every 5 seconds), so perhaps a percentage of the total number of points is better? In my experiments I have seen anything between 17 and 3500 points. It obviously depends on the time frame and how often that specific spread occurred in the past 4 years.
Quoting Riskcuit
You can see my 3 entries on this pair as price continued to ramp up. On all of these, I was targeting the 200MA as my exit at "fair value". But price was running pretty hard so it became obvious that exiting at my current target would result in a small loss.

So in order to increase my probability of profit on this trade, I rolled my target to the 4HR chart's "fair value" (800 MA on the 1HR chart) and closed at about 50% max profit for an overall profit on the position.
How do you roll-over/manage the trade when the 800 MA is on the opposite side of your positions and keep running away?
No greed. No fear. Just maths.
Quoting PipMeUp
{quote} How do you roll over/manage the trade when the 800 MA is on the opposite side of your positions and keep running away?
If this is the case, you have been way too aggressive trading the 200MA. You should really only be rolling to the 800 if you are caught in a really large move (the kind that occurs <5% of the time for any sample), which would essentially guarantee the 800 MA won't be on the wrong side. Additionally, you need to manage your risk/size as if every move was one of the <5% moves, which means, staying small, so average down against big moves is not capital intensive, or stressful.

The trading should be very smooth and relaxing. A good rule to help this is to ensure no single entry uses more than 1:1 leverage. This is harder to do if you are trading smaller TFs but on the 1HR & up charts, this provides plenty of opportunity for profits.

For example, here is a CHF/JPY trade I entered last night that looks like it's going to go pretty well... BUT it could still turn at any moment. Currently, the 800MA is on the wrong side, but I won't be rolling to the 800MA unless the pair sinks to at least 160.00 or even lower, and only if the move down is so strong that I get over my skis on the 200MA and want to try and get out at BE.
Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot from 2023-09-01 13-12-50.png Size: 24 KB



The thing with rolling is, that there isn't a hard rule of when exactly to do it, and ideally, you very rarely ever do need to do it. It's really just a way to extend the duration with the hopes of getting out at BE or a little better for the times when you may have gotten a little aggressive a little too soon. It shouldn't be a regular occurrence and if you can get out at the 200MA with even jut a small loss, that is probably better to do than roll to the 800.
ri · skuht
Here are my results for August with this strategy, The Month started slow but ended well

Many will turn their nose up at 4% (which is a lower monthly return than usual) but 1.04^12 = 60% annualized. So I'll take 4% from stress-free, low-intensity trading any time!

Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot from 2023-09-01 13-29-34.png Size: 5 KB


Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot from 2023-09-01 13-32-33.png Size: 10 KB


Lastly, look at the breakdown of the trades.
Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot from 2023-09-01 13-33-49.png Size: 8 KB



Booked over 1000 pips with a win rate of 78% and I had an average loss of 36 pips and an average win of 47 pips. That's money dude!
ri · skuht
Quoting Riskcuit
Here are my results for August with this strategy, The Month started slow but ended well Many will turn their nose up at 4% (which is a lower monthly return than usual) but 1.04^12 = 60% annualized. So I'll take 4% from stress-free, low-intensity trading any time! {image} {image} Lastly, look at the breakdown of the trades. {image} Booked over 1000 pips with a win rate of 78% and I had an average loss of 36 pips and an average win of 47 pips. That's money dude!
Congrats!

Around August 20, you had DD of around 10-11 %, which is not bad...would you cut positions if DD was 15%, 20%?

Did you take only 36 trades or you counted each time you took profit/exited on the basket of trades as one trade?
Quoting Riskcuit
Here are my results for August with this strategy, The Month started slow but ended well Many will turn their nose up at 4% (which is a lower monthly return than usual) but 1.04^12 = 60% annualized. So I'll take 4% from stress-free, low-intensity trading any time! {image} {image} Lastly, look at the breakdown of the trades. {image} Booked over 1000 pips with a win rate of 78% and I had an average loss of 36 pips and an average win of 47 pips. That's money dude!
I have similar statistics so far trading your ideas. I trade on M15 so might become more stressful. This week was easy.
1.02^200
Quoting driven18
{quote} Around August 20, you had DD 10%, which is not bad...would you cut positions if DD was 15%, 20%? How many individual trades you have taken?
There were 36 total trades (unique orders/entries) for the month as shown in the last pic. I would definitely be defending positions a lot harder if DD was approaching 15%-20%. Something to understand is DD will always be pretty large relative to run-up. If we are looking to make 4% a month, we should expect to have draws ATLEAST the same size, if not double or triple depending on how high our rate is. Many trades will spend the majority of their time in draw because we are waiting for returns to fair value, we aren't trying to predict turns in the market.

