Breaking Bad : 21-W EMA
Updated: Jun 16
Does this Indicator really decide Bull / Bear for ฿itcoin?
What happens after Bitcoin (BTC) breaks below the 21-Weekly Exponential Moving Average (21-W EMA)? This event recently occurred! This sign has become famous for indicating Bull and Bear markets because it always worked in the past. But in a world where the Composite Man knows everyone is watching an indicator, will it be as reliable a discriminator? Are we de facto in a Bear Market now that price has not recovered to this indicator?
This article is a deep dive into the 21-Weekly EMA, including its fundamentals and how it has played out for Bitcoin. There are few aspects that make these principles important for our times. Recalling that Wyckoff taught us the importance of learning from past behavior, we will review each of the significant events when Bitcoin's price played with the 21-W EMA. Don't worry - there are only a few cases to look at.
In these case studies, you'll see amazing repetitions: patterns or fractals. Understanding how the patterns realize - or are violated - will allow you to step in or out of positions. You will then be able to use this to manage your coins and risk/reward appetite in the coming weeks.
For all of Bitcoin's history, there's been a few secret TA Rules. Knowing what do with the Golden Cross, Death Cross, PPO Crosses, and the 21-W EMA are the basics anyone needed to know to make basic trades in Bitcoin. Here's one basic rule:
21-W EMA Logic for Bull Runs
IF BTC price bounced off OR rebounded above the 21-W EMA, THEN Bitcoin would continue to be in a Bull Market, ELSE
If BTC price closed with one (1) ENTIRE weekly candle below this indicator, THEN BTC would go into a Bear Market.
From the figure above, one can see the 100% reliability of this process during previous runs. In fact, it's so well know that many crypto gurus released videos explaining "When to Sell Bitcoin" using this rule. However, during the previous two Bull Runs, this was not a well known fact. The cryptoverse was small, and such rules of trading were being formed. But today, this rule is almost thought of as a Law!
Recently, we explored two actual Laws of investing. Remembering Wyckoff's Two Laws, we are reminded that price action never follows the same course as it did in the past and that makes the market never behave the same way. Should we expect this rule to continue to indicate the fate of Bitcoin?
If one used this rule alone, then they should sell and walk away. Perhaps not today, but soon. Then again, we can also remember the analytical importance of comparing today's price action to past behavior. Are there reasons why this rule might no longer be in effect?
What are Moving Averages?
To begin with, an exponential moving average (EMA) and the simple moving average (SMA) are both technical indicators that use past data to generate a smooth trend line for the price of an asset by taking one type of average of prices across a time frame. The effective difference between the two moving averages is that EMA places a greater weight on recent prices, while SMA places equal weight on all data points. This is why the EMA line turns more quickly than the SMA line. Meanwhile, the weighting of the EMA is more short term, and so zooming out to a weekly chart gives a wider view.
How are time frames chosen?
What is so special about "21"? After all, there are 3-, 5-, 10-, 20, 21, 30-, 50-, 51-, 100-, 200-, 250- and even 600-averages. Why use one over another? The truth is there is no universal answer. There are useful conventions though. For instance:
US market time frames: The moving averages were meant to represent special time frames, where five trading days meant a single stock market week, ten made up two weeks, fifty weeks was about a year, etc. This resulted in the standard set for Wall Street became: 10, 20, 50, 100 and 200. A cross between the 50-D MA and 200-D MA then became the Golden / Death Cross. Note, these terms only apply to these number on daily time frames. This is simply a Convention.
200-MA Buy/Sell Rule: Boomer's learned a 200-MA Rule, whereby "Long-term investors" in DOW, S&P, or particular stocks use the 200-D MA (40-Week). They were taught to simply dollar cost average, buy and hold if their asset was above the MA, and switch to bonds when price was below. As it turns out the choice of a 200-D MA is arbitrary and quite sub-optimal in the days when STONKS ONLY GO UP. The old rule doesn't work anymore.
Short time frames: Day Traders and Scalpers learned to choose an MA/EMA for their own time frame. Specifically, they would focus on the average time they wished to hold a trade for scalping or trading.
Japanese market time frames: While Wyckoff had released his trading strategies for American assets, a trading sensei on the other side of the world was making parallel strides. Goichi Hosoda developed the concept of Ichimoku Cloud in the1930s for the Japanese markets. The schedule was different in Japan because Saturdays were normal business / trading days. As such, the timeline horizons were 9, 26 & 52. Nine represents a week and a half, while the numbers 26 and 52 represent one and two months, respectively. However, for 24/7 markets - like cryptocurrency - many traders adjust the Ichimoku settings to reflect the 24/7 markets, changing (9, 26, 52) to (10, 30, 60). To get a longer term chart, traders adjust the settings to (20, 60, 120).
