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.

M.a.c.d. scan

Hey guys,

I'm trying to do a Scan for MACD that looks for a similar pattern to this. but, preferably not using the exact numbers on the High end. Something that starts High, and then is CURRENTLY Just below 0 Like this, but maybe with a Little bit of tolerance so that it catches more stocks. But, generally, starting high, then coming to just below 0.

Kinda the general gyst of this though. Starting high, then ending just below 0. Any thoughts?


  • Click Picture.
  • markdmarkd mod
    edited May 2023
    Something like

    and [0 x MACD Line(12,26,0)]
    and [20 days ago max(5, MACD Line(12,26,9)) > 0]
    and [max(25, MACD Line(12,26,9)) = 20 days ago max(5, MACD Line(12,26,9))]

    So that says, today, MACD crossed below zero (or literally, zero crossed above MACD - same thing),

    and in a 5 day period 20 days ago, the best MACD value was greater than zero,

    and, the best MACD value in the last 25 days was the same as the MACD value in that 5 day period (in other words, that 5 day max was a peak, or a point somewhere in the middle of a down sloping Line).

    I haven't tested this, but it should be something like that. You might have to play with the parameters.

    If you do, the first max value (parameter) in the third line (25) should equal the max value (5) in the second line plus the days ago value (20) in the second line. The "day ago" side of the third line should be the same as in the second line.

    Pattern scanning is pretty hit or miss - you only get charts that match the scan exactly, some (many!) of which are not what you really want, and you don't get charts that may vary only slightly from what you scanned for, but are very much what your are looking for. You can work on tweaks forever, but at some point you have to settle for what seems best, not keep looking for perfect. Also, as the market moves along, you may not get very many hits and then more than you can handle - that's because most stocks follow the general market pretty closely, pattern wise.
  • Cool. Thanks markd i think i can work with that. It will be really nice when AI takes one look at a pattern and can Find matches, Kinda like ChatGPT. i realize that's quite far off, but 'ya never know. thanks again.
  • markdmarkd mod
    edited May 2023
    It will be interesting to see how markets adapt to AI trading - or how AI will adapt to markets. Assuming there will be multiple AI machines trading, how well will they come up with strategies to anticipate each other? And how will having no fear affect their capital allocation decisions? How long they will let a trade run for or against them before closing out? Presumably it will be based on past market behaviors. Initially, those will be human behaviors, but eventually the AI machines will be using statistics produced by each other's behaviors. I imagine, too, that initially humans will write their rules of engagement, but since they are self-learning, those rules will be re-written. Then what? Will that make the markets more volatile, or less? Will humans be able to trade profitably against the machines?

  • yeah, it scares me if too many people begin to use it, then there will be no way of Anticipating price movements. But, would b nice if it worked for me-and-you. :blush: Am i just being greedy, idk. Who Doesn't want to make more money? Who Doesn't want an Edge? But, also it could make Jump-on-the-bandwagon trading even more ubiquitous; buying NFLX, aapl, msft and all those Fang-types. It's probably a double-edged sword that should be used Wisely.
  • lmkwinlmkwin ✭✭
    Two that are currently claiming to be using it exclusively are AIEQ and Let Bob (LETB). AIEQ says that IBM's Watson is making all the decisions. "Let Bob" has a hgher expense ratio and is currently showing 100% cash as its holdings.

    "About AIEQ ®. Artificial intelligence powered ETF that utilizes IBM Watson to equal a team of 1,000 research analysts, traders and quants working around the clock. The first actively managed ETF to fully utilize artificial intelligence as a method for stock selection. Analyzing millions of data points across news, social media, industry and ..."

  • You Right. I Wouldn't want to go up against IBM Watson, any day of the week. I tell ya that much. I'll have to look into this a little more. Sounds interesting. I am of the mind though, that all these Algorithms by the Big Dogs like goldman sachs or whatever are already Mucking things up a bit, too much.
    Just when you think you've Mastered human Psychology...
    Adds another dimension.
    Interesting Stuff, tho for Sure.
  • lmkwinlmkwin ✭✭
    Looks like Watson is under performing the SPX by a wide margin in the 18 month chart from above. Tough environment for a computer to be investing obviously. Tough environment for people too. Of course, they have only been publicly available for that period or less. Let Bob is 100% cash at the moment but that isn't entirely AI run, hence the higher expense ratio. Dogs gotta eat.

    These flavor du jour ETFs are a gimmick to get you to believe that you can't compete with the machines and big power players so you best join them. Marketing is a powerful tool.
  • markdmarkd mod
    edited May 2023
    From the description, it seems AIEQ is just WAY to complicated. True understanding ought to be based on really simple principles - like E=mc2.

