$btc - htf data analysis
There is a 98.4% chance bitcoin does not go below 50k on a volatility adjusted basis.
In the light of our last post on #bitcoin's infamous electricity cost metric, where we called the bottom once again before a whopping 37% move all the way to 83k from the very bottom of 60k, called out live, all due to one of our most important signals passing by, I decided to go deeper into the analysis.
With all the random numbers thrown around, vague calls and loud celebrations of how the bears "called" this entire move proudly, and with that same conviction, expecting 50k and below, I think what people need the most right now, is at least one solid metric + data shared, describing how that happening, is a highly unlikely chance.
Good data and strong data in general is hard to dispute, but I still give the kind disclaimer that this is just my lens applied to that strong data. There are multiple ways to interpret data. With this one though, no matter the lens, interpretations are quite narrow and I think that's the very way to approach data analysis in trading.
I always find it quite funny when someone posts a chart of 3 data points, then concludes that the 4th one is a guarantee, whilst anyone who followed high-school statistics, knows otherwise, how 3 times 100% chance, doesn't mean 4th time, in a probabilistic world.
So with this post, I like to offer strong data, as well as explaining the logic behind the data (to remove the black box data-only effect), of why I am so confident we don't go below 50k.
Thank you in advance for this more extensive read. I am sure you will enjoy and some of you may feel some nostalgia every time I share a post like this given my historic reputation on these.
Without any time wasting further, let's get to it.
The logic
This one is about the miners electricity cost to produce 1 $btc. This is a vital metric. Now I know there is a lot of controverse around miners and their impact, but there is still an inflation of 0.84 per year on $btc to date since the last halving (about 164,000 BTC per year). Seems negligible but at the current price of Bitcoin (61k), that still equates to 10.7 billion dollars per year. So every year, 1/6th of the entire supply of @MicroStrategy 's entire holdings gets released into the hands of the miners, and with $btc's thin liquidity existing to this date, you wouldn't want to see that dumped on the market, certainly not every year. So yes, the miners still have a very important impact that can't be underestimated.
Put differently, that equates to @MicroStrategy's entire holdings being sold every 6 years (1.5 cycles long). If that doesn't put a different swing on the significance of Saylor's actual influence on the market, I don't know what will. And I believe I have convinced you now how impactful and in control the miners still are today. (In fact, I don't need to convince you, the production cost floor speaks for itself, still until today.)
So you don't want the miners to sell (which they mostly do, slowly, to keep their business running). But due to the current situation, they can't do that anymore, because the market has hit rare conditions, only happening a few times every cycle.
That is, the price has dropped below the average weighted electricity cost to produce one bitcoin:native per kWh.
Significant? Maybe. Let's put some logic behind it: Not only does that mean that the miners can't sell their $btc for a profit. It also means that it is simply cheaper to just log into a CEX (large funds: OTC) and buy 1 Bitcoin, instead of going through the pain of mining 1 Bitcoin. So not only does this make the miners (the people controlling $btc) not want to sell, it also makes them want to buy, because it is cheaper to just buy instead of mine them. And although I am not saying that is what they do, it is a large pressure and narrative on the market, which has driven price north without any deeper revisiting each and every time in history.
That is the logic behind why this works. Should we believe it blindly? Never. Successful trading and analysis is always a combination of data + logic and cross verifying two., never of just one or the other. But let's call it an assumption (assumption 1).
I could not write on one slate the amount of charts and videos and posts I see on X every day, only covering aspects of just one of the two, in mere lazy manner too, just throwing numbers around, or using complex risk metrics or equations without any logic behind it. It hurts the seeing eye.
All power to them however. Many are learning, many are adapting and many don't even trade, they just DCA and draw some charts telling everyone how they are "mostly right".
Short rant aside, logic by itself is strong and often missing, but we need data to verify logic (hypothesis) correctly.
Collecting data
How to do this? It's very simple. To compare how price compares against the continuous band of production/electricity cost, all it takes is simply mapping it out on a price-time chart on tradingview, which is the purple band represented below, starting with the production cost and the electricity cost as the floor. The production cost is higher due to mining equipment, and that cost also varies since mining rigs are tuned to performance, therefore cost, that is why a wide band appears. The electricity cost is the floor because that is disregarding capex into mining equipment, and electricity cost per kWh (worlds average) doesn't vary much over time.
What we also notice is how the elec/prod cost rises over time, due to two drivers:
➡️The halving (every halving, it becomes more difficult to mine 1 BTC, giving a large jump)
➡️ General competition (adoption driven, more miners = more competition for blocks).
Both feed the eternal adoption cycle of bitcoin and rising floor price (unless abandonment, the opposite of adoption happens, let's hope not, but there are clear signs it's not happening).
So, mapping out the elec/prod cost and simply comparing how far price bottom above or below each time it visited, gives us a statistical reference to where price will bottom now (or where it is unlikely to go now).
It is indeed a mere statistic, because volatility is somewhat statistically driven, intertwined with cycles.
