The ETH/BTC Trading Pair
Directional price predictions through parametric analysis
The Ethereum community has been very excited the past two weeks about the technical breakout of the ether/bitcoin trading pair. Breaking out of a symmetric triangle tends to be a big deal on its own, but today I want to talk about the hidden underlying sentiment that is driving the price action.
I refer to my Trading bitcoin exchange flows during a bull run article frequently (I think it may be my strongest piece of analysis), but I’ll summarize it quickly before extending the analysis to model the ETH/BTC pair.
In a healthy market: higher prices encourage holders to take profits.
When the price of an asset goes up holders are more likely to sell.
When the price of an asset drops holders are less likely to sell.
During a selloff the effects are diminished because confident investors see it as a buying opportunity.
In an unhealthy market: lower prices encourage holders to take losses:
When the price of an asset goes up holders feel relief that the market isn’t about to collapse and are less likely to sell.
When the price of an asset drops holders are afraid and become more likely to sell because they fear the market may be about to collapse.
During a selloff the effects are amplified as scared market participants rush for the exit.
With that thesis in mind, we can use exchange inflows and outflows as a proxy for holder sentiment and measure the evolving correlation between net exchange flows and price changes to gauge how fearful and volatile the market will be.
By analyzing the exchange flow vs. price correlations for both bitcoin and ethereum and then taking a difference to examine their relative values, we can eliminate the biggest hurdle to using this analysis to model ethereum—it’s correlation to bitcoin.
Seen in the figure below, throughout 2021 the relative correlations have been a strong predictor of forward price directionality. When the shaded area in the figure is green, the relative health of ethereum (measured by exchange flow vs. price correlations) is healthier than it is for bitcoin. When the shaded area is red, bitcoin is the healthier asset.
It should be noted that the model broke down in the midst of extreme price swings directly after the Chinese cryptocurrency mining ban. Further analysis excludes data from May 11, 2021 to July 1, 2021.
It is essential to analysis with this model to consider not only the absolute value of the relative correlations, but to also examine the trend in the change of the correlation1.
The current situation is a ETH-BTC Up-Trending Difference in Correlation of 0.28. That’s a lot of jargon, but it simplifies into the figure below. What the data tells us is that in the short-term (modelled with two-week forward returns) there is very little historic downside and a lot of potential upside.
Stepping a little further into the data, on 96 up-trending + positive data points, 75 have seen gains for the ETH/BTC pair vs only 21 seeing losses2.
That’s a 78% success rate—a rare feat in financial analysis.
In the 21 fails, the average loss was 2.8%.
In the 75 successes, the average gain was 14.0%.
The average return including successes and failures was +10.3%.
In addition, with ether’s high beta to bitcoin, by holding ether you maintain essentially all of your exposure to bitcoin’s potential price appreciation. Contrastingly, by holding bitcoin you are far less exposed to gains in the price of ether.
As a final note, post EIP-1559 we have now seen a significantly flippening in Bitcoin vs Ethereum dynamics. The average inflation rate for ether tokens is now comfortably below bitcoin’s inflation rate and will likely stay there until the merge when ethereum will turn deflationary.
Trends in this model are defined as changes in relative correlation of at-least 0.1 over the proceeding 7 days. They are broken into up-trending, down-trending, and flat-trending.
The bitcoin halving cycle is defined as beginning May 11, 2020. The data analysis is up to December 7, 2021.