New Members: Be sure to confirm your email address by clicking on the link that was sent to your email inbox. You will not be able to post messages until you click that link.

Problems with scanning for PnF patterns

Dear friends, I have some questions relating to scans using PnF patterns and understanding the results/implementing the scans:

- I have quite often a result accompanied by an indication "No PnF pattern found". How is this possible when using PnF scan parameters?
- The patterns found are sometimes far left in the chart and less useful
- Which PnF base is used by the engine (traditional, percentage...)?

My impression is that the Pnf scan is either not working reliably, or I have not figured out yet how to use it proberly. How are your experiences?

My favourite pattern is PnF double top breakout followed by a pullback, even below/above reversal point. Do you have any hints on search parameters? Are there other parameters than those which can be selected?

I would appreciate any hint! Thank you, Rob

Comments

  • lmkwinlmkwin ✭✭
    edited September 2020
    can you post your scan code? it's a lot easier to assist you with trouble shooting.

    I use PnF in scans. The PnF patterns are based on the Traditional scaling.

    If you go to the Predefined Scans page, there are a number of PnF scans at the bottom of the page

    https://stockcharts.com/def/servlet/SC.scan

    When you open one, in the upper right there is a "Click here to edit this scan". You can see the scan code for that scan there. If you are a Xtra or Pro member, you can modify and save how you see fit.

    This is a great way to learn how to build scans.
  • Thank you for your support, LMKWIN!

    I actually use the standard scan definitions with only little modifications to keep it simple.
    My problem shows up with the predefined scans as well. If you for example check the PnF bear scan, the results are quite late, usually way after 3 box reversal with a column of x, which is often too late for a good entry point. Taking the bear trap as example, I would be happy to have results prior to the reversal, i.e. in the Os. There are various results such as long tail up or a top breakout as well, but I look for the Os after the break in the Xs (and vice versa).


    Do you know any work around here to capture the bear trap timely? Thank you!
  • ... a good example of what I am looking for is the BTU bear trap on Sept. 16, showing timely up today using the PnF bear trap scan, whereas a patern from July is less useful.
  • lmkwinlmkwin ✭✭
    edited September 2020
    a Trap isn't a Trap until it reverses, so you can't "catch" a trap sooner than the reversal. A trap starts with a multiple bottom or top break, like a triple top or triple bottom. Perhaps you can look at those patterns instead. These are traps before the reversal. Of course, the stock may not reverse



    If you want to get only the NEW patterns from today, you would modify the scan code to say that the pattern today is True and it wasn't True yesterday.

    Below is the scan code that your BTU example shows on.

    [type = stock] AND [exchange = NYSE] AND [Daily SMA(20,Daily Volume) > 40000] AND [Daily PnF Bear Trap = 1]

    The "Daily PnF Bear Trap = 1" is the same as saying Daily PnF Bear Trap is True. The scan returns all of the symbols that the current last PnF pattern is Bear Trap regardless of when it first appeared.

    To get the ones that are True today and weren't True yesterday (or whatever threshold you want) you would modify the code to say

    [type = stock] AND [exchange = NYSE] AND [Daily SMA(20,Daily Volume) > 40000] AND [Daily PnF Bear Trap = 1] AND [1 day ago Daily PnF Bear Trap = 0]

    You could also use the True/ False by doing this:

    [type = stock] AND [exchange = NYSE] AND [Daily SMA(20,Daily Volume) > 40000] AND [Daily PnF Bear Trap is True] AND [1 day ago Daily PnF Bear Trap is False]

    In the scan engine there are also other options like

    and [PnF Chart In Xs is true]

    If you want to see symbols that are in X's today and weren't in X's yesterday, you'd modify the code in the same way we did for the Bear Trap. Today is True, Yesterday was False.

    There's also one for Box Count

    and [P&F Box Count > 10]

    I utilize this one more than most any other as I look for something called a Dynamic CSS pattern. You can search the forum for discussion on that.