Also at that time, I probably had positions in anywhere from 6-8 pairs, so 10% spread across that many pairs is pretty low per pair (and once you are in multiple pairs and have rolled trades, you are likely trading some currencies both ways so the drawdown between those positions will level off/hedge each other out a bit).
ri · skuht
Quoting Riskcuit
{quote} There were 36 total trades (unique orders/entries) for the month as shown in the last pic. I would definitely be defending positions a lot harder if DD was approaching 15%-20%. Something to understand is DD will always be pretty large relative to run-up. If we are looking to make 4% a month, we should expect to have draws ATLEAST the same size, if not double or triple depending on how high our rate is. Many trades will spend the majority of their time in draw because we are waiting for returns to fair value, we aren't trying to predict turns...
All below is based on your numbers for August, the month you deem to be satisfactory at 4% in profit.

Let's say that average profit is 4% a month, regardless how many pairs you trade. It is easy to get to 15% DD at which point you may have to cut the whole multiple pair position. That would take close to 4 months to get it back, considering every next month is on average is 4% Win.

If you increase position size, your DD will be even larger meaning you may have to cut it quicker. 15% DD a month with 4% in profit is a tough sell.

Those are the facts...still much work to be done.
Quoting Riskcuit
{quote} If this is the case, you have been way too aggressive trading the 200MA. You should really only be rolling to the 800 if you are caught in a really large move (the kind that occurs <5% of the time for any sample), which would essentially guarantee the 800 MA won't be on the wrong side
Not what I meant. Say you enter somewhere in the orange bubble:
Attached Image (click to enlarge)
Click to Enlarge Name: example-1.png Size: 14 KB

But theta is consummed and you add somewhere in the second orange bubble:
Attached Image (click to enlarge)
Click to Enlarge Name: example-2.png Size: 16 KB

How do you manage/mitigate this?
Attached Image (click to enlarge)
Click to Enlarge Name: example-3.png Size: 24 KB

In this example the price had a nice pullback to close on the red MA800 at -450 pips but it might not have happened. The max DD was -850 pips on this example. The winners are relatively much smaller.
Note: There are 760 H1 bar between the first entry and the MA800 crossing the price.
No greed. No fear. Just maths.
Quoting driven18
{quote} All below is based on your numbers for August, the month you deem to be satisfactory at 4% in profit. Let's say that average profit is 4% a month, regardless how many pairs you trade. It is easy to get to 15% DD at which point you may have to cut the whole multiple pair position. That would take close to 4 months to get it back, considering every next month is on average is 4% Win. If you increase position size, your DD will be even larger meaning you may have to cut it quicker. 15% DD a month with 4% in profit is a tough sell. Those are...
The risk may not be as tolerable for others but I find it to be fine/easy to deal with.

Also if you understand risk/reward and probabilities it should be fairly obvious that if your win rate is above 50%, your drawdowns will almost always be larger than your run ups for any sample period. If that isn’t intuitive for you, I’d recommend reading more on probability theory.
ri · skuht
Quoting PipMeUp
{quote} Not what I meant. Say you enter somewhere in the orange bubble: {image} But theta is consummed and you add somewhere in the second orange bubble: {image} How do you manage/mitigate this? {image} In this example the price had a nice pullback to close on the red MA800 at -450 pips but it might not have happened. The max DD was -850 pips on this example. The winners are relatively much smaller. Note: There are 760 H1 bar between the first entry and the MA800 crossing the price.
Yeah, so this is an instance where you'd almost certainly end up with a net loser across all the positions for this move, whether you rolled or not. This large of a move is a <3% (probably even a less than 1%) of the time move, and these are the ones you need the be careful about.

As I have said many times, this isn't some 100% win martingale strategy. You will have occasional losers, and they can be very large (this is why I mentioned VAR being very large, and you should assume every trade can draw down to VAR).

Below is your image marked up, The black lines show where theta needs to eat your opportunity away before you should consider re-adding. It also assumes you rolled after the 3rd entry (which it really probably would have been better to just take the net loss on the return to 200 after the 3rd entry on all 3 orders).



