This is not a quick read. I don’t apologise for that – it’s a really good post and it may teach you a lot about long-term BTC trends.
In a sea of crypto analysts, it’s important that you are able to differentiate the good from the bad from the ugly. The good and the ugly are fine (I’m both), but the bad are – by definition –
The New Year has heralded the dreaded return of price predictions, exacerbated in this case by the upcoming infamous Bitcoin halving. I tell you again, all the halving affects is short-to medium term price volatility, not long-term price! These two posts remain more relevant than ever:
I don’t bother myself with short-term Bitcoin price movements, I’m an investor, not a trader. What matters to me is not that Bitcoin has gone up (or down) today, but rather that it will have gone up a lot five years from now!
With this in mind, I focus more on long-term charting than on short-term charting. Long-term charting is more predictable, accurate – and to investors such as myself – useful.
If you are a trader, then by all means trade – but to me BTC is an investment asset. Traders help with liquidity and volume, so they do help the overall credibility of BTC a bit, but it is the investors who
BELIEVE in BTC and who are actively driving the adoption of crypto. Those who have the belief in BTC – enough belief to hold it as a long-term asset – they are the ones who are really driving the adoption of crypto and the
creation of a better world for us all. But enough about that, let’s get to the interesting stuff!
Round Pegs in Square Holes:
I often publish long-term BTC charts. On the majority of these charts you will find my “Long-term base trendline”, such as in this chart from the end of December post: “Bitcoin – Approaching 2020”.
But what is this line exactly? What does it mean? Is my line better than anyone else’s line?
Without the answers to these questions, my line is just another random line published by just another crypto analyst. It’s equal in status to any other similar line you may know of.
With the answers to these questions my line becomes a veritable tour de force – an answer to questions yet to be asked and a Sword of Damocles above the head of lesser analysts!
This line is thus very important to me, I stake not only my reputation on it, but also my own investment strategy.
Caveats: NEVER have I had a line on a chart which stays fixed indefinitely. It’s true that this particular line is about as fixed as they come, but it DOES move from time to time. So far the movements have always been very small – more “adjustments” than “movements” – call it “fine tuning”. Just note that the line is not 100% fixed and is consequently not 100% accurate. Also note that while it is valid 99% of the time, there ARE times that price can momentarily dip below it when BTC has serious negative momentum. Such dips never last, but they have occurred in the past, and will probably do so again in the future. That being said, I have great faith in this line and have spent so much time on it that I am absolutely satisfied with how accurately it is dialled in right now.
But not everybody would agree with my line – and they have strong reasons not to. This post shows you why, and then explains why I have chosen what I have chosen. As always – YOU must decide for yourself!
While I choose to use a straight line, there are those who prefer curved lines. They are in the majority.
[Pause for quick rant]:
When has “Majority” ever equated to “Right”? Anyone with an understanding of politics, the media or propaganda can quickly point out the logical flaws in using populism as a basis or substitute for correctness!
Don’t ignore that rant, it’s important. Now let’s look at curved lines:
There is considerable evidence to show that logarithmic curves can be made to fit the history of BTC price. You will often see charts such as this one:
On the face of it that’s a great looking chart. The indicator lines do what they should: they hit the tops of the peaks and the lows of the troughs. They are consistent curves which account for all prices during the time period, and they narrow together over time – accounting for reduced volatility as the market matures.
It all sounds so very credible – right?
It really does, so much so in fact, that I was quite impressed the first time I saw such curves used (about two years ago).
But taking something at face value is not a Bit Brain trait, especially if there are flaws in the logic or rationale behind it.
Look: the model is good. I am genuinely impressed by it and I think it does a pretty good job of accounting for BTC price at the moment. But I am concerned because I have found flaws with it and I don’t know how well it will hold up over time.
I DON’T KNOW the right answer here, because I can’t tell the future. I look at the data available to me – ALL the data – and then I try to make the best judgement calls I can based on that. These curved lines don’t account for all the data, which is where the problem come in. I KNOW that I’m not always right, but I have a higher chance of being right because I’m not looking for the simple answer. I’m not looking for that one curve that fits all the data. Similarly, I’m also not ignoring all the data that doesn’t fit, or making overly broad price brackets to try to account for all anomalies (here’s looking at you stock-to-flow!)