    Hope this helps?
  • Thank you so much for your input and time you spent on my questions.

    I have found your contributions in the forum - your CSS approach is really interesting and I think I understand the reasoning behind it. I have also found your book suggestions and will follow them.

    I am slowly leaving the traditional charting world where I have been for many years, and am in PnF education mode to develop a PnF system reflecting my market understanding (I look e.g. for the end of the accumulation zone (long), preferably for a pullback following a breakout, or for "irrational" pullbacks, and appreciate any hint, also with regards to educational sources.

    Thank you again, Robert
  • lmkwinlmkwin ✭✭
    edited September 2020
    A good blog on StockCharts.com to follow is Wyckoff Power Charting. Wyckoff used PnF charts in his work. In a couple of episodes on scanning for Wyckoff patterns, they provided an excellent suggestion of using the multiple top and bottom breakouts as a starter and then review the charts for subsequent activity to see actionable situations. Excellent suggestion. Multiple tops would be triple top, quad top, quint top, etc, and the spreads, spread triple top, spread quad top. The multiple tops are an indication of consolidation. Their additional analysis is conducted on time charts.

    There is also a new author, Prashant Shah. He has a great book on PnF that, I call a tremendous bridge for those that are transitioning from traditional time charts to noiseless charts. Noiseless charts are PnF, Renko, Kagi, and 3 line break charts. They only reflect "significant" price movement. How significant? That is all in your settings and personal time preferences.

    I've been a student of PnF for over a decade now, and he introduced some concepts and terms I had never heard before, but have impacted my understanding and analysis. He is active in the India stockmarkets, owns a trading software company and does educational blogs and videos, as well as books. I posted a thread on this forum when he does a new Twitter thread on TA. Below is a link to many. I find his explanations very interesting and educational.

    https://www.definedge.com/market-concepts-simplified/


    On StockCharts, you can't default to a PnF chart so I use a Renko chart as my Default chart. Renko charts are approximate equivalents to a PnF 1 box reversal chart.

    I call learning PnF is like learning a new language or sport or skill. It takes time and practice and patience. Took me a few years to get comfortable enough. You need to know yourself and what you are looking for (which you defined in your post), and then find what it looks like in the charts.

    My favorite patterns are Poles. The High Pole was identified by the late great Earl Blumenthal. A pioneer in PnF world. The Low Pole was identified by the late great Mike Burke.

    Congrats on taking the steps to start down the PnF path.
  • Thank you for you inputs, and I really appreciate your willingness to share your knowledge.

    I watched the behaviour of some stocks I know on PnF buy signals signals during market hours today and had some possibly interesting observations regarding institutional buying, but need to verify this over a couple of trading days.

    May I ask you two questions:
    - are you using the percentage style, i.e. 1% x2, or are you using the traditional style?
    - would you be willing to share more details on your dynamic CSS concept? I would be really interested in learning how you implement it. I guess I understand the reasoning behind it (capturing the end of the accumulation after sell-off)

    It feels like discovering the quantum nature of light after spending many years with wave theories :-)

    Thanks, Rob

  • @Robertplayitloud, I use both. The Traditional scale is what is readily available on most websites and is used as the scale for scanning and screening on StockCharts.com. It would be interesting to be able to scan on % scale or other scales but you can't. I did a free trial on Updata, the software that Jeremy DuPlessis helped develop. They ONLY use the % trend scale, but they are severely limited, in my opinion, when it comes to patterns and complex patterns. I passed on signing up after the free trial. All of my inquiries as to screening options were "we can probably do that with custom programming" i.e. additional charges.

    I "find" my securities using scans, therefore, on the Traditional scale. I then review the charts using % trend. I find % trend an excellent additional tool. I also recommend using CLOSE ONLY prices as one of your analysis options. High/Low and Close Only. Close Only removes more "noise" from the charts.

