PTS Divergence Finder Sell
The PTS Divergence Finder Sell Indicator by Roger Medcalf of Precision Trading Systems.
Bearish - Sell Indications only.
First and foremost, I have been asked many times why I did not provide a sell divergence indicator while happily providing a buy signal divergence finder for many years.
I gave the answer that sell divergences are less reliable than the buy divergences, which is still true.
Some solutions to change this were found, not by giving in to peer pressure or by modification of this indicator which I made more than fifteen years ago, but by changing the default settings to be more strict.
I also noticed that Trading View calculates indicators very fast.
The nature of fear and greed are entirely different as fear is fast and instinct driven at market tops as the opposite emotions of fear and euphoria can instantly lead a human brain into the survival mode of fight or flight.
In the bottoming or oversold conditions in markets greed probogates slowly in buyers as they consider picking up value purchases at market lows with a mindset of having a low expectation of success.
This is what causes the asymmetry in market tops versus bottoms. Therefore the asymmetric settings of the buy and sell versions of this product are now explained for clarity.
I have decided to release the sell divergence indicator with "stricter" default settings.
The Demand Index length used is 55 and the difference it needs to trigger a signal is 12. These of course are user adjustable. The strictness means there are less bad signals.
The results are many tops and intermediate high points defined with pinpoint accuracy. As expected there are some disastrous signals in the midst of violent up trends which a trader can lose on if not using risk management and stops. Likewise it frequently finds the exact top.
How does the PTS Divergence Finder Sell Indicator work?
The PTS Divergence Finder Sell Indicator accurately measures the number of divergences which have occurred in Demand Index, which is a volume based indicator.
This is a histogram style indicator for subgraph two, which plots spikes which appear like stalagmites coming up from the base.
The indicator examines multiple lookback periods of the volume based Demand Index Indicator for the length that you specify. It finds high points in prices where the DI is not making a new "local" high and missing it by the "difference" setting you input.
The Demand Index is called via the library file and the intelligent code sets volume to 1 if no volume is found.
For this reason it will “work” (producing a meaningful plot) on Forex or indices without volume but performance is going to be slightly less than optimal as the valuable dimension of volume is missing.
It is therefore best to focus on instruments like stocks and futures and cryptos that have large volumes and lots of participants.
Liquid markets where many people are “voting” on the market direction give the best results.
A total of twenty look back periods are scanned on every bar and these are hard coded and non adjustable. The length of Demand Index is user adjustable but it is suggested not to wander too far below the default setting of length 55.
The second user adjustable field is “difference” and this represents the difference between Demand Index now and Demand Index “N” bars ago. (N being 20 different look back periods of various periods)
You will understand that a length 18 Demand Index produces a much more volatile plot than a 80 period plot.
For this reason you can find short lengths of “DI” and small “difference” values will produce many more signals of divergences as there is higher volatility in the underlying indicator.
You will observe this when you use it. You can set it to give hundreds of insignificant values but it is best used so you just see the significant ones by following the guidelines below.
Consider this like using a 20 period MA on a 30 second chart compared to a 20 period MA on an hourly chart.
Clearly the hourly MA change of direction is much more meaningful and important.
Suggested settings for various lengths:
Demand Index lengths less than 12 are not generally recommended for finding "good" sell divergences
DI Length 20 = difference of 20 – 35
DI Length 30 = difference of 15 – 33
DI Length 40 = difference of 13 – 31
DI Length 50 = difference of 10 – 19
DI Length 60 = difference of 9 – 15
DI Length 70 = difference of 8 – 14
DI Length 80 = difference of 8 – 13
DI Length 90 = difference of 7 – 12
DI Length 100 = difference of 7 – 11
DI Length 110 to DI Length 200 = difference of 6 – 10
DI Length > 200 = difference of less than 6
You can use decimal places of 0.24 or 0.65 if using lengths > 500
This indicator goes up to length 1000.
As I designed this myself and have been using it for more than fifteen years you can trust me when I suggest to stay reasonably close to the default settings.
Output relevance.
The minimum value is zero which means there are no divergences found, you can then find values from 1 to 20 which is a count of the number of instances found.
Paradoxically it is not so significance if the number is very high or very low as a major top occurring on a multi month high may just show a reading of 1, but some minor mid up trend rally might show a reading of 9.
Be suspicious if you see too many large readings of 12 to 18 reoccurring as it is likely that the indicator is plotted on a market in a very long term and rapid up trend which is dangerous to be shorting.
Execution of trades.
Exercise caution with this product.
Risk control is essential and risking more than 1% to 1.5% of your capital from entry price to stop would not be advised.
As with hunting, firing out lots of small trades in a shot gun approach will lead to better results than gambling all on the first signal you see.
There is much more chance of hitting a bird with a shot gun than a canon and the ammunition is much cheaper.
Always always use a stop loss. Something like 3 to 7 times a fifty period average true range for example.