But Bit Brain, what’s wrong with these curved lines? What don’t they do?
I’m glad you asked…
They simply don’t fit all the observed data. Spot the problem on the chart below (which goes back to mid-2010 as opposed to the mid-2011 chart seen earlier):
It may not look like much, but it’s there at the beginning. The curve simply doesn’t fit all available data.
Can ANY curve fit the all data? As an experiment, I tried to construct my own curve to fit all the data, or at least the data with respect to the low end of the BTC prices over time (similar to the bottom curve on the chart seen above). Here is what I got:
That’s the best I could do.
THERE IS NO CURVED LINE THAT CAN ACCOUNT FOR ALL THE DATA IN THE HISTORY OF BTC
The line I drew is about as close as you can get to an all-data-encompassing-curved line for BTC base price over time. It’s no good because the BTC price is way below it near the beginning, but it’s still about the best you can do. This is how it looks next to the curved log lines that are commonly used:
“Houston, we’ve had a problem.”
Remember that there is another year and a half of (unavailable on my charts) data before this time (mid-2010) which only makes the problem far worse!
The commonly used logarithmic curves don’t fit. So what now? How bad is the problem? Is it really a problem? How will this play out in the future?
Once again I DON’T KNOW, but I DO KNOW some other things and I will use those to show you why I use what I do.
Summing up the Round Pegs:
It is my suspicion that the curved lines are working semi-well for a moment in time. The curves are not too far off reality, even though they can’t account for outlying information, so they look good. But I suspect that they will begin to degrade at increasing speed over time. Anomalies and/or a price trend shift away from the curves may well render them useless in times to come. I suspect that the curves will hold true for the next five years or so, but doubt that they will last for a decade. During this time (if I am right), they will become increasingly useless as prediction tools, because their lower and upper limits will begin to stray away from the lowest points of bear markets and the highest points of bull markets.
I believe that the curves fit a snapshot of time: for now that time is June 2011 to date. I suspect that they may have useful a shelf-life of approximately 15 years.
While we’re at it…
I mentioned stock-to-flow earlier. It’s not strictly related to these log curves, but Stock-to-flow is wrong in the same way that I say the curves above are wrong, but only more so – MUCH more! In addition to this, there are FUNDAMENTAL REASONS WHY STOCK-TO-FLOW IS WRONG! Those reasons are the same as other supply-sided arguments which I debunk in my “Bitcoin Halving” posts, as linked to at the beginning of this article. Go check them out.
Before I let Stock-to-flow off the hook and move on, let’s take a look at a chart or two:
Having trouble getting all the data to fit? Just make ridiculously broad price bands to encapsulate all the data! Ignore the fact that some data STILL doesn’t fit your model!
Or how about this, from PlanB themselves: Remember that EACH VERTICAL BLOCK represents a factor of 10x. So if an arrow is one block high, then PlanB’s predicted price is ten times too high or too low!
The largest of those arrows shows that the real price is about 14 times higher than the predicted one! I mean, come ON! Why would anyone use this? Remember my rant earlier about the difference between being popular and being right? Now you know why I say that…
Right, now here comes the tricky part: I’ve ripped the arguments of the majority apart, now I have to justify and explain my own standpoint. Here goes:
The key to my long-term trendline is this: I don’t try to fit all the data into one all-encompassing line.
I do not believe there is a way to fit all the data neatly into one indicator (short of using ridiculously wide price bands to account for all anomalies). I look at the markets and their history, I take the fundamentals into account, and then I try to project what is left in the form of usable indicators.
Below is another look at my long-term base trendline, this time seen in its entirety. Note where it begins: mid-2013.
I have often stated that I do not trust BTC data from the early years of BTC trading. I normally state that mid-2013 is the time that I start to find data sufficiently representative to provide useful statistical value.