    My usual review is Traditional 3 box reversal. Then % trend 3%x2 both High/Low and Close Only. I may also review a 3 or 5% x 1. I use the 3%x2 close only as my final decision maker on a buy.

    The CSS stands for Consecutive Sell Signals. A sell signal is a bearish indication. Multiple sell signals usually confirm a bearish move. Something to keep in mind with PnF is that the names Sell signal and Buy signal are not intended to be directions as to what you should do. You will not do well, over time, if you buy all buy signals and sell all sell signals. They are just names for a pattern. Patterns are divided up into two sides. Bullish and Bearish,

    So in my "system" I look for at least 3 consecutive sell signals to imply significant bearishness. After 3 or more CSS I then look for a column of X's that is at least 7X's tall. In most charts, depending on the price levels, a column of 3, 4, and 5 in length are pretty common. A column of 6 is less common. A column of 7 even less common. The longer columns are usually where a change of character occurs.

    So after a period of significant distribution or bearishness, I look for a period of significant accumulation or bullishness. A CSS followed by a 7xx (or greater) column. I also substitute 2 columns of 6xx as long as the 2nd column of 6xx has a higher bottom than the 1st column of 6xx. The 2 columns of 6xx need not be consecutive columns.

    After these bearish and bullish indications are shown I will then review the charts on the alternate scales and look for another reversal or "confirmation". The confirmation is a reversal to O's and then back to X's. The confirmation can be on the alternate scales, but I need to see the initial bullish indication be backed up by additional bullish indications.

    Afterall, that initial 7xx could just be short covering or some other head fake news/rumor. It's preferable if the last or the 2nd to last sell signal in the CSS series have the longest column of O's in the series of Sell Signals. This is usually a sign of exhaustion.

    Hope this helps answers your questions.

    I do have a public chartlist on StockCharts that lists Dynamic CSS patterns that occurred in the last few months, from the SML, MID and QQQ universe.

    https://stockcharts.com/public/1175948




  • Dear Andrew, many thanks!

    Your CSS approach is really interesting. I can imagine that the RS (symbol:sector) picks up at the 6xx or 7xx, which could be another confirmation, which, however, is perhaps inherently included in the PnF chart, anyway.

    Thank you for sharing your chart list! Interestingly, there are some stocks that I follow as well, so I hope that I can better understad the way you deploy PnF. I hope I can return the favour somehow!

    With my best wishes, Robert
  • I used to incorporate RS into the analysis. I do show ratio on my time charts but it's a secondary or tertiary at best, in my opinion. The price, the potential via price objectives and expected resistance and support, the DNA of the security via seasonality, etc play more into my decision making. A security that hasn't shown the capacity to increase in value for extended periods of time would be an example of DNA. What is the history of the symbol?

    The public ChartList is fully updated (new symbols added and removed) to keep the recent patterns showing once a month. The list is re-sorted regularly, usually by 1 month return, but sometimes using a scan to sort via the Rank By statement. Sometimes I'll use a weighted RS as the rank by.

    If you "Follow" the list you get an email from StockCharts reflecting when the list was updated. If you do follow it, and you find it helpful or interesting, vote for it.

    To "return the favour" you can keep us updated on how your transition is going. I'm sure many posters and lurkers would find it useful and helpful. And of course, keep asking questions and sharing answers on this board.


  • lmkwinlmkwin ✭✭
    edited September 2020
    I also should mention that I consider a High Pole to be a sell signal, and use it in my CSS count.

    In Prashant Shah's book he talks about patterns and their reliability or backtested performance on various scales. The 100% pole on a 3% or greater scale was one of the best in terms of performance. A 100% pole is a retracement of the entire pole column. A traditional pole is defined as a greater than 50% retracement.

    If you look at the example chart I posted earlier you may be able to see a couple Dynamic CSS patterns. 3 or more consecutive sell signals followed by a column of X's at least 7X's high.
  • I have a very similiar approach - my stock DNA is a SMA over RS on weekly, which unfortunately cannot be implemented yet in Stockcharts.