    On the other hand, Malkiel argued the markets are random in the long run. Can AI deal with random ?

    It may be that Let Bob really is smart and it is in 100% cash because it has realized it can't.
  • Well, if I were a Betting-man, which I'm Not, I would buy AIEQ Only when it is above that S/t moving average (Maybe sell, on two -- perhaps back to back -- Red days)
  • I suspect it does well in 'Broad' market Upmoves only; Doesn't appear to be Setup to take advantage of Down movements, at all...
  • With LETB -- Most Likely a beatles Fan - I would Say buy it when it is Slightly below the sht tm moving average on a Daily Chart, or Slightly ABOVE on a Weekly chart. And, just Hold for Time being or poss. if MACD turns negative.
  • lmkwinlmkwin ✭✭
    I don't buy ETFs for very simple reasons. The holdings and selection process, and the management/rebalancing. I am much more comfortable assembling my own ETF of securities that fit my criteria. An author on a StockCharts blog recently talked about AI stocks.

    Biggest position in AIEQ, at the moment, is PLTR they added at 11.70. My guess is that AEIQ is a perpetual window dressing machine. Nothing wrong with buying PLTR at 11.70. But to make it the largest position (over 5%) out of 143 stocks is a bit odd in my opinion. Many of the stocks in AIEQ have decent looking reversal looking charts, so maybe Watson is on to a solid system and just needs a cooperative market. Who doesn't?

    I only consider ETFs that I can't duplicate or are outside of my SOP, like leveraged funds or some commodity plays or stuff like that. Even then, I still don't like them for very simple reasons.
  • markdmarkd mod
    edited May 2023
    PLTR at 11.70 looks like a breakout play.

    A Wyckoff play would have been would have been in Dec-Jan between $6-$8 - new lows with accumulation rising steadily since May 2022 (Force 52 weeks). The second week in May 22 is huge volume with a tail - meaning price at that level is getting supported big time. The last week of Dec 22 breaks the May low to flush out the stops. A few more weeks of accumulation, a breakout capped in Feb on huge volume, a range and then the "real" breakout.

    Maybe someone should come up with an AI Wyckoff engine.
  • was joking about the Beatles, btw. I actually Like that definition better ha
  • PLTR is an macd Play - From Oct. 31st to Dec. 30th down, Dec. 30th to Feb 17th up, Feb. 17th to May 5th down, Then May 5th to ... Up.
  • I'd wait for the Handle
  • plenty of Green volume, though. Agree, with the comments on ETF's they seldom track the underlying as they should, particularly doubles and inverses. In my experience anyways
  • lmkwinlmkwin ✭✭
    I believe that Wyckoff analysis wouldn't have started until after the sideways action, that started in May 22, was more defined as a sideways action out of the downward trend channel. The Selling Climax in early May and Automatic Reaction in June/July set the trading range approximately 6.5 to 9.5. The stock moved out of the down channel in late July/early August and then re-entered the trading range. A spring in December confirmed the bottom of the trading range. A spring in late Jan/early Feb, where the price finally broke above the prior quarter's high (green band is 9 week Price Channel), confirmed the top of the range. Volatility contracted from Feb to May '23. The breakout above 9.25-9.50ish would have been a Wyckoff entry point in my opinion. After it had a move above the upper blue/cyan line and came back to test it. The blue/cyan lines are chandelier exits with 30 week short settings.

    Note the Selling Climax in May '22 showed as a change on the KST on Volume. The panel below that showed short covering as evidenced by the rising CMF EMA(blue line). Also note the OBV didn't start to rise until 2023.

    Panel below that is the BB Width(10,2) with price channel highlighting contraction and expansion.

    On the Price panel I show the ZigZag indicator at 15%. Looking at any 4 lines of the ZigZag you will be able to see W's and M's. The stochastic line that moves from bottom to top and back again, is the stoch for the PPO. The change in direction indicates a possible good entry/exit. Flat spots indicate continued trend in that direction or sideways action in that movement and may indicate a possible good re-entry/exit.. That same stoch PPO is also shown in the top CCI panel where you can see the CCI moving out of oversold and into what Stan Weinstein would call a late stage 1 early stage 2 in Jan/Feb '23

    Of course, ALL of this is not really neccessary as it's all available to see on the Point and Figure chart.