The reasoning behind this is simple: Prior to 2013 Bitcoin simply wasn’t big enough, I’ve said this before but it bears repeating. There were not enough people invested in Bitcoin. They were not financial investors, they were not trying to to find fair price or make winning trades. They were tech geeks, cryptographic geeks, computer scientists and programmers who were playing with a new toy – taking a long shot that had little chance of going mainstream. The further you go back, the more true this becomes.
Any big buy had a radical effect on the price of BTC – which traded for less than $10 for most of that period and for less than $0.50 prior to 2011. You have to understand that even though logarithms do an excellent job of accounting for both small and large numbers in the same data-set, those numbers have to be consistent in order to provide good data without corrupting the model. Early BTC data is NOT consistent enough! It’s not something such as cell division which happens at a predictable rate and then to leads smooth, chartable exponential growth. Filling an equation with anomaly prone data taken from an overly-small data set is scientifically incorrect and will not yield a correct result! You have to have a sufficiently large sample size in order for the statistics to become valid. In the early days of BTC, the sample size was simply too small.
I’m really not comfortable with the way that logarithmic BTC price explanations try to incorporate all this old garbage data into their models. That’s why I ignore the early data entirely on my charts. From a fundamental perspective I haven’t ignored it, I’ve realised that it’s not proper data. I know why the data is bad and I know which data is bad. Thus my method incorporates all the given data, while remaining accurate by filtering out that which is not relevant – at least that’s the theory – a theory which is better than any mainstream ones I know of.
So that you don’t have to take me at face value, let me show you a chart or two which should help you make up your own mind on this topic:
The chart below expands on one of the charts we saw previously; the one with the “This is a problem!” section at the start of the log curve data. Here we can see that region in context, along with the regions which come after it (which I described earlier). The log curves are overlaid on the chart too.
My own model doesn’t use either of the first two regions.
Below I have removed the log lines and replaced them with an ascending channel, the base of which forms my long-term base trendline.
Starting to make sense now?
But wait, I’m not done yet…
I watch other major markets too. No market exists in isolation, they tend to influence one another. Some display similarities and correlations – we have such a case which effects BTC.
Early growth of BTC
BTC price shot up very rapidly in the early days – too rapidly to fit comfortably into accurate price models. Now: whether you believe me or not about the lack of BTC data in the early days making it impossible to accurately analyse and model that data, you will have to pay attention to the next point:
Bitcoin is often compared to Gold (whether a certain grouchy anti-BTC gold bug likes it or not). Both store value well, both are constantly climbing in value, albeit with some retracements and corrections when the fiat-based markets are booming. So here is an interesting little fact for you all: between the time of the Bitcoin Genesis Block and the start of 2013, Gold flew up in value. Gold went from approximately $830 at the time of the Bitcoin Genesis block, to over $1900 in August 2011, and then slowly dropped to around $1300 at the time I started trusting Bitcoin price data in mid 2013.
If market pressures were such that Gold got a huge boost, then such pressures may well have influenced Bitcoin too. Whether is was people looking for a store of value or just having money available with which to speculate is irrelevant; Gold went up very fast so it makes sense that Bitcoin should have done the same (much as Gold and Bitcoin have been doing since mid-December 2019).
The chart below illustrates this graphically:
It’s not just Gold – I used Gold because it’s easier for most people to see how it compares to BTC. You can look at the S&P, or the Dow or other large indices / commodities / etc. You’ll see the same thing.
Explanation / Summary
Here’s what I think happened: BTC was born after the market crash of 2008. It was born into the rapid recovery phase of that crash. Combined with its inherent early growth, this supercharged BTC in the early days – creating unrealistic and unsustainable growth rates for most of the first four years of its existence.
As can be seen from looking at Gold, the hype phase died away throughout the duration of 2012 and things went back to normal, which is why I only start to trust BTC after it settled down in 2013. It’s hard to say exactly when BTC data becomes trustworthy, there is no clearly defined line. I normally choose to use 1 July 2013 – the halfway mark. The reality is that it could have been a few months earlier, but not as early as 1 January 2013.
To put thing into perspective, BTC had a total market cap of roughly $150 mil on 1 January 2013. That grew to well over $1Bn in less than three months! At that stage BTC got noticed and became a serious investment asset. It then corrected downwards to below $1Bn by mid-year – which is when I start using the data – before starting to climb again. I attribute the climbing after June 2013 to organic BTC growth instead of to external market driving factors; which is why BTC then continued to climb – but at a different rate to what it had previously – while Gold continued to lose value and stagnate for the rest of the decade.