    First observations after a couple of months:

    - PnF acts like a controllable quantizer and surpresses a lot of noise. The typical stages such as accumulation and mark-up can easily be recognised, springs are surpressed, and the signals are crystal clear.
    - I am still verifying PnF using traditional charts, but it is like hindsight analysis, unbelievable.
    - I also have the impression that MMs cannot easily target the PnF stop loss levels with their long wicks.
    - I wished I could process the PnF data to some extent (annotations, prop. indicators, etc). I also wished I could store global settings in Stockcharts that are used across my account with different browsers/computers.

    I do not understand how the MAs are calculated over PnA, but I spotted some MAs on your charts. How are you using them?

    I will keep you updated on my transition progress, but I hope not to bother you too much. You have already helped me a lot.
  • a few responses

    " my stock DNA is a SMA over RS on weekly, which unfortunately cannot be implemented yet in Stockcharts."

    What is the sticking point? You can get most things to work on StockCharts

    Your - comments
    Observation 1 Not necessarily. It all depends on box size and reversal and type used. You can make a PnF chart very "noisy".

    Observation 2 You can also review on % trend. Depending on the security price it's often better. Default box size Traditional scale on a stock from $20-100 per share is 1. A box on a 25 dollar stock is approximately 4%. A box on a 85 dollar stock is a little over 1%. So apples to apples against a time chart is usually better to use the % trend to compare. I haven't mentioned the month markers. These also should play into analysis. Connecting a line through the last 3 month markers will also provide some insights as to direction.

    Observation 3 If using High/Low charts you can see where the wicks are for the most part. Stop Losses are an interesting topic. I don't use them as, although they can reduce your risk, they will reduce your reward.

    Observation 4 I would love annotations as well as Horizontal Lines and Dividend in the price header. Have made suggestions like that to support. You can save your preferences for ChartStyles that will remain on your account on any browser. You can create default Time chart as well as custom CandleGlance and Market and Industry Summary charts, GalleryView, as well as any other type/style you like, up to the 50 chartstyle limit.

    You can also change your default PnF chart style. Make your changes, click on Update and then click on Save as Default. Your new PnF chartstyle will be your default when you click on a PnF chart as well as it will be the default PnF used in GalleryView.

    "I do not understand how the MAs are calculated over...."
    In Prashant Shah's book he uses common technical indicators on PnF charts. In PnF world, the indicators are structured based on the PnF chart. RSI is based on boxes filled. You mention MA. The Moving average on a PnF chart is based on the mid point value of each column. On my charts the 1, 7, 9, and 20 column MA are shown. The 1 column traces out the midpoint of each column. This can be useful in determining where a 50% retracement lies as well as a great short term indication of where price support may lie. I look for W's and M's in that line. W's are bullish, M's are bearish. The 7 and 9 MA are intermediate indicators. You'll note how well they define/illustrate strong trends. The 20 is a longer term and often approximates where the traditional Bullish Support and Bearish Resistance trend lines would fall. Also fun to note the interaction of the 1 with the 7/9 and the 20.

    "I hope not to bother you too much"
    You can try but I'm never bothered sharing experience with PnF and StockCharts.com and technical analysis.
  • Dear Andrew,

    after countless permutations and verifications, I have arrived at the following initial set up, which reflects a lot of your input:

    - Scan and pre-selection: traditional 3 box reversal, in addition 3%x2, and 1%x2. PnF MA: 1, 8 and 21 (I can't do without Fibonacci somewhere :-))
    - Scan parameters: initially double top breakout to better understand the results
    - I still use bar chart price action for verification, CCI for divergence, along with some standard MAs, and the Wyckoff principles
    - I use trailing SL so well chosen entry levels are essential. PnF 1%x2 is helpful here

    The reason why I am looking at PnF is that the traditional bar charts appear to form an undetermined system for mastering the increasing market complexity with additional parameters such as Covid 19 and Robinhood :-).