One key note: volatility adjustment is important.
On that note, collecting how far the wick goes below the band each time in absolute sense, is not sensible enough. It is to its simplicity elegant, but price also needs to be adjusted for volatility because for example daily 40% up and down moves Today are far less likely than back when $btc was priced 1$ per coin. Anyone who ever traded microcaps or penny stocks, knows what I mean. So since we are using the entire population as back test data, we must adjust price for volatility.
How to do this? By price law books, the relation depends on liquidity (how thin it is), the operators controlling the markets and the time in the year, day, week. But in general terms, liquidity thickness is linearly proportional to volatility and volatility scales inversely with the fourth root of price.
What does the latter mean? If price doubles, it means volatility decreases with the fourth root of 2, which is 1.1892...
So when price wicks below the band in say 2015 for 20%, that means today, when adjusted for volatility, that difference should be 61k/whatever the price was in 2015. E.g. $61000/$250 = 244. Fourth root of 244: 3.95. Which means the 20% wick should be accounted for as a 5.06% wick in the data.
Keep in mind, this is a relatively rough assumption (assumption 2), but one backed by price-liquidity-volatility laws.
So throughout the entire history, we collect these data points of how far below or above the wick went relative to the electricity cost at that time, and compare that to the chances of reaching 50k now, by comparing how much further price has to go below the current low, which is 61.1k, conveniently aligning with the exact electricity cost of 1 $btc today.
Using 61.1k as the in-real-time of writing this post, that puts 50k: about -18% below that.
That sums up how to collect the data.
With assumptions again renamed below...
➡️ Assumption 1: the logic of miners' impact
➡️ Assumption 2: volatility decreases with the fourth root of price (market cap).
➡️ Assumption 3: normal distribution of random volatility differences around a given price point...
... we are ready to collect the data.
Data Analysis
Next, let's look at the data, let's look at the history, where I will be taking every single data point which has reached inside the production cost band as a high timeframe bottom data-point. Because frankly, as it speaks for itself, it has been a high timeframe bottom every single time.
Below, are all the data points, sorted by date (Monday starting the weekly candle), % wicked below (-) or above (+) the lower edge of the band, and its normalized %, normalized by the square root of volatility (assumption 2).
Date │ % wick (-) or (+) lower band │ Normalized %
➡️12 Jan 2015 │ -12.46% │ -2.83%
➡️17 Aug 2015 │ -26.41% │ -5.99%
➡️1 Aug 2016 │ +1.54% │ +0.45%
➡️9 Jan 2017 │ - 12.44% │ -4.30%
➡️20 Mar 2017 │ - 7.21% │ -2.56%
➡️10 April 2017 │ -11.26% │ -4.42%
➡️10 Dec 2018 │ -26.80% │ - 13.81%
➡️9 Mar 2020 │-26.45% │ -14.38%
➡️9 Sept 2020 │ - 9.22% │ -5.98%
➡️ 7 Nov 2022 │ -0.67% │ -0.47%
➡️10 Dec 2024 │ +8.47% │ +7.74%
➡️2 Feb 2026 │-5.45% │ -5.44%
Using the volatility-liquidity adjusted %'s into a mean and assuming they are normally distributed, which, in argument with a Poisson distribution, is acceptable. Both distributions lead to similar results, but a normal distribution is more lenient towards random events revolving around a centreline (here, the bottom line of the production cost band), hence my choice.
The mean is -4.33%. The sample standard deviation is 5.99%. However, we chose every single low so we opt for the population sdev, since we do indeed have a sample of the entire population. This sdev is 5.74%.
Within this population, the z-score of 17.16%, which is the excursion needed from the current low of 61.1k (which also aligns with the perfect bottom of the band), to reach 50k, is -2.14. This equates by law of statistics: to 1.6% chance of reaching 50k, a low chance.
What if we use the non-price adjusted volatility %'s?
Then the mean is −10.70% and the sdev is 11.69%. In this case, a -17.16% lower excursion from the current low of 61.1k, to 50k, has a z-score of -0.55. This aligns with 29%.
Conclusion
The chances of never reaching 50k or below are 71% when not adjusting for price-volatility and only 98.4% when adjusting for price-volatility. Let's be realistic, and choose the exact middle between both chances, which is 84.7%. Still a very high chance, more than enough to look for aggressive involvement.
So personally, regardless of whether my assumptions are correct (98.4% chance of no 50k), or are not (71% chance). I personally believe expecting lower than 50k is hopeful and wishful thinking to its peak. And this valley is just a mere opportunity for the bears to be loud and proud again, before absolutely missing the chance of lifetime opportune buying prices once again.
We take a look at the timeline, we take a look who is clearly and loudly bearish, who is loudly bullish, we mark them on the chart, and realize when extrapolated to the entire world, both are majority disfavouring proper data, in my humble view.
We look some months down the line, and see where we will be, and whether talking generally bullishly, or generally bearishly was the smartest move to flourish in the world of crypto finance.