Whilst it is often possible that a spike appears exactly at the precise high of the week or year and could be the only one you see all year it is risky just to short it or sell it instantly as some markets produce several failed signals which continue to rally higher.
The safest and least risky method is to wait for the trend to begin falling after you see a divergence. This is subjective to your own definition of how to measure the trend as “falling” but I would suggest waiting for a 8-20 period Exponential average to turn down before selling.
Once the trade is entered you can implement a trailing stop to allow maximum potential gains and if your style is one of wanting to take quick profits then it is wise to take only some partial profits and give the sell off a chance to go lower and exit the remainder when the trend changes. If the move was picked up near the absolute top it could be a very large collapse in the downtrend.
Sometimes you might wait up to twenty five bars after the divergence is seen before the trend begins falling. Much longer than this an it gradually negates the signal as it shows the buyers have become stronger and the safest decision is to stay out of the market.
It is not unusual for the divergences to mark the exact high of a market and this high can lead to a large move down.
There are however frequent “failed divergences” and these can be treated in the same technical analysis manner as a failed head and shoulders or failed double top where the failure to fall indicates a likelihood of a continuation higher, meaning it is time to cut a loss.
This indicator only gives sell signals. Every single signal will be given in some degree or another in an up trend at the highest high price.
Market selection is important.
Avoid markets in an endless up trend. Look for ebbs and flows in a major down trend.
Best results are on liquid markets in a good long term down trend that has frequent rallies, you can observe the past signals and often history repeats with the good previous signals tending to indicate that future signals may also be good. (This is not certain of course)
This is also true of a market showing several historically bad divergence signals leading to more bad signals.
If the past performance of this indicator is poor on the market you are viewing, then move to another market until one is found where the readings show good price dips after the signals in historical data.
Time frames.
This product can be applied to any time frame of market but be aware as is stated above, the slower time frames yield more valid signals and shorter time frames lead to more randomness and noise ridden plots of lower significance.
That said, it provides a valid reason to enter a trade and can give good results providing good stops and risk control are used. I have seen plenty of valid signals on 30 second charts right up to weekly charts.
Idiosyncrasies.
It can often be seen that multiple divergences occur over a range of ten to thirty or so bars during a very gentle spiky kind of rally.
This can be treated in the same way as above - waiting for the trend to fall after the last divergence occurs is the way to play it.
Groups of divergences can indicate some patient insider selling patterns in anticipation of some bad news they might know.
Thanks for reading this and please read it a few more times to fully understand the points mentioned.
After that please spend some time changing settings and markets to fully appreciate how it operates.
Roger Medcalf - Precision Trading Systems
Bearish - Sell Indications only.
First and foremost, I have been asked many times why I did not provide a sell divergence indicator while happily providing a buy signal divergence finder for many years.
I gave the answer that sell divergences are less reliable than the buy divergences, which is still true.
Some solutions to change this were found, not by giving in to peer pressure or by modification of this indicator which I made more than fifteen years ago, but by changing the default settings to be more strict.
I also noticed that Trading View calculates indicators very fast.
The nature of fear and greed are entirely different as fear is fast and instinct driven at market tops as the opposite emotions of fear and euphoria can instantly lead a human brain into the survival mode of fight or flight.
In the bottoming or oversold conditions in markets greed probogates slowly in buyers as they consider picking up value purchases at market lows with a mindset of having a low expectation of success.
This is what causes the asymmetry in market tops versus bottoms. Therefore the asymmetric settings of the buy and sell versions of this product are now explained for clarity.
I have decided to release the sell divergence indicator with "stricter" default settings.
The Demand Index length used is 55 and the difference it needs to trigger a signal is 12. These of course are user adjustable. The strictness means there are less bad signals.
The results are many tops and intermediate high points defined with pinpoint accuracy. As expected there are some disastrous signals in the midst of violent up trends which a trader can lose on if not using risk management and stops. Likewise it frequently finds the exact top.
How does the PTS Divergence Finder Sell Indicator work?
The PTS Divergence Finder Sell Indicator accurately measures the number of divergences which have occurred in Demand Index, which is a volume based indicator.
This is a histogram style indicator for subgraph two, which plots spikes which appear like stalagmites coming up from the base.
The indicator examines multiple lookback periods of the volume based Demand Index Indicator for the length that you specify. It finds high points in prices where the DI is not making a new "local" high and missing it by the "difference" setting you input.
The Demand Index is called via the library file and the intelligent code sets volume to 1 if no volume is found.
For this reason it will “work” (producing a meaningful plot) on Forex or indices without volume but performance is going to be slightly less than optimal as the valuable dimension of volume is missing.
It is therefore best to focus on instruments like stocks and futures and cryptos that have large volumes and lots of participants.
Liquid markets where many people are “voting” on the market direction give the best results.
A total of twenty look back periods are scanned on every bar and these are hard coded and non adjustable. The length of Demand Index is user adjustable but it is suggested not to wander too far below the default setting of length 55.