    I will keep you updated.

    Best wishes, Rob
  • markdmarkd mod
    edited September 2020
    Very interesting discussion. I'm not a PnF person, although I get why people like it.

    I noticed in Rob's last answer he mentioned using the bar chart for verification and applying Wyckoff principles.

    I have a scan that picks up on the possible presence of Wyckoff-like accumulation, at least in some instances - it will pick up duds as well of course, but at least some of the results can be set ups for the approach of the mark-up phase, which should appear on the PnF charts.

    [[exchange is NYSE] or [exchange is NASD] ]

    and [
    [group is CommunicationServicesSector]
    or [group is ConsumerStaplesSector]
    or [group is CyclicalsSector]
    or [group is EnergySector]
    or [group is FinancialSector]
    or [group is HealthCareSector]
    or [group is IndustrialSector]
    or [group is MaterialsSector]
    or [group is RealEstateSector]
    or [group is UtilitiesSector]

    ]
    and [min(21, Fast Stoch %K(251,1)) < 20] // line 1

    and [sma(63, close) < 20 days ago sma(63, close)] // line 2
    and [close < sma(63, close)] // line 3

    and [sma(63, Force(251)) < 63 days ago sma(63, Force(251))] // line 4
    and [Force(251) > sma(63, FORCE(251))] // line 5

    Line 1 says the stock has recently been in the lower 20 % of its one year range - in other words, it's potentially sold out.

    Lines 2 and 3 say price is below the falling MA 63 - in other words, price hasn't yet recovered from a long fall.

    Lines 4 and 5 say accumulated volume (according to the Force indicator with a one year look back) is above its falling (or possibly turning up) medium term MA (63 days, or quarterly).

    If long term price is falling but long term Force is generally rising, it means that despite prices falling, buying is stronger than selling, which is the hallmark of Wyckoff accumulation.

    I want to repeat the results will turn up a lot of junk - stuff that is falling and isn't coming back - you can sort of tell those by how disorderly the chart looks, choppy instead of wavy - but some - maybe 1 in 10 ? - should have the potential to double or triple or more within a reasonable time, maybe a few months. But even good ones can appear to be on the way and then fail - smart money makes mistakes, too. The phase of the market matters, too. You will see more after a correction than now - but even now, sectors go in and out of favor.

    Here's an example of one that worked (from Dec 2018), that this scan should have found in the circled area (it actually came up on a scan for high up volume on a falling stock).

    Notice the sector RS is strong (top panel) and the industry RS in the sector is strong (second panel) but the stock's RS in its industry is weak (third panel). The fourth panel is Force, and you can see it is above the green 63 MA. It would take some guts to buy this while the price is going down, but when it turns up, you would be more likely to believe it because the "Force" is with you.



  • Dear MARKD, I am looking for stocks that had or have a rest for a while. PnF feels like a transformation of the price-time series into a "boxed" domain, similiar to a Fourier transformation of a time domain signal into frequency domain, hopefully providing additional clues.

    Thank you for sharing your very interesting scan! I magnified the stock you shared on hourly and even below in the accumulation stage. A perfect example of accumulation with a nice shake out in December, followed by a NRB on 26th December, 2:30pm, at exploding volume (insider buying?).

  • markdmarkd mod
    edited September 2020
    Exploding volume is usually due to news. ARPO was a newish drug stock, for which announcements are almost always a big deal. The big volume you mentioned could have been institutions (although this is a low price low volume stock, so maybe small funds, not big ones), but not necessarily the insiders doing the accumulating (by insiders, I don't mean company executives, I mean the fund or funds operating in the stock).

    According to Wyckoff, insiders accumulate quietly, first letting sellers come to them (most of November), then bidding up, but quietly, as supply dries up. If others catch on that something is happening, the operators cap rallies as long as they are still accumulating their stake. In November and early December, the red and black bars with higher highs - bars that open up but close down (red) or open up and close up but below the open (black) - and the long wicks, are operators selling short to discourage outsider buying before mark-up begins.