    The throw over of the down channel in May '22. The trading range being established by the automatic reaction.
    The concept of parrallel trend line shows the move out of the down channel in July/August '22. The springs are evident with the move above and below the trading range. The contracting volatility is shown by the decreasing column lengths

  • lmkwinlmkwin ✭✭
    I should also mention that there is a Dynamic CSS pattern on that PnF chart with 3 or more Consective Sell Signals (CSS) followed by a column of X that is at least 7X in length. In the case of a stock price under $20 that 7X is reduced to 6X requirement.

    1st sell signal occurred after the movement out of the down channel in July/August '22. The numbers on the PnF chart are the month markers. 1 is January, 2 February.... up to 9 is September. A,B, and C are October to December. 2nd sell signal occurred in November, 3rd consecutive sell signal occurred in December. 6xx column occurred in January '23.

    I often find myself answering "After a Dynamic CSS Pattern" to the question "When would have been a good time to get involved with this security?". Full disclosure, I have no position in PLTR.
  • markdmarkd mod
    edited May 2023
    "The breakout above 9.25-9.50ish would have been a Wyckoff entry point"

    Yes, it would be the safe entry point. An aggressive entry point would have been after the successful Dec test of the May low, if you felt that the evidence at that point was sufficient to indicate professional interest. The evidence would have been the the May low - massive volume leaving a tail at a new low in a long down trend indicates professional buying (short covering and/or value players) and that (most often) means the end of the trend. You could enter there (if the down trend is over the risk is minimal), but it's hardly an ideal entry because while you know negative interest is exhausted, you don't yet know whether positive interest will develop. So, you could be sitting on dead money. So, whether you enter there depends on how long you want to wait. It turned out to be most of a year. Force shows that during that time there was a lot of net buying in the ranges, and little net selling on the way down to the Dec test (see my chart link above). If the selling has dried up at new lows, there is no one left to sell, so again, it's a reasonably safe buy. The accumulation in Force suggests there really is buying interest and the stock will get "hard to buy", so when buying does appear, price should do well. Especially in a low priced stock, the lower your entry point, the fewer points you will need to make multiples of your entry price.

    I always thought the main point of Wyckoff was the principle of accumulation without raising prices (and at tops, distribution without lowering prices). The details - signs of strength, etc. are interesting but they don't always occur, they don't always appear in the same order, they can be ambiguous, and they can be hard to identify in real time. But the logic of "hidden" accumulation or distribution by knowledgeable players seems sound.

    P.S. I find the Force method of calculating net volume more refined and logical than OBV or CMF. But that's another discussion.
  • lmkwinlmkwin ✭✭
    I agree with you @markd . That December test, led to the Dynamic CSS column in January and that afforded a lower buy point range of approximately 7.5-8.5. My "Wyckoff" analysis was based more on what I imagine Bruce Fraser and Roman Bogomazov would be saying if they were to analyze the price action. I don't think they would be recommending the aggressive buy at that point as it was moving out of Phase B and into Phase C. I think they look intently at Phase C for possible action points as it moves in to Phase D.

    They would probaby use a chart like this

    Bruce shared one of his simple go-to scans to look for possible candidates for further analysis

    and[weekly close > weekly sma(39)] // the weekly close is greater than the 39 week SMA
    and [3 weeks ago weekly sma(39) < weekly sma(39)] // 39 week SMA is rising in the last 3 weeks

    Rank by [SCTR]

    Then he says to not look at the highest SCTRs as they are in well established trends, but to look at the lower SCTR levels for developing trends.

    I don't know why you wouldn't just put a SCTR cap in but perhaps it was just for display on the show.

    //SCTR filter

    and [SCTR > 50] and [SCTR< 80] //SCTR is greater than 50 and less than 80

    Grayson also says to look for the developing SCTR strength and not at the highest SCTRs as they are the most extended and likely to pullback.

    I don't like or use SCTR as it puts too much weight, in my opinion, on being extended from moving averages. The further extended, the higher the score. And the gap from "long term" to "medium-term" to "short-term" is not balanced. And giving 10% weight to short term is not going to help identify developing trend, in my opinion.

    Joe Rabil covered PLTR on his recent show with his multi period ACP analysis. PLTR looked like it could be good for a trade if the weekly ADX moved up and the monthly macd confirmed the weekly and the daily close stayed above the 18ma it could be ..... he doesn't look at SCTR.

  • Well, there are advantages to waiting. Your capital isn't lying idle while you wait for the move to develop.