Fibonacci Levels as Time Periods: Traders who favor Fibonacci tend to choose a modified set of time frame periods. They use 5-, 13-, 21-, 34-, 55-, 89-, 144-, & 233- time frames. Since most of these numbers are slightly larger than the US market time frame numbers, Simple Moving Averages are usually replaced with Exponential Moving Averages. Some of these work out nicely for cryptocurrency with its 24/7 activity. For instance, 21-days is three weeks. Twenty one weeks is 147 days or about 40% of a year. But for others, the numbers are relatively close to their MA counterparts: 10 vs 13, 50 vs 55, 100 vs 89, & 200 vs 233. Therefore, the two that usually get the most attention are the 21- and 34- time frames, as they are relatively unique. Finally, we have a reason for "21" besides blackjack!
In other words, each interval number had a place in history that defined it and still their choice is rather arbitrary. What matters most for any indicator is whether the market respects it. Do traders, bots, and the Composite Man use it to signal Resistance and Support? And yet Fibonacci became dominant within the cryptoverse very early. Many major retracements and extension follow Fibonacci patterns. In fact, Trading View has at least ten Fibonacci tools! Is it any wonder that 21 became such a respected number in Bitcoin and cryptocurrency?
Reflecting on Bitcoin today
Is there a case to be made for the 21-W EMA Logic for Bull Runs being irrelevant today? Consider that this indicator is time-based. Above I noted three examples where time frames were adjusted because of timing: US markets schedule, Japanese market schedule, and 24-hour trading schedule. Therefore, an adjustment in time will affect its relevance. Meanwhile, the current uptrend was faster than previous ones. Bitcoin left the 21-W EMA last September and did not touch it since, spending a record time away from the indicator. During this run the MA was outrun and never acted as support! Therefore, it's possible that breaking it recently and not recovering does not mean a Bear Market.
Old Wall Rules
But do markets break rules and survive? We've already discussed one outdated investing rule: the 200-MA Buy/Sell Rule. Here are a few other investing paradigms that seem outdated today:
Financial News is a great place to get stock tips.
Moving Averages tell you when to buy/sell.
60/40 Stocks & Bond split.
Subtract your age from 100 to determine the percentage of your portfolio that should be in stocks vs bonds.
You can’t hold a real estate investment in an IRA.
The Buffett Indicator - the ratio of total US stock market valuation to GDP - is the best way to pick stocks.
A Shiller PE (CAPE) Ratio above 30 is a bubble.
Interest rates cannot go negative.
Companies that do not make profit cannot survive.
Both unemployment and inflation metrics have changed in their methods so much that they cannot be used to compare against previous crisis events.
Governments cannot print money to their hearts content.
As markets evolve, old and established rules are thrown out and become irrelevant. Perhaps the 21-W EMA Rule is being challenged today.
First Bull Run
Instead of focusing on the Rule itself, let's explore how Bitcoin has acted after it interacted with the 21-W EMA. During the first Bull Run, this EMA acted as fast support at least six times. In this next section, we will cover the two instances that were not fast bounces: the successful 2013 mid-run test and failed the 2014 post top test.
The first thing to note for this run is we see a standard Wyckoff mark p with a throw-over defining the top during a blow off top distribution phase. There are ten two upward re-tests and two major distribution events before mark down for re-accumulation. The re-accumulation period lasted about three months, followed by a ~10X mark up.
The markdown began as the third downward leg of the distribution phase. This move was an a-b-c corrective wave with massive selling volume. The Selling Climax (SC) was in early July, and the Automatic Rally (AR) retraced over 80%, but settled around the Golden Ratio (61.8%). The AR was so aggressive, it rallied past the PS level! This allowed price get above the 21-W EMA just in time, but it was short-lived. The short-term uptrend ended nine days later, and a Secondary Test retraced the AR to the Golden Ratio (61.8%) in three days. The Composite Man bought up both of these sell-offs, as price again retook the 21-W EMA - just in time. After these tests, price action became dull, with a slow grind upwards and a crazy sell-off, fake out that acted as a Spring in late September.
Looking into the RSI allows for a view of buyer behavior. From the peak price, notice lower lows and lower highs in momentum, meeting a firm resistance line. RSI makes its ultimate low at the SC event. The fast AR event breaks the RSI resistance and finds 45 as support. Notice that AR tops when RSI is between 55 and 65. Even through there is a sell off, the higher lows find support and momentum slowly grows.
BTC spent about six months trading in this re-accumulation event. From April to October, BTC traded around $130. This the Bull Run mid-cycle Consolidation Zone.