    In general, swing traders and investors don't have to worry about market makers or high frequency trading - their influence is mostly short term. Insiders are not a problem either. As long as trades are reported and show up on the chart timely, technicians have what they need to interpret what is going on - the laws of supply and demand have not been repealed and a weight of the evidence approach can still work. That said, the new insider trading laws, restricting the flow of fundamental information, have made technical trading harder because there is less informed buying and selling for technicians to detect in advance of news. So everybody gets surprised at once and many more lose than would have otherwise.
  • "and the long wicks, are operators selling short to discourage outsider buying before mark-up begins"

    I asked myself many times where this buying pressure is comming from, and now there is a perfect explanation, thank you!

    May I ask you how you can spot the operator's action and recognise after mark-up that a distribution stage is a real distribution stage and not just another accumulation stage? As far as I understand the distribution stage, stocks get choppy, volume picks up and upper wicks occur. Many traders place their SL somewhere below support, and operators enjoy fishing even at the alledged distribution stage. This was by the way one of the reasons for me to learn and deploy PnF.

    I am rather new to Stockcharts and am discovering this universe. There is so much profound knowledge and experience in this forum and on stockcharts.com, I wished I joined 10 years ago.

  • markdmarkd mod
    edited September 2020
    Well, you can't necessarily distinguish between "intermediate" distribution and "final" distribution.

    It's worth keeping in mind that Wyckoff-style operators account for only a small portion of all activity in the market. Just because a pool (today, a fund or group of funds) chooses to distribute when price hits a target profit doesn't mean that the stock itself has run its course. If the earnings outlook is still good, others may accumulate what the pool distributes and the stock continues up.

    Also, it helps to note that a stock can go up without accumulation and it can go down without distribution. Many stocks just move with the market. They aren't interesting enough to gain a following so no one makes a play to control them.

    On the other hand, some stocks are so widely owned that no one can get control of them. For those stocks, it's possible that Wyckoff's description of operator activity more or less coincides with "natural" accumulation or distribution - in other words, as institutions gradually change their view, they begin to accumulate or distribute collectively, but not in deliberate co-ordination - they just come to similar conclusions as the fundamental picture (interest rates, inflation, employment, consumer sentiment, etc.) develops. You are more likely to see pure Wyckoff patterns develop in small and mid-size stocks with a niche following, not market bellwethers the like the FAANGs.

    Note that while stocks seem to have repetitive (but widely varying) patterns, stock price movements do NOT follow natural laws. This isn't physics. Time between significant market events can be very elastic - stretching out over long periods or snapping in an instant. So it's very hard to get it right. So when you get burned, that doesn't mean someone is manipulating the market. The main thing is to protect yourself from things going wrong. There's a saying - watch your losses and the profits will take care of themselves.

    One way to see distribution is essentially the mirror of what happens in accumulation - big volume doesn't move the price in the direction of the trend.

    In ARPO above, there is a huge volume red bar in November, in a down trend, but it DOESN'T make a new low. If there is no new low on big volume in a down trend, then there MUST be (knowledgeable) buyers. Later on, in mid-December there is another strong volume bar that closes UP. Then again at the end of December a high volume bar again closes up and above several down closes. That last could be smart buyers, but more likely new (and mistaken) short sellers panicking out as the operators finish up their buying at new lows.

    So, in distribution the opposite happens. Big volume does NOT make a new high, as you expect it would in an up trend (or, it closes up, but range is very compressed for the volume, or it closes up but with a long wick, or it closes down, or it closes up, just not a new close, etc.) So, there must be knowledgeable (but not omniscient)sellers taking advantage of enthusiastic buying to dump pieces of their position. As with the accumulation stage, where operators will keep prices from getting out of their buying range until they can acquire their full position, in distribution, operators buy buy buy to keep prices from falling too far until they have finished unloading.

