TL;DR - We may be finally overcoming the $60k. This week I update the model and describe some historical price action as an analogy to the current price action. Pi-cycle Top has will be invalidated if the price does not crash in the next few days… when and how will the Bitcoin Storm Tracker model be invalidated?
The model bounds have formed an “L” shape, indicating a strong floor around $50k, but lots of potential for upside.
The statistics are not sure when the next move up will be.
Current price action is comparable to July 2017 historical data. While not predictive, it motivates the hypothesis that strong HODLers may be finalizing their accumulation.
I will consider a fitted model “invalidated” if the daily close falls under the 95% confidence bound for 2 days in a row. The oldest model that is still valid is from March 13th but will be invalidated if we don’t move well past $60k this week.
“All this has happened before, and it will all happen again...”
— Peter Pan, 1953, Battlestar Gallactica, 2004-2009
I would like to remind readers this post is a purely educational exercise. I’m fitting models based on data and am probably wrong. Don’t construe this analysis as investment advice, I am not an investment advisor and you should seek one out before making decisions with your money.
The Forecast
The main change to the model this week has been an expansion of the uncertainty bounds. The upper and lower bounds have begun to form an “L” shape as the lower bound predicts a floor of around ~$50k and dropping slightly, while the upper bound has now become nearly vertical. If the price takes off, it appears the next local correction may occur around $80k-$90k sometime around the end of April or into May.
The uncertainty bounds are important, and while frustrating for some, expanded bounds describe how uncertain the price action made the model. It’s hard to predict when exactly the current consolidation will end, but there is reasonable confidence there will not be a daily close below ~$45k within the next month, but with much more potential for upside.
Analysis: Recap
The price action this week was within that expected by last week’s model. As discussed above, the current model has expanded uncertainty bounds, forming a rather flat lower bound for the foreseeable future.
Critical Date
As we may be continuing up, the critical date has started to creep closer in time again. What we look for in this plot is for the critical date to stabilize, indicating the model has “locked on” to the super-exponential growth. Since the critical date has started to come down (to an earlier date), we can interpret this as potential re-heating of the market. We must wait to see if it will return to the June-July 2021 timeframe.
Philosophy: To-date Model Fit vs. 2017 Hindsight
It’s important to understand the impact of “microbubbles” on the LPPL model fit as more price data comes in. Let’s take a look at how the model fit would look if we use the 2017 bubble price data, where we have the luxury of hindsight. The following illustration shows a visualization of these fits. I’ve hand-drawn the LPPL model fits here for illustrative purposes, but you can cross-reference refer this paper to get exact values.
The key takeaway is that if you use the “to-date” data for the first half of the bubble and fit the LPPL model (bottom plot), the model makes an overly optimistic prediction because it tries to fit to the “microbubbles” in each pump upwards. If we have the luxury of hindsight (top and bottom plots), the LPPL model does not predict as aggressive price action for the first half of the bull run.
This might explain why the predictions made back in February did not anticipate we would stay around the $60k level for so long.
What this means is that if the current position in the bull cycle is similar to where it was in July 2017, the “to-date” fits that I’ve been sharing may be overly optimistic. As more price data comes in, the fit will improve. I’ll call a fit invalidated if the daily price is below the 95% confidence bounds for two days in a row. After improving these confidence bounds at the beginning of March, the March 13th forecast is just about to be invalidated. Of course, looking at this week’s fit above will provide the best approximation of these bounds based on all of the available price data.
With that in mind, the current fit’s lower bound remains relatively constant and actually trends slightly downward at around or just below $50k for at least the next month. The upper bound of course predicts more upside, and the asymmetry (“L” shape) of these bounds is bullish. Let’s see what happens!
Philosophy: All of this has happened before…
Stepping away from the quantitative analysis, I also like to use this newsletter to talk about regular trends and patterns that do not fit well within the existing statistical model. In previous weeks I’ve described how there appears to be a lower bound for the price, which can be explained by adoption and Metcalfe’s law. Can we see more clues to where adoption is happening by looking at historical patterns in the price?
Essentially the idea is that there is a “Strong Resistance/Support” level for each cycle. Each cycle can be broken down into three phases:
Test new cycle resistance ~3x the previous cycle’s ATH
Break the resistance and undergo a super-exponential growth up to the new ATH for the cycle
Crash and reach support at ~3.5x the previous cycle’s ATH
If this pattern repeats itself, we might just now be testing this resistance. Once it’s broken that might indicate we enter the super-exponential FOMO stage over the next few months. When this occurs, the LPPL model will be essential to understand the probability of the cycle top. Regardless, this points to a cycle bottom of ~$60k-$70k. Assuming, of course, this pattern repeats. This has narrowed my original cycle bottom range prediction of $40k-$100k.
I’ve been calling this my “Illuminati Confirmed” chart. I ran out of red-string and newspaper clippings, but the above will have to suffice. I’ve compared the current cycle to 2017 before, and exact matching is dubious, but as the saying goes:
“History does not repeat itself, but it often rhymes.”
— Mark Twain, maybe1
Speaking from my own experience, I was first made aware of Bitcoin after it “crashed” from $30. Like most people, I thought it was dying and/or highly speculative. I didn’t follow it that closely, but I watched as people lost money to the Mt. Gox fiasco, and saw the price in the ~$200 range for what felt like years. I did not buy any until 2017 after it broke the resistance level plotted above in 2017. At that point, I believe I was converted into a “Strong HODLer” and those lucky enough to do so hold onto their coins even during a crash, creating an even stronger support level. It took me 2 cycles to “get it” and I’m extremely curious to see if this pattern repeats for future cycles.
Please think of this as a first step to trying to come up with a hypothesis for the next cycle’s bottom. It would certainly be exciting if we are just now finishing a strong HODLer accumulation phase and we have only upside from here to look forward to.
That’s it for this week, thanks for reading!
https://quoteinvestigator.com/2014/01/12/history-rhymes/
I appreciate the research taken every week for this.
Thanks so much! I look forward to these every week. The repeating cycle chart is how many people are predicting things may develop in the next months and years. Sometimes that's enough to make it happen.