Replication in price action manipulation

For those of you who remember Matrix movies, you might remember Agent smith replicating many times to increase his efficiency and to be omni-present at multiple places. When we talk about the concept of replication lets first break down the basics:

If this is the basic definition of replication, what makes replication different from a "pattern"? For example, in markets, people often bring up term patterns. "This is the pattern I trade" and they present you with a visual representation. Patterns are typically meant to have quite a wide variety of what identifies an X pattern. In other words, an orange is an orange but a hybrid between apple and orange is also an orange in most traders' books when it comes to identifying the "orange" pattern. There is a lot of sloppiness and rule-bending towards subjectivity about how patterns are often identified.

With replications at least for how I use them the rules are far more strict. So strict that some traders might say "There is no way to have that narrow identifying scheme in markets, as markets don't work like math. Nothing copies itself so identically as within a replication." While yes true-pure replication as within biology or chemistry which could be up to 99% will not happen in markets, the similarity can get actually quite up there every so often above 85s%.

In my view (not scientific but more of laymans phrasing) difference between pattern and replication is that pattern can have wider time exposure for how far back we look for matching samples between two charts that we identify as potentially the same patterns.

With replications the time horizon is very tight, or in most cases for me just within a single ticker on the same day. So the focus of replication is to find highly repeatable behavior on the same ticker, done by SAME players. And that is the key in my view.

Its not so much about a pattern, as it is about "read on the opponent" as per tactical sports, competitive games, warfare etc.

"Too many chefs can spoil the rigged soup"

Any pattern across a stretched time horizon can be formed by different market makers or institutions, leading to more likely discrepancy and slightly different outcomes or different visual looks of chart patterns.

It goes back to the saying that if you have too many chefs that are cooking the soup, the net result over 100 samples will be much more likely diverse, uncontrolled and randomized, than if you just singled out 1 chef to do the task. Assuming all those 100 chefs (and the 1) know what they are doing. It's the same with market makers.

This is why in my view replications can be a significant insight into price action because:

-If we have repeatable behavior on the same ticker within the same day, it is statistically much more likely being done by the same MMs / riggers, than if we have patterns/tickers spread out 1 or 3 months apart. That's just a mathematical guarantee.

You might wonder why this matters, to have one market maker instead of many? Well lets return to the point about the cooking the rigged soup above. If you had to make a bet with predictable outcome of the soup, where would it be more likely to establish an edge:

-scenario where 100 chefs cook soup every time over 100 samples

-scenario with 1 chef cooks soup over 100 samples

Obviously its abstract question because a lot of it about the process is not defined well, but its something to get the argument going into the article and why the lean towards single rigger/chef is better for outcome predictions.

So why would we that much prefer to have "only one" market maker, the same guy? Because if you develop the read on that same "guy" it is possible to extract edge if that participant keeps using same trick.

To illustrate this very briefly let's use the poker analogy. Think of a poker player that you develop read on, but he doesnt know it, and he is very repetitive. He should be walking away from the table as he has a read on from opponent. The same rules apply to poker as to markets. Some players you will have read on behind the poker table, and some you won't. Some will have certain traits that make it very unreadable for you whether they are bluffing or not, and some will be for you readable. For some market makers you just won't get a read on, over the hours the ticker trades. And some you'll get a read on right away, especially if it is REPETITIVE behavior.

So to return to the prior point, what is this about "single market market" or "same guy behind poker table"? Well if you have that kind of leading participant that you have a read on, this might be leading to series of wins behind a poker table, or trades in markets. To achieve that it brings us to the prior point mentioned above. It needs to happen on the same ticker, on the same day, to increase the chances of the "same guy" orchestrating it all. Or a group of them, either way, if it's a single market maker or a group that works as an orchestrated wolf pack it doesn't matter. The outcome is the same for us.

This is where replications stand above patterns. They allow you to zone into, it's all about "read on the opponent" unlike pattern trading where traders are completely detached of who and why the price is moving as it is within possible pattern. Replications go a step above that, we try to zoom closer on the market maker, rather than sticking with "random market pattern". The downside? It's not a common thing. More about that lower under the rule of rarity.

Replications require (highly) manipulated price action

Placing "highly" into brackets because manipulations probably happen on a daily basis everywhere in markets, it's just that in many cases the manipulated activity does not have a leading role in price outcomes, or does not represent more than 10% of total order flow.