    You can sometimes see the effects of accumulation and distribution in the Force(251) indicator. In ARPO, Force did not make new lows with price. In a distributing stock, you would expect to see new high prices without a new high in Force. What's remarkable is how long prices can continue to rise (or fall) after the first indications of distribution (or accumulation). But the divergence is at least a warning to give more weight to bearish (or bullish) activity when it finally comes along.

    A final point to make - operators also sell when they accumulate and they also buy when they distribute. In order to control a stock, an operator must have a significant amount of both cash AND stock. Even though they are building a position to sell long term, they also need stock in the accumulation phase for the short term so they can short to contain prices in their range. And even though they are building cash in the distribution phase, they need to use some cash to prop prices up within their range. So, operators are always both buyers and sellers, and always need to replenish their cash position and stock positions for operations - so you will always see them (e.g. running stops) on both sides of the market regardless of the market stage.

  • Thank you!
  • Another effective way to find changes in character is to look at Selling Climaxes and Buying Climaxes. But to review them several weeks after they happen. Tops and Bottoms take time to form. A climax can be indicative of a change beginning and in hindsight. You can run scans to find climaxes that occurred back in time.
  • Dear Andrew,

    this is exactly what I tried over the last couple of days, with your and Mark's input. Tried several searches for double bottoms after sell-off and double tops after extended moves, using both, PnF and the traditional search as well.

    BCDA for excample shows in PnF traditional 3 a nice double bottom, and you can see in PnF 3%x2 a nice double top brakout. I took BCDA as example to tweak my PnF and traditional search, and was less successful.

    BCDA has no double bottom breakout pattern at traditional 3 reversal, so I simply tried "and [double bottom is true]". There was no error message but the scan did not find BCDA:

    [type is stock] and [sma(20,volume) > 10000] and [market cap < 100]
    and [ [exchange is NYSE]
    or [exchange is NASDAQ]
    or [exchange is AMEX] ]

    and [ double bottom is true]

    I then tried

    and [ p&F box count >6]
    and [ PnF Chart In Os is true]

    to find the prior sell-off column ofO's, but I guess p&F box count looks for the most recent column count. I have not yot found a possibility to search for the bottomed sell-off. I can find the ongoing sell-off or extension with PnF count. Sure,

    and [ p&F box count =3]
    and [ PnF Chart In Os is true]

    worked well but this search is rather unspecific, as the stock can be within the 3% reversal, or even not.

    I then tried to find the double top breakout prior to the mark up stage, with

    and [ PnF Chart In Os is true]
    and [double top breakout is true]

    which did not work, neither (which is a little bit confusing as some other scans show double top breakouts that took place at a certain date...). I have not found yet a possibility to specifically address different columns in a matrix fashion, soch as column i, column i-1 etc. in PnF. I would also be happy if there was a possibility to select the PnF style for searches to look inside the traditional columns.


    I therefore think of connecting both, the PnF search and the traditional search in one search to reduce screening time.

    Unfortunately, I am not a that familiar with using too many or a combination of various indicators and the way they work together for obtaining more accurate scans ( I found combinations of SCTR, Stochastics, Force etc. on this forum, but I have not understood how these combinations work so I am not using them blindly).

    Sure, I use RSI for oversold and overbought, PCTDIFF fo find stocks that rest near a MA ( I use MA 8, 21, 35, 55, 63, 100 and 200, depending on stock and time frame), but the question is how to find stocks at the reversal. I am swing trader in different time frames, by I guess that time frames can be adjusted using (traditional) parameters.

    I would be happy to use both, the PnF search and the traditional search in one scan for bottoming/topping or reversing stocks order to reduce my screening time, and would appreciate any hint.


