March 14, 2021: Inflation, MMT and Bitcoin 🖨️💸
Part 2 of a 2 part series on Inflation, MMT, and Bitcoin
In part one of this series, we talked a lot about the hot topic of inflation and money printing. Some people claim the Fed’s policy is an experiment in MMT. Yet inflation isn’t actually tolerated within MMT.
Today, the Fed doesn’t deny inflation is here. The debate now is simply “how much.” There’s plenty of fun trickery the government can use to manipulate inflation calculations. They can simply claim steak and dog food are substitutable products. Or maybe they the prices are temporarily offline.
We’re not going to look too much at whether inflation numbers are manipulated, assuming the most scrutinized nation on the planet could only cook the books for so long before it showed up in other indicators. Also, the Fed is currently admitting to a 2.26% inflation rate, the highest in years, so even if they’re lying it’s at least directionally honest.
The real concern — if some people are getting $1400 worth of stimmy, what happens if they find their cost of living has increased by $1500 or more? What to do?
Time to bring Bitcoin into the conversation. Bitcoin supply is finite. This makes it the Hegelian dialectical antithesis to federal money printing. It intuitively makes sense as a hedge. It’s a narrative commonly pushed by bitcoiners:
Meanwhile the popular press is propagandizing against this narrative.
The popular press tends to be wrong about all things Bitcoin, so your independent thought alarms should be blaring. But what do the data say?
Better Call FRED
The FRED economic data are great because they have a large amount of consistently measured data that’s easy to import and manipulate. FRED is also great because they cited Curve in published research:
Using FRED data, we can directly compare inflation metrics and bitcoin price to see if there’s any trend at all. The great part here is it’s reproducible with a few lines of code, so anybody who likes can check my work and build on it easily.
First the hypothesis. If investors saw BTC is an inflation hedge, we would expect to see prices go up anything inflation indicators picked up. We can use a few basic concepts:
correlation: 1 if the change in prices are perfectly positively correlated, -1 if negatively correlated, 0 if no correlation
beta: does the asset move directionally with the base asset? 1 if they move about the same. >1 if it moves directionally in a more exaggerated manner, <0 if they move against each other.
For example, look at FRED data for S&P daily closing price (SP500) and the Dow Jones (DJIA) closing price. We expect these would mostly move in the same direction — on most days all markets close green or red. We always consider the % change when looking at these:
This is what a correlation of 97.2% looks like. Perfectly linear. Beautiful.
What about beta? Sure, beta is generally used to compare the directional movement of two assets. But it’s just the covariance divided by the variance, so in principle we can use it for anything. For instance, here’s how it looks for the Dow vs the S&P:
This is a beta of 1.02, as the two indices basically move parallel (beta hovers close to 1) over the past five years. This serves as your sanity check to see what things look like when they work perfectly in unison.
Gold
OK, let’s use Gold as our primary benchmark when considering inflation. Gold is considered the classic hedge against inflation. We can compare gold price (GOLDAMGBD228NLBM) to the 10-year breakeven inflation rate (T10YIE). Let’s see what this looks like in terms of the raw prices:
We only care about daily change though:
This is what a 21% correlation of daily change percentage looks like. It’s not as obviously linear as our S&P chart above. Still, we see gold prices rising when inflation peaked, and often dropping when inflation rates dropped. What does the beta look like? We overlay the SP500/DJIA (blue) beta, which was near 1, for reference.
So, in fact gold doesn’t particularly move with the inflation rate at all, except for a brief spike in 2020. When inflation rate moved from 0.5% back to 1.5% over the course of 2020, gold also ran up way faster.
This is what the classic hedge against inflation looks like under scrutiny. Now let’s do Bitcoin!
Bitcoin
How does Bitcoin price daily change correlate with inflation? We can’t see much with the raw price charts, since Bitcoin’s going through a boom cycle right now.
Hence we always use % change:
Gross, what did I just look at? Modern art? Nope… that’s what a lowly 5% correlation looks like. It’s essentially random.
Here’s the beta chart like we showed above for gold, with S&P vs DJIA filling in to demonstrate a beta of nearly 1.
Hmm… sort of looks like our gold beta chart. It’s a lot of beta near zero, but it sort of tried to jumped up a bit during 2020 (though not to the same degree as gold). What on earth was happening?
If BTC prices had much to do with inflation, it’s at best a severely lagging indicator. One could try to argue that in late 2020 BTC took over as gold among investors as a hedge, but the evidence isn’t really showing up in the rolling beta.
Correlations
The funny thing with this analysis, is that Bitcoin price movements are rather poorly correlated with any other asset or indicator.
