Projecting Bitcoin's Price with Popularity (Google Trends)
A Tale of Two Charts
The chart below shows the monthly average price of bitcoin since October 2013.
This second chart shows the Google Trends Score by month for the search term "bitcoin", also since October 2013.
When I first saw the second chart, I thought it looked remarkably similar to the first. Sure enough, it does. Below are the two charts together. The darker line shows the price according to the left-side axis and the lighter line shows the Google Trends Score according to the right-side axis.
It turns out that bitcoin's price and its Google Trends Score are quite correlated.
Specifically, they have had a 72% positive correlation since October 2013. This compares favorably to bitcoin's 26% correlation with the stock market and 24% correlation with gold over the last year.
I should note that the correlation between bitcoin's price and its Google Trends Score was much stronger up until both price and score peaked in December 2017 (98%) than it has been since (59%).
Nonetheless, the purpose of this analysis is not to argue about the level of correlation nor to assert causation one way or another (though it seems logical that the price and popularity feed off of each other). Rather, the purpose of this analysis is to make a speculative projection regarding what bitcoin's price could be if and when its popularity (as measured by its Google Trends Score) returns to and surpasses its previous all-time high.
In order to make this projection, we first must consider the historical ratio between bitcoin's average price and its Google Trends Score on a monthly basis. Let us refer to this as the price:score ratio for short.
Bitcoin's Price:Score Ratio
The chart below shows how bitcoin's price:score ratio has increased over the last several years.
For a ratio that has been relatively volatile historically, it has remained unusually consistent for the last two months (June & July) at 729, just 4% short of its all-time high of 758 in November 2019.
Bitcoin's price:score ratio has been above 645 for 9 of the last 12 months after having only reached such a level once before, and barely (646 in October 2018).
The 3 months in the last year during which the ratio was lower than 645 can be fairly easily explained. They took place this past March, April, and May as hype grew in anticipation of bitcoin's once every four years halving event on May 11, 2020. Additionally, people perhaps were searching more than usual for information about alternative assets (including bitcoin) when the stock market crashed ~30% in the early days of the lockdown resulting from the pandemic. Those two factors together seem sufficient to explain the uptick in bitcoin's Google Trends Score over that period. On the other side of the price:score ratio, bitcoin's price fell along with the market, and so with the price going down and the score going up, the decreased price:score ratio over those few months makes sense.
Since the stock market substantially recovered and the halving event came and went without materially moving bitcoin's price, the Google Trends Score has returned to the pre-pandemic, pre-halving-hype range. For the last two months (June & July), bitcoin's monthly Google Trends Score has stayed steady at 13, which is equal to its average from April 2018 (after the peak) through February 2020 (before the halving hype and market crash).
For some additional context, it should be noted that bitcoin's Google Trends Score has only been equal to or greater than 21 in 7 months since bitcoin's inception and they all took place consecutively surrounding bitcoin's last major surge from September 2017 through March 2018 (with scores chronologically of 21, 22, 44, 100, 55, 39, and 23).
Flying Under The Radar
Sufficed to say, bitcoin's average score of 13 since the last surge, excluding the three aforementioned months from earlier this year (14 if we include them), shows objectively that bitcoin is flying under the radar (in terms of its popularity as a search term at least). Of course, it may not seem this way to people who follow crypto closely and daily but the algorithmic and mass-data collecting Google Trends Score should be trusted over any single person's speculative and subjective observation about the current level of hype and attention.
If you, like me, believe that bitcoin is and has been flying under the radar for more than two years now, the natural question to ask is this. What do we expect to happen when that is no longer the case? In other words, what happens to bitcoin's price if and when it reaches the pinnacle of its next hype cycle? How high might bitcoin fly the next time we see a situation similar to that of December 2017?
Projecting Bitcoin's Price
I believe a reasonable response to this question can be made by applying bitcoin's average price:score ratio over the last twelve months (643) to a Google Trends Score of 100 to estimate what bitcoin's price would be if the price:score ratio stays about constant and bitcoin's popularity in terms of search returns to its all-time high. The price implied by this projection is about $65,000 as is demonstrated by the maximum value in the top right corner of the chart below. Basically, if one believes that bitcoin will at least at some point return to the level of popularity it attained in December 2017, this methodology would lead that person to expect bitcoin to attain a price of at least $60,000 at that time.
This estimate can be made less conservative in two ways. The first is to assume that the price:score ratio will continue to increase as it has all along (6% month-over-month), rather than calculating the price using the LTM average of the ratio. The second is to assume that bitcoin's popularity as a search term will well surpass its previous all-time high set in December 2017 just as that all-time was more than twice as great as any high before it and more than 8 times greater than the all-time high before May 2017 (which was set in December 2013, perhaps not coincidentally exactly 4 years or one halving period prior to the latest all-time high). I hesitate to call either of these two less conservative assumptions aggressive because they are not only reasonable but actually probable if one is to simply assume that the historical trends will continue in terms of the score and the ratio.
As such, I will close by presenting a chart which one can use to approximate the price implied by their own inputs for the Google Trends Score and the price:score ratio. In this chart, I show the ratio for the last 12 months, 2 years, 3 years, and so on, and use the average year-over-year growth rate between those figures (16%) to project the next twelve months ("NTM"), next two years ("N2Y"), and next three years ("N3Y") ratios.
The highest price projected on the chart is right around $300,000 based on a Google Trends Score of 3 times the previous all-time high (300) and a projected N3Y price:score ratio of 996 (the LTM average of 643 increased 16% 3 times for 3 years). Of course, this maximum price projection is limited only arbitrarily by the maximums I have decided to present on this chart for both the score and the ratio.
To Be Continued...
Many people have employed many methodologies in attempting to accurately project bitcoin's price. For example, PlanB's stock to flow ratio looks at scarcity to make one of the more convincing price projections I've seen. My price:score ratio looks at popularity in endeavoring to do the same. This analysis was not intended to be totally comprehensive. My intention was merely to introduce this methodology for others to critique. As such, my hope is that this becomes widely enough read so that I may receive ample critical feedback for further consideration and refinement of the methodology.
As always, I welcome all outreach to jake@blogofjake.com and any comments on Twitter @blogofjake