It took about 200 days in 2013 for BTC to retake its previous high of $259. Technically, this was not a Bear Market because price only spent a few days below the 200-MA and a Death Cross was narrowly averted. However, for investors at the time the double bottom, near 80% crash felt like the end. And yet, in July BTC began to use the 50-MA as support, while the Composite Man continued its re-accumulation after decline campaign. A one day crash in October served as a spring, and prices rocketed off. It took only eight weeks to mark up BTC's price by 10X. As price growth accelerated, throw over occurred triggered the Composite Man's Distribution phase.
After initially finding support on the 21-W EMA in mid-December, a loss of momentum and a failure to make new highs in early Jan led to an a-b-c correction. SC occurred with a 65% draw-down from the peak of $1163. The Automatic Rally took six days and retraced 75% the b-c move. The AR brought price back above the 21-W EMA, which seemed to act as support for the next days. However, the following weeks took price below the 21-W EMA and the Second Test didn't find support above SC.
The daily RSI gives a more detailed diagnostic of the end of the Bull Run. Notice the peak in momentum occurred with the blow off top, while the following three tests for a new high each found lower highs in RSI. A resistance line formed in RSI that wasn't broken until the SC. The AR broke the RSI downtrend, yet BTC was sold off around the 55 RSI level. Notice the makings of a pattern? Then the Second Test made a lower low in RSI. This signals to professional traders and the Composite Man that the campaign is coming to and end. Subsequently, the fast rally failed to re-take the 21-W EMA and made a double-top in RSI.
The following weeks took price below the 21-W EMA and the Second Test didn't find support above SC. Price continued to make lower highs and lows. About 100-days later, a Bull Trap occurred before price capitulated over 400+ days, down to 87% from its peak, to $152.
Notice this low is almost the same as the high from the 2013 mid-cycle. The Bull Run's mid-cycle peak served as support during the next Bear Market. This is another pattern that played out in the aftermath of the second Bull Run too.
Second Bull Run
During the second Bull Run, the 21-W EMA was tested repeatedly and acted as fast support at least seven times. In this next section, we will cover the two instances that were not quick bounces: the deep and penetrating, yet successful 2016 mid-run test and failed the 2018 post top test.
2016-2017 Mid-Run Tests
As price appreciated up to the previous all time high, it operated within a channel with several, quick blow off tops, mark downs, and re-accumulation zones. Each mark up phase was fast, making higher highs and higher lows in both price and RSI. Each peak typically had three attempts to make higher highs in RSI before mark down began. Sell-offs were fast and severe, typically 30 to 45% over a few days. And yet, each time the 21-W EMA served as strong support. Price only closed two days under this indicator! Blow off tops were then marked down, as RSI made lower highs.
It's quite easy to see the trends in RSI, with >80 in the daily being a clear sell signal. Recall the pattern we have observed in RSI finding resistance around 55? During this run up, each climb from a low in RSI also was experienced a sell off between 55 and 65. Notice that unlike a peak, RSI makes higher lows and then makes higher highs. There's that pattern again!
Notice also that above the previous ATH, BTC spends many months trading side ways. From May to September, BTC continued to trade around $2700. There were two 40+% corrections over three months, making this the Bull Run mid-cycle Consolidation Zone.
2018 Post Top Test
After BTC's price broke the previous ATH, it entered a different, accelerated price structure. Then, it interacted with the 21-W EMA only twice before its peak. As before, throw over triggered distribution and a triple test for higher highs. Bearish divergence in RSI lead to a quick, 50% mark down from the blow off top peak over 30 days. At this point, price found support on the 21-W EMA (image below, point 'a'). In each of the previous peaks, the sell-off ended at the 21-W EMA. Yet in this case, the RSI had still not gotten down to the highly over-sold level. A quick bounce and follow through of the a-b-c- correction post-throw over lowered price to an SC well below the 21-W EMA. Even while there was a fast bouncing automatic rally, the weekly candle closed with price well below the 21-W EMA for the first time in over 800 days. The automatic rally overtook PS and retraced 92% of b-c. In fact, it to attempted to rally further.
Again, the daily RSI level of 55 served as a major sell signal. Unlike 2014 peak, the second test made a higher low. And yet, because the 21-W EMA was so close to price, the second test led to price again closing below the 21-W EMA. After which, the rally off the ST made a lower high while RSI again found major resistance around 55. Then, RSI failed to make a higher low after the SC. Again, this tells the Composite Man and professional traders that the campaign is ending. Lower lows and lower highs continued for another 270 days, taking price down 85% from its peak.
Capitulation found support at around $3100. This level was extremely significant during the mark up. It had served as a resistance level that resulted in a mid-cycle, 40% correction, 50+ day (June - July) crash. It also served as support during the following 40% correction (September). This yields yet another pattern to recognize. A major resistance level during the accelerated, post-ATH bull run serves as support during the following bear market. So far, every Bull Run required a major crash to help establish support levels for the Bear Market that occurs afterwards.