    Thank you, Robert







  • A couple of suggestions and observations


    BCDA is a lottery ticket. Not a lot of trading history. At the top of each PnF chart it shows the pattern. In the case of BCDA, the reversal to X's negates the prior pattern. I believe the prior pattern would have been a descending triple/ quad bottom breakdown. So you can see why your 1st scan wouldn't pick up BCDA.

    A way to capture prior patterns is to cast a wider net. The code below will return a Low Pole pattern that was true in each of the days selected. You can pick and chose whatever patterns and days you would prefer in this manner.

    and [[10 days ago PnF Low Pole is true] or [15 days ago PnF Low Pole is true] or [20 days ago PnF Low Pole is true] or [25 days ago PnF Low Pole is true] or [30 days ago PnF Low Pole is true] or [35 days ago PnF Low Pole is true] or [40 days ago PnF Low Pole is true] or [45 days ago PnF Low Pole is true] or [50 days ago PnF Low Pole is true] or [55 days ago PnF Low Pole is true] or [60 days ago PnF Low Pole is true] or [65 days ago PnF Low Pole is true]]

    Will it catch every low pole in the last 15-65 days? No it won't, but you can get the idea of how to cast a wider net from it.

    "and [ p&F box count >6]
    and [ PnF Chart In Os is true]"

    as you can see in the chart, the current PnF box count is less than 6, so it wouldn't pick that up.

    "and [ PnF Chart In Os is true]
    and [double top breakout is true]"
    Once again, the pattern is negated by the reversals

    But the stock is still on a buy signal or a sell signal. There is a scan PnF option for this. Both Buy and sell

    and [PnF Buy Signal is true]

    https://school.stockcharts.com/doku.php?id=chart_analysis:pnf_charts:pnf_alerts


    I concur with your statement about understanding indicators to find out how they work. TA is all about learning and understanding what the indicators are telling you and if you should listen.




  • edited September 2020
    Thank you! I selected BCDA for my "experiments" a difficult stock in a difficult sector, which I am not trading, anyway.

    I think I understand now a little bit better how the search engine is working: a signal given by a pattern is valid until negation such as reversal. A pattern sich as low pole is not a signal, and it is valid as long it is there. With this in mind the results make sense. Is this correct?


    Today I tested what I have learned from you and Mark so far in my trading account. I have the impression that I feel the price action of the stocks better, and I definitely see more clearly the structure of a stock with PnF. PnF is also a very good verification of what I see on traditional charts, anyway.

    Prashant's book is my next challenge :-)

    Thank you so much for your patience!

    Best wishes, Robert

  • I think you're getting it. A lowpole is a pattern but you can't have a LowPole with the current column in O's. Same with a Double Top Breakout. Can't have one where the current column is in O's.

    O's are pullbacks, or consolidation or continuation. O's lead to Opportunity. O's are good, as are X's. O's lead to bottoms, and X's lead to tops. There has never been a bottom that isn't indicated by an O, nor a top that wasn't an X.

  • Thank you Andrew. My PnF scans are now working - it took me a while to understand the PnF search engine, and I now get the results I am looking for.

    Once again please allow me to express my gratitude for your and Mark's support and patience.
  • Am I correct in assuming that the SMAs on a scan are different than the SMAs on a PnF Chart Overlay. The Moving average on a PnF chart is based on the mid point value of each column. A 1 would indicate the mid point value of 1 column. A 7 would indicate the mid point value averaged over 7 columns, etc.
  • lmkwinlmkwin ✭✭
    You are correct. The SMA on a PnF chart is not available scan engine data. You are also correct on your description of the SMA's on Pnf charts.

    Depending on the price activity of the chart you can sometimes get apples to apples closeness with a time chart SMA's but, it's usually a "puts you in the ballpark" close. In GalleryView you can set up custom styles on the charts. In your PnF chart, if you click on the Save as Default link, not only is the version on the screen saved as your default PnF chart, it's also used as the PnF chartstyle in GalleryView.
Sign In or Register to comment.