In basic terms: BTC only matured as an investment asset in early to mid-2013. Data before that time can be ignored – as long as you understand why it can be ignored!
Notes on Square Holes:
My straight line approach to the BTC base trend comes with a few caveats of its own. While my line is normally well labelled, it is commonly misunderstood. The Long-Term Base Trendline I show is NOT a price prediction! It is a BASE PRICE which BTC should remain ABOVE most of the time. Please bear that in mind if using my data or posts for information.
As I stated earlier, it can change over time. The changes will probably be hardly noticeable, but may occur. If BTC were to make a definitive drop below $7000 today and stay there, then my long-term trendline would have to either be significantly adjusted, or thrown away altogether. Here’s an interesting implication derived from that:
You won’t see Bitcoin selling for below $7000 again. Bit Brain has spoken.
(That excludes a whale tanking the price on individual exchanges, price display errors, Bitcoin being replaced by something better 10 years from now, etc)
I won’t go into the details because this post is long enough, but continued constant growth of BTC can also be largely attributed to the S-curve of technology adoption. I wrote about the S-curve in one of my earliest STEEM posts two years ago: check it out here “BITCOIN PRICE PREDICTIONS – Chart Display – Part 3 – The S curve (More good news!)”.
Footnotes and Conclusion
There is also a lot I can say about the nature of logarithms and using them in charting. I spent hours running common logs vs natural logs scenarios in preparing for this article – research which I am choosing not to include in the interests of keeping things on topic. But here are the sentient points:
- The BTC price curve MIGHT be a natural log function, but this would only apply to more recent data (mid-2013 and beyond) and not before.
- There is no log function that works for all BTC price data.
- Log functions straighten out over time. Using a straight line might be an approximation of an log function, we will only know this after a few more years once either the logs or the straight lines start to diverge from the data observed.
- This is what you get when you plot the natural logs of the base values of BTC on a chart .
- You can see that it’s not a straight line (as a correct log function that explains all BTC price action would be), but that using straight lines over specific periods might work.
- This ties in with what I said earlier about BTC normalising towards mid-2013.
- The x-axis is the number of days since the genesis block.
- My straight lines ARE actually log lines because they are only straight on Log (base 10) charts.
- No I’m not elaborating on that now – this is a blog, not a mathematics class!
If you’ve made it this far then well done – you may print this medal out and frame it as a reminder of your fortitude:
The truth is that right now there isn’t much difference between my straight line and the curved log lines, but over time this WILL change! In 2026 the commonly used base log line puts BTC crossing $500 000 while my trendline will puts BTC crossing $1 000 000 per coin. That’s a factor of two and is no longer negligible!
Continuing to put round pegs (log lines) into straight holes (straight trendlines) will work for a few more years, but if I am right about the holes being straight, then eventually those round pegs will no longer fit.
Yes I know I should have broken this post up. Once again I DIDN’T break up a long article because it loses it’s coherent logical flow and then you end up with two half stories instead on one whole one.
I hope that this article has achieved its purpose: to provide you with the UNDERSTANDING of why I use a straight trendline as my BTC base. I hope it shows you the difference between an analyst that develops and refines his own methods – questioning everything along the way, and those who use what is already available and merely try to fit lines and curves to what they see on the charts. I may not always be right, but I have a damn good understanding of the fundamentals what I do and how it works. When I am wrong, I make sure that I find out why and I take steps to correct that (and God knows I do make mistakes!)
So off you go now, the ball is in your court. Believe quantity or believe quality. No – you can’t take the middle ground – that’s just being wrong from both perspectives! I hoped you enjoyed reading this post as much as what I enjoyed writing it.
See you again soon.
Yours in crypto
All charts made by Bit Brain with TradingView
“The secret to success: find out where people are going and get there first”
~ Mark Twain
“Crypto does not require institutional investment to succeed; institutions require crypto investments to remain successful”
~ Bit Brain