The second user adjustable field is “difference” and this represents the difference between Demand Index now and Demand Index “N” bars ago. (N being 20 different look back periods of various periods)
You will understand that a length 18 Demand Index produces a much more volatile plot than a 80 period plot.
For this reason you can find short lengths of “DI” and small “difference” values will produce many more signals of divergences as there is higher volatility in the underlying indicator.
You will observe this when you use it. You can set it to give hundreds of insignificant values but it is best used so you just see the significant ones by following the guidelines below.
Consider this like using a 20 period MA on a 30 second chart compared to a 20 period MA on an hourly chart.
Clearly the hourly MA change of direction is much more meaningful and important.
Suggested settings for various lengths:
Demand Index lengths less than 12 are not generally recommended for finding "good" sell divergences
DI Length 20 = difference of 20 – 35
DI Length 30 = difference of 15 – 33
DI Length 40 = difference of 13 – 31
DI Length 50 = difference of 10 – 19
DI Length 60 = difference of 9 – 15
DI Length 70 = difference of 8 – 14
DI Length 80 = difference of 8 – 13
DI Length 90 = difference of 7 – 12
DI Length 100 = difference of 7 – 11
DI Length 110 to DI Length 200 = difference of 6 – 10
DI Length > 200 = difference of less than 6
You can use decimal places of 0.24 or 0.65 if using lengths > 500
This indicator goes up to length 1000.
As I designed this myself and have been using it for more than fifteen years you can trust me when I suggest to stay reasonably close to the default settings.
Output relevance.
The minimum value is zero which means there are no divergences found, you can then find values from 1 to 20 which is a count of the number of instances found.
Paradoxically it is not so significance if the number is very high or very low as a major top occurring on a multi month high may just show a reading of 1, but some minor mid up trend rally might show a reading of 9.
Be suspicious if you see too many large readings of 12 to 18 reoccurring as it is likely that the indicator is plotted on a market in a very long term and rapid up trend which is dangerous to be shorting.
Execution of trades.
Exercise caution with this product.
Risk control is essential and risking more than 1% to 1.5% of your capital from entry price to stop would not be advised.
As with hunting, firing out lots of small trades in a shot gun approach will lead to better results than gambling all on the first signal you see.
There is much more chance of hitting a bird with a shot gun than a canon and the ammunition is much cheaper.
Always always use a stop loss. Something like 3 to 7 times a fifty period average true range for example.
Whilst it is often possible that a spike appears exactly at the precise high of the week or year and could be the only one you see all year it is risky just to short it or sell it instantly as some markets produce several failed signals which continue to rally higher.
The safest and least risky method is to wait for the trend to begin falling after you see a divergence. This is subjective to your own definition of how to measure the trend as “falling” but I would suggest waiting for a 8-20 period Exponential average to turn down before selling.
Once the trade is entered you can implement a trailing stop to allow maximum potential gains and if your style is one of wanting to take quick profits then it is wise to take only some partial profits and give the sell off a chance to go lower and exit the remainder when the trend changes. If the move was picked up near the absolute top it could be a very large collapse in the downtrend.
Sometimes you might wait up to twenty five bars after the divergence is seen before the trend begins falling. Much longer than this an it gradually negates the signal as it shows the buyers have become stronger and the safest decision is to stay out of the market.
It is not unusual for the divergences to mark the exact high of a market and this high can lead to a large move down.
There are however frequent “failed divergences” and these can be treated in the same technical analysis manner as a failed head and shoulders or failed double top where the failure to fall indicates a likelihood of a continuation higher, meaning it is time to cut a loss.
This indicator only gives sell signals. Every single signal will be given in some degree or another in an up trend at the highest high price.
Market selection is important.
Avoid markets in an endless up trend. Look for ebbs and flows in a major down trend.
Best results are on liquid markets in a good long term down trend that has frequent rallies, you can observe the past signals and often history repeats with the good previous signals tending to indicate that future signals may also be good. (This is not certain of course)
This is also true of a market showing several historically bad divergence signals leading to more bad signals.
If the past performance of this indicator is poor on the market you are viewing, then move to another market until one is found where the readings show good price dips after the signals in historical data.
Time frames.
This product can be applied to any time frame of market but be aware as is stated above, the slower time frames yield more valid signals and shorter time frames lead to more randomness and noise ridden plots of lower significance.
That said, it provides a valid reason to enter a trade and can give good results providing good stops and risk control are used. I have seen plenty of valid signals on 30 second charts right up to weekly charts.
Idiosyncrasies.
It can often be seen that multiple divergences occur over a range of ten to thirty or so bars during a very gentle spiky kind of rally.
This can be treated in the same way as above - waiting for the trend to fall after the last divergence occurs is the way to play it.
Groups of divergences can indicate some patient insider selling patterns in anticipation of some bad news they might know.
Thanks for reading this and please read it a few more times to fully understand the points mentioned.
After that please spend some time changing settings and markets to fully appreciate how it operates.
Roger Medcalf - Precision Trading Systems