The 2019, three month rally was no slouch. With 5X returns, a lot of people were introduced to Bitcoin over this period. However, the following "Covid Crash" down to $3.5k makes this move seem more like a Bull Trap. And yet, it shares all the same characteristics of the previous runs.
A fast mark-up period ended with throw over and the triggering of a distribution event. Like the accelerated bull runs, price did not touch the 21-W EMA during the quick mark up. After falling below the trend channel, it was tested from below This again shows the Composite Man's work, because we again see three tests for higher prices after a blow off top and bearish divergence on the RSI. Distribution was triggered, and yet again an a-b-c- correction after the throw over. During the distribution event and mark-down, the 21-W EMA isn't even respected at all. It's almost as though it was irrelevant.
Unlike during previous runs, the SC response is weak. There didn't seem to be an Automatic Rally. Twenty-seven days later, a lower low is created and what seems to be a short-squeeze finally resulted in a rally (R) that successfully retraces about 80% of the b-c move. Again, we find another pattern! After every single SC is arrived at, a significant rally occurs, retracing b-c! Notice also the 55-65 RSI level served as a sell trigger. Finally, a failure to make a higher low in RSI broadcasts the campaign is ending. There's still money to be made by day traders, but lower highs and lower lows ensue. The peak of R served as major resistance, and the bullish sentiment of many market participants in Feb. 2020 led to what seemed like a rug pull as market plummeted. And yet, simply knowing these patterns could have allowed an investor to wait to scoop up bargains in March. After all, Bitcoin's Bull Runs had not ever begun without a double test of the lows - yet another pattern.
2021 - Today
After finding the 21-W EMA as resistance in late April 2020, BTC only touched this indicator twice during this upward swing. Both of these touches were in September, while the first signs of strength were amassing. During the mark up and distribution phases, the 21-W EMA was never approached and was again irreverent.
During the mid-May mark D=down event, the indicator acted as support on two days. There seemed to be massive buying at PS - which coincided with this indicator. Meanwhile, breaking below this level triggered market-side capitulation, bringing BTC down 50% from it's recent peak. Price has remained below this indicator for almost two weeks, while sentiment has turned quite Bearish.
The Composite Man seemed to step in around $30k, and the fastest AR of yet was achieved. A single day, 45% retracement of b-c However, this is quite uncharacteristic. It was much faster than normal, and with very little follow through. Higher lows have been reached, but lower highs are not being made. Moreover, daily RSI had not yet reached the 55-65 level, as it keeps finding resistance on a downward line. In other words, this rally bears little resemblance to the previous ones and can hardly be called over. Therefore, let us treat this level as a preliminary AR, with the recent retests as a preliminary ST. Meanwhile, the 21-W EMA inches lower and lower while price compresses in a symmetric wedge.
While many believe the 21-W EMA is the indicator to use when making their Bitcoin investing decisions, this seems to simply be due to their study of previous Bull Runs. In truth, there is nothing magical about this indicator. That said, there's nothing Bullish about Bitcoin being below this indicator. And yet we have seen that since the 2018 Bear market, this moving average has held little significance. Perhaps the Composite Man knows all eyes are on it and is using that information against retail traders.
On the other hand, there are many other, lesser known patterns that repeat. Here is a list I've covered within this article.
Throw-over triggers Distribution
Distribution events have three (3) tests for higher highs
The third test is a back-test of the lower-bound, up-trend line
Distribution ends with an a-b-c Correction
SC events have massive volume and long wicks, ending the down-trend
Retracement off SC are typically 50 to 78% of b-c. If that occurred soon, BTC would rally to $45-53k
The AR occurs with daily RSI around 55 and takes many days to realize
Price up swings from low RSI often find resistance and are sold off when RSI ranges from 55 to 65, with 55 being the norm after SC events
Bear Markets occurred when:
The ST broke the SC
The RSI at the ST made a lower low
The RSI after the ST failed to make a higher high
Mid-Bull Run corrections occur at Resistance levels that act as support in subsequent Bear Markets. Assuming BTC makes new all time highs, the low of the future Bear Market should be above $60k, but might be as low as $40k.
Bull Runs have mid-cycle Consolidation Zones that take months and have significant draw down, where in many investors believe the Mark Up Campaign to be over and a Bear Market is imminent
The two Bull Run peaked and had a Bear Market bottom of 78 and 87% off their peak. If this run's peak is behind us and at $61,327, then the bear market lows would be expected to be $8k to 14k
Bear Markets have always had two tests of the lows before Mark Up began in earnest
But what do you think? Has the 21-W EMA Have you found some patterns that I've left off? When moon?