In certain cases of especially highly volatile instruments like crypto altcoins or smallcap stocks, the manipulations, and their order flow can become the leading flow on the asset, it is the leading determining factor pushing the price. This is what we want if replication were to be the useful tactic of choice.

We want as janky movements, with:

-hard rejects and pinbar candles,

-with as many fake breakouts,

-with as many support breakdowns that reclaim fast backup,

-with as many big spikes followed by fast collapses.

Basically, the price movements that don't make sense to be driven by random markets, instead suggesting being driven by large manipulators is what we want to have right environment for replications. This will happen every so often on one ticker. 

Replication needs certain trapping behavior that fits into, this is key. There is no point in seeking replications within some easygoing uptrend that fits the most basic price action guide for dummies. We want as cunning moves as possible, to:

1. Identify rigger (participant that dominates order flow so much they can push price from anywhere to anywhere even if it doesn't make sense).

2. To have the context of identifying one trap. Why? Because on replication the whole point is that the rigger will repeat the same trap two, three, or sometimes even more than 4 times within 8 hours of the NY session.

3. Example below replication on ticker IBIO:

To highlight what i meant about "janky" movements that signal rigged action, and what we need for replication potential:

Not all small-caps will have such price action, displaying junky manipulation activity. There are often days when you see a Disney version of price action. Ticker that pops up after opening and then goes into a smooth fade. A dream version of movement for some on the short side. Or a ticker that you buy in a strong market and it just keeps squeezing up into parabolic moves until the end of the day. This kind of price action will not lead to a usecase within replications and rigged activity. Ideally what we want is RANGE. Or a softer trend. If the price escapes too fast to either of the extremes it will not get enough order flow trapping done and will represent a lack of replication opportunities. So range or softer trend is our friend such as example on tickers KAVL and ZAP below where both tickers are constricted into backside trend for two hours:

Notice clean replication of movements on both tickers above (happening within the same day or a few days apart.

Same controlled walk-down with lower highs, the same first higher high within a trend that ends up squeezing, and it only squeezes above the highest high within the backside trend, tops out there, and sinks deep.

Price action clues to identify replications

It is very important to start with: It's all about price action. This approach does not scale to indicators more or less because it is more about details rather than smoothing out. Therefore it's impossible to constitute this with anything else but naked PA / TA for the usecase of replications. This is especially because no indicator is good at identifying manipulation price activities, and the core concept that replications are leaning on is manipulated activity. This should explain why there is mis-match between indicator+replication use.

Structures, legs and abnormalities

There are three ways how you should split your technical analysis oversight on daily basis:

-Structures (range structure, wedge triangle, Type1, etc)

-Legs (up move in price usually consisting of predominantly bullish candles for example)

-Single candle abnormalities (pin bars/fake breakouts/fast reclaims

-Observing price structures to identify similarities (1 hour of price actions structure on the left vs another 1 hour of PA structure on the right for example)

-Observing legs within price action to identify similarities (bull legs to bull legs and bear legs to bear legs and how similar or not they might be for example)

-Observing abnormalities such as perhaps halts into key liquidation levels (under major supports or above major resistances where stop hunting might take place, or big pinbar candles or fake breakouts around critical levels and if they repeat). We need to deploy the eye to the price action that specifically breaks down PA into multiple layers that helps us to create variables. Once we got the variables on chart down, then we apply it where it really matters: Comparison of those variables or comparison of zones.

Compare left to the right, on continuous basis through the day

This is where the core of the replication strategy comes into play. Everything we have outlined above is a method of what to look for in PA to build a scheme of how things can be compared. The next step we have to take is to then actually compare them on an intraday basis.

This means watching price action flows and charts, identifying the zones of price structures, and comparing left to the right.

What I tend to usually do is split the ticker into 6-8 zone segments across the day (always from a 1-minute time frame), and then compare those zones together (in no particular order, zone 1 can be compared to zone 6 for example). It's all about identifying "Are they using the same trap that they have been using before"? That is the gist of the strategy. If we have enough clues to overlap, we would be taking directional exposure to benefit from riggers' upcoming directional moves, assuming they will repeat same trick with same outcome.