Attached Image (click to enlarge)
Click to Enlarge Name: Untitled design (1).png Size: 31 KB



As I mentioned in the first post about rolling, when you roll, you typically want to take risk off after doing so by reducing your positions (like I showed in GBP/NZD). So after that 3rd entry (and you roll to 800 MA), you are looking to close that 3rd entry at BE or small profit to reduce risk, because you know you have just extended your duration by rolling, and should be expecting the drawdown to continue (this is why you rolled in the first place). You had many chances to also as it was trading around that 3rd entry for at least 12 or so hours a few days after the entry (when it returned to the 200MA).


Also, once you roll, the management technique of using theta rolls to the higher MA as well, so you wouldn't re-add for a while after the 3rd entry. So in this instance, assuming you were super mechanical, you'd end up with 2-3 losers and a winner, but an overall net loss on position. It happens, this is why you need to stay small so you can weather these storms that occur 1-3% of the time.

Lastly, typically when you are in these moves, you know it, so you are more cautious to add in general as you know you are in a tail move that can run run run. In these cases, you typically aren't trying to add as soon as possible at each possible entry. You should think, "This thing is running, even though I technically could add here by XYZ rules, I probably shouldn't and should look to at a higher price, which will almost surely occur because I rolled the trade, extending my duration, therefore extending the possible levels price can reach in either direction”.

You never roll the trade because you think you are timing a turn in the market. It's the opposite, you roll the trade because you have no clue when it will turn, and you don't think it will be soon, so you are being less aggressive once you roll.

Rolling is not a get-out-of-jail-free card and if you struggle with risk management, you probably shouldn't do it as it can cause more harm than good if you use it too aggressively. You roll for a last-ditch effort to reduce the loss you will take/give you a chance to turn it into a winner or BE. That doesn't mean it works in your favor 100% of the time. If folks are looking for something like this, they will never find it in the trading business.
ri · skuht
Quoting Riskcuit
{quote} The risk may not be as tolerable for others but I find it to be fine/easy to deal with. Also if you understand risk/reward and probabilities it should be fairly obvious that if your win rate is above 50%, your drawdowns will almost always be larger than your run ups for any sample period. If that isn’t intuitive for you, I’d recommend reading more on probability theory.
Probability theory, You are wrong here and I will be gentile, it is Friday, any technique that has 4 times DD(I know you have 3 times DD but we have to account for WTF moments) of the potential monthly profit, is technique that is very close to failure. If you have $100,000 to trade and you are $16,000 negative in DD hoping to make $4,000 at the end of the month, the probability is you will not place another trade.
Quoting driven18
{quote} Probability theory, You are wrong here and I will be gentile, it is Friday, any technique that has 4 times DD(I know you have 3 times DD but we have to account for WTF moments) of the potential monthly profit, is technique that is very close to failure. If you have $100,000 to trade and you are $16,000 negative in DD hoping to make $4,000 at the end of the month, the probability is you will not place another trade.
Attached Image (click to enlarge)
Click to Enlarge Name: cover1.jpg Size: 32 KB
ri · skuht
Quoting Riskcuit
{quote} {image}
I am not getting it. Anyway, Good Luck to us All!
Thank you riskcuit for this nice and simple strategy. I tested it on a small live account. I used too much risk, I know , but I know what I do, since I am trading since over 15 years now. Here is the result from the last 3 weeks:

Attached Image (click to enlarge)
Click to Enlarge Name: IMG_1214.jpeg Size: 570 KB


i can only suggest everyone to live test this strategy. Especially I like riscuits analogy to a real market store . A whole different view to the markets. But as he said, the power of this strategy is to trade often and many markets. The losers, which will appear are than not big anymore… thanks again riscuit.
the only place success comes before work is in the dictionary
Quoting kleybenny
Thank you riskcuit for this nice and simple strategy. I tested it on a small live account. I used too much risk, I know , but I know what I do, since I am trading since over 15 years now. Here is the result from the last 3 weeks: {image} i can only suggest everyone to live test this strategy. Especially I like riscuits analogy to a real market store . A whole different view to the markets. But as he said, the power of this strategy is to trade often and many markets. The losers, which will appear are than not big anymore… thanks...
how many pairs are you trading and what timeframe?
Seek first to understand, then to be understood
Mean reversion has a higher-than-expected relative drawdown based on the posts here - something amiss.

Try VWAP or TWAP with progressive entry(ies) instead. It helps to reduce floating dd at the moment.

Additionally, mean-regression model for smaller accounts sub 250k can look at higher timeframe time-bars to follow the trend while regressing back to WAP.

Cheers,
PDX
PDX adjusts risk, not scales..
0 traders viewing now