    I don't doubt Fraser/Bogomazov's knowledge of Wyckoff, but they seem to use it as a method for interpreting ALL market action in any stock. Wyckoff's intent, I believe was to spot "operator" activity on specific stocks, and not to explain movements of the entire market. It wasn't meant to replace Dow theory. Many (most!) stocks move in sympathy with the market, without any special help (or hindrance) from "operators". I doubt very much that any operator (fund, or syndicate of funds) has the clout to move, say, Apple or Google or Microsoft. Those stocks respond to changing perceptions of broad fundamentals (economic and company specific) which are too widely known to allow a cabal to operate secretly. But funds can move small and mid caps (story stocks, before the story goes public) and that's where I think Wyckoff is most useful.
  • Well, if Warren Buffet & Bill Gates Announced that they were Selling massive Amounts of OXY, at the Same Time...
  • Haha joking.
  • Haven't really used Wycoff that much, even tho i find it interesting. Any books or materials you particularly recommend, on the subject?
  • "Three Skills of Top Trading" by Hank Pruden chapters 4-10 explain Wyckoff pretty well. Probably available on Amazon.

    Wyckoff's original writings have been - maybe still are - available in pdf form on the internet. Try a google search. But, some pdf sites are malicious, so it's probably best not to click on anything unless you have some kind of antivirus/malware subscription running on your machine.

    Also, Bruce Fraser's blog may have some links to source materials
  • lmkwinlmkwin ✭✭
    A top Wyckoff principle is to apply the methodolgy to the determine the trend of the market and then select securities aligned with the trend. The composite operator motives and actions is what he was looking to figure out in his analysis.

    I wonder if a multiperiod correlation could be useful, but sliding it back instead of using different correlation values. Like correlation 100 > 0.7 and 50 days ago correlation 100 > 0.7 etc to find things highly in step securities with the market movements over various periods. Maybe then apply a PctRelative in the same manner.

    I believe that the Dynamic CSS pattern on Point and Figure charts that I follow may be a Wyckoff type approach with the requirements it uses. Signficant distribution followed by significant accumulation. It would be great if added a couple of filters for PnF charts. One would be the ability to scan for consecutive or multiple sell or buy signals. I have this on another site (DorseyWright). It's very helpful. It would be nice if they added a couple filters for SharpCharts as well. One would be regarding relative strength and new highs. I can get a list of these from another site (IBD). I think that is an area where StockCharts can improve a bit. They have the RRG but you can't scan on it. The RRG ratio line is close to a Mansfield Relative Strength The have the Mansfield RS filter available on

    Maybe those will be in the huge, huge, huge update coming.
  • markdmarkd mod
    edited May 2023
    Actually, in practice, the Wyckoff approach requires some subtlety not discussed in the usual explanations of the method. (but I agree with your point about Dynamic CSS).

    I agree the sector and the industry should show relative strength, but part of the subterfuge of the operators is to hide the stock's strength until they are ready. They want to accumulate at low prices, so they will discourage relative strength in the stock itself.

    In the link above, the top three panels show different relative strengths - the sector vs the market (xlk:$spx), the industry vs the sector (djussw:xlk), and the stock vs. its industry (pltr:djussw).

    XLK shows strength in Jan 2023 by crossing it's 3 mo moving average and staying above it.

    DJUSSW (dow jones software industry index) does the same in Mar 2023.

    But PLTR's RS looks weak between Mar and May - crossing under its MA until the May breakout.

    Notice PLTR first broke out in 3rd week of Feb 2023, but was swatted down - see the long wick on high volume after tall up bar. That is the signature of operators containing the price to discourage other "non-syndicate" operators or the public from running away with the price. The following range and down leg is an attempt to shake them out and disguise their interest in the stock.

    The operators have to do this to maintain control of the price up to their target price. "Control the price" - really "manage the price" means having enough stock to short to contain rallies (as they did in Feb 2023), and enough cash to support selloffs (as they did April into May 2023). They need to hit the price target because they have financed their operations by borrowing millions and have a nut to crack before they make money themselves. If other operators get control, they may decide to sell out before the syndicate hits its target. If that happens, the selling may attract a following that the syndicate cannot absorb, they don't hit their target and end up in debt.
    (P.S. a deep dive into P&F can tell you the target - unfortunately, it doesn't tell you how long it will take to get there or the draw downs on the way).

    The intent of the syndicates' operations is to "paint the chart" to make PLTR looks like a weak sister until they are ready - but in fact it's a prime candidate. The indicators don't tell you that, but recognizing the syndicate's footprints does (the long tail on high volume at a new low on the weekly chart - link in previous post, the Feb swat down on high volume, the rising long term Force showing accumulation, not confirming new low prices). Well, it helps; syndicates are not clairvoyant and they do sometimes fail.

Sign In or Register to comment.