An example of what I mean about segregating price action into zones (I do this every single day, on every single ticker) is using ticker FNGR, an older example from 2 years ago. The chart drawings are messy, probably one of my late-day fast recap works with a slightly autistic art style, but should be good enough to get some example going:

So the idea is as the price develops through the day, we split the tickers chart into zones/segments based on what makes sense to be bulked into each zone.

For example, the distribution structure could be one single zone. Type 4 structure could be another. Vertical breakout squeeze could be another zone.

And then we compare if similarities in the current latest zone match some of the prior ones. Ideally in replication, we want as many zones to match as possible, and more than 50% of them is a minimum. In fact, we want those numbers to be quite higher. The above example on FNGR goes quite high up to 80% perhaps.

Realtime method to identify replication

Let's use example of a WKEY ticker recently to break down how the process of identifying replication looks likein realtime. None ofthe insights are useful if trader has no idea how to apply the method realtime. Below is breakdown of WKEY ticker across multiple segments of the day as a trader would see it, chunk by chunk.

Use either GIF or seperated images below as study case. All separated images are from same ticker on the day WKEY just split per hours as how trader would see the chart develop as day went by and how realtime analysis would have been applied.

Full summary of WKEY and some realtime personal comments so that it doesnt stay in purely theoretical field:

  As per the attached chart, we can see they repeated type1/Type4 trap 3 times. Consolidation with lower highs swipe into a squeeze and fake breakout and then sending it all the way back down into demand clearout, ending with halt down.

Many longs on ticker like that will enter long into strength and the majority of shorts will be covering into HOD breach especially as halting right at that level adds extra heat. Heat is a real decision driver to cover often. This is why you have to think like a market maker to avoid being on the wrong side of the ticker and understand the game they are playing with you. It's all about targeted liquidations by someone who is potentially locking the float on the ticker. So they shake out as many longs before the squeeze as possible via those traps while locking up the float around the VPOC average.

Side note for WKEY: I thought also the catalyst was a decent one, leading to a good chance of some upside move, because DXYZ and TSLA are recently hot and the catalyst on WKEY plays into that via SpaceX semiconductors. But early on the catalyst by itself is never enough, bias should never be built heavily just on that. Once they started replicating behavior with same type of traps that was the real indication that indeed playing the long side might be a good idea. Hindsight shorts would have worked too well.

Split chart into zones and compare zones

As you can see it is mostly about zones, or 1. breaking chart down into segments and then 2. comparing the segments of the chart to see repeatable patterns.

This is why we should never think of chart of a ticker as one whole piece, instead think of it as series of larger and smaller patterns, sometimes clearer sometimes not at all. If we always segment chart into zones and try to find matches, at some point some ticker will gives us the clue for replications. This is especially useful on rigged - manipulated tickers. Which basically brings us to the most important point of why runnign this exercise of "tracking replications".

From my experience watching over the years hundreds of traders, rigged tickers tend to be most difficult to read and to trade for most traders, especially on the short side. Because the short side is less forgiving as rigged tickers will usually have a gradual uptrend especially if rigged over entire day.

To establish a read on the opponent (market maker) just like a solid player behind a poker table would want to establish a read on his main opponent sitting behind the table, all this is necessary to improve the chances of taking the right trades. This is where replications fit in. They can help aiding the read on manipulated tickers once every so often.

Two types of replications

-Intraday replication (WKEY and FNGR in article) where the ticker repeats its own behaviors multiple times, and the market makers who are controlling the ticker are for some reason repeating the same trick exactly the same.

-Day-to-day replication (EVTL/BTCS/DATS in article) where two or three tickers within the span of 2 or 3 days get replicated behaviors. Usually belonging to the same theme or sector.

Intraday replications such as WKEY are far better because it is easier to tell how they are re-using the flows, unlike in day-to-day ones where two different tickers are much more likely to have different volumes, different levels of price, and different details.

Replications between two tickers within same day

Replications do not apply just on same ticker, altough it is much prefered to focus on those. They as well happen on two different tickers (ticker B doing what ticker A did yesterday) if they are trading within short time span difference, such as 1-3 days apart (no more than 3).

Below we have such an example between tickers PHIO and DRMA from several months ago. I have highlighted only the structure after both tickers have squeezed and the microshelf was formed.

Notice how much similarities is between those two.

1. Same type4 of structure with pinbar poke under support before they squeezed it.

2. They squeeze both but its a fakeout that ends up failing before it breaks HOD.

3. They both sunk very fast.

   
   

   Notice all those are identical characteristics for each structure. This is where it gets us back to replication and "who is doing this"? It is likely same market maker, as chances of this to happen by coincidence within 2 days to such a same extent is not very high at all. When i say "same market maker" i should again point out that this includes team of market makers that orchestrate together. Either single entity, or multiple entities working as single unit. Its the same.

From a practical perspective, to extract edge out of replications you need to memorize charts well sometimes (if its not happening on same ticker). Details.

For example to take a short on DRMA above you would have to remember PHIO from the day before well what it did, so you could extract that fake rotation short trade on DRMA. If replication is happening on same ticker then of course we do not need any good memory as entire chart is always in front of us. For cross-ticker-replications however that does apply.

Again to highlight. Replications and patterns in my view are not the same, because there is a distinctive difference of time horizon. Pattern can happen at any point in time, spread out as far as possible. For replication we seek specifically tight time horizon, as we aim/assume the same market maker is behind the action. Meaning replication needs to happen within same day, or maximum 1-3 days apart.

How close to the same can it get?

One of the better older examples is from 2022 when we had the last legs of the crypto theme where 3 tickers within 2 days apart had perfect replications, identical charts (have personally traded each). Same sector, same charts, same time. Whats the chance that those arent being run by the same MMs?

We could split those charts into zones as shown on prior lesson above, and we could notice:

1. first zone matches on all 3 tickers with fake breakout,

2. so does 2nd zone with consolidation,

3. so does the 3rd HOD clearout zone with secondary fbo,

4. and so does the 4th zone with heavier unwind.

I have intentionally hidden away from the chart all key components such as: -price -name of ticker/stock For a reason, because if we strip that away, notice that its not easy to tell the difference between the charts anymore. Sure there are still plenty of details different on each, but on rough glance those charts are 3 different tickers within 2 days apart, driven by crypto sector at the time (EVTL/BTCS/DATS).

Rule of rarity

Replications do not happen often. Speaking about small-caps specifically, from my years of observations it's probably 1 in a 50 sample rule. In some months less (1 in 20) and in some months more (closer to 1 in 100). Making 50 is possibly a good average. Meaning that on random tickers (gappers or movers) in smallcaps, replication will happen on one ticker, and then for 20 or 30 or 40 tickers it will not happen. The frequency is largely dependant on market activity. In hot markets it goes down to 20 (good) and in cold market higher to 40 (bad).

Most short sellers will much rather prefer edges around easy fading tickers with no-brainer trading methods that require shorting something right at open. Bull traders who prefer to click and see something either go up 200% and make 10R or go down 20% and stop out. Simple no hassle methods that dont even require one to look at chart. Well...replications are as opposite to that as it gets. Replications are more for traders who like to watch price action behaviors from close-ups instead and are good at spotting similarities.

Conclusion: Apple and orange are kinda the same if you zoom out a lot. If you zoom very close, they arent the same at all.

Spotting replications is about using the concept of mirroring and comparing. And going way overboard to what many traders would assume is rational to expect from "similarity of movements". Some traders are skeptical that chart PA can really repeat nearly 80% the same on the same ticker, yet it does occasionally. However here is the catch:

There is a big echo of nothing for a long while....before replication happens again. Replications are spread out in the distribution of samples as:

In terms of how many tickers on a random day will actually go for clear replication of its price action. Few tickers per month and that might be in most cases. And then a whole bunch of nothing untill next time.

Trading replications requires a trader who has an eye constantly up seeking for them, else it's easy to miss it. It requires patience and tons of selectivity to not too eagerly match things together. Understanding how traps in markets work is essential. It wont be up for everyones cup of tea.

Personally, this is for me the cleanest edge one can get in the market, especially smallcap stocks or crypto which are most often rigged. The term I often like to use is "the closest you can get to cheating in trading is by trading replications". If you like trading manipulated action, if you are pattern seeking type of personality, and have a lot of patience to wait for a right opportunity even catching a few of those rare samples on a monthly basis can be significant even that frequency of occurrence is not high (as per personal long trades on WKEY):