For example, we saw above what a 97% correlation looked like between the DJIA and S&P. This is like a holy grail in financial indicators. How does DJIA compare with other indicators? Here we look at daily changes of DJIA vs S&P, along with several other FRED daily indexes. FRED has thousands of indicators, so we had to do some cherry picking here for what we thought looked interesting and/or relevant.
The stock market has a lot of fairly strong correlations — here we consider the time horizon of 2017 - present like above:
We saw earlier DJIA running 97% with S&P, and it shows strong correlations with the Volatility Index (-62%) and BofA High Yield Index (+59%). It even moves +31% with the inflation rate, suggesting some possible correlation or perhaps even causation (ie stocks as hedge against inflation).
What about gold? Nothing quite so strong as the stock market, but we do observe some movement in tandem, at least the 20-30% correlation range:
Gold prices move noticeably upwards whenever the dollar gets weaker, which makes some intuitive sense. The strongest correlation with gold price is the dollar index (-31%). It also moves somewhat in tandem with the US/Euro exchange rate (+27%) and China/US exchange rate (-21%) — both of which directionally indicate rising gold price on weaker dollar. We also see gold move in tandem with the 10 year inflation rate about 21% of the time, again one of the stronger indicators.
Now compare the above correlation heatmaps with Bitcoin price movements.
The incredibly light pastel shading here suggest Bitcoin price is almost completely uncorrelated with any other indicator. It’s a fully independent beast.
The closest correlation I could find among daily indicators on FRED was the stock market, yet they move together just 16% of the time. Accordingly, it had a 15% negative correlation with VIX, which had previously shown to correlate with stock prices.
Everything else displayed less than a 10% correlation. This makes it tough to argue there’s any sort of relationship between Bitcoin and inflation — or really Bitcoin and any existing financial indicator, going back as far as 2017.
Well, what about monthly indicators? FRED has a lot more data that only gets reported once a month, so this gives us a lot more potential indicators. Here we expand our time horizon back to 2014 to give us a larger sample size.
OK, maybe looking at this set of data we can possibly pick out some indicators? There’s pretty much zero correlation the 10-year inflation rate (T10YIEM), but there does seem to be some apparent movement with possibly related measures. The 10-Year Treasury Inflation-Indexed Security, which dips when inflation is high, shows about 19% correlation with BTC price. We also note 18% correlation with changes to the Consumer Price Index (CPIAUCSL), often used as a gauge of inflation.
Also a handful of personal consumer metrics seem to run slightly in tandem — +18% Employment Level (CE160V), +16% Personal Consumer Expenditures (PCEPILFE), and -15% Personal Savings Rate (PSAVERT). This tells the story that when retail users have a more secure situation (higher employment, lower savings) then the price of bitcoin goes up.
My first thought was that maybe this was because this time horizon went back to the 2014-2017 range, where Bitcoin was cheap and not heavily an institutional game. But looking through the data, it seemed as if the bulk of this correlation in these instances was coming from 2017 onwards. By way of illustration, 2014 to present for CPI demonstrates an 18% correlation:
Looking just at 2017 onward shows a higher 27% correlation:
Really, the effect here is to overweight that 80% candle in September 2017, when CPI also was noticeably increasing quite fast. Still, looking from 2018 onward and we still observe a 17% monthly correlation, which is fairly consistent over this period.
Unfortunately, it wasn’t trivial to run these analyses on monthly stock data, since too often the markets were closed on the first of the month when other data was reported, and I didn’t have time to backfill the gaps.
Takeaway
Pressed to interpret all this — is there any strong evidence that investors turn to Bitcoin as a hedge against inflation? Not yet. For the past four years, here’s how various daily indicators tended to move with inflation:
Whenever inflation actually rose, investors bought stocks, oil, some gold. Bitcoin moved essentially zero.
We do see some slight evidence that Bitcoin prices tend to rise when conditions change for retail investors. When employment rates and personal expenditures rise, savings rates drop, and CPI increases, Bitcoin prices tend to go up. You could use this to cobble together an argument that retail investors make a turn to Bitcoin as a hedge against perceived inflation. It’s a bit thin, but there’s at least a little data to work with.
That’s not to say Bitcoin is actually a poor hedge against inflation. In terms of real returns, the best place to have your money this past decade would have been to mine the Genesis block, then cash it out for about one day to buy Gamestop.
Still, if you want arguments of what causes movements in bitcoin, you’d probably be better off looking at fundamentals like hashrate and halvening:
For more info, check our live market data at https://curvemarketcap.com/ or our subscribe to our daily newsletter at https://curve.substack.com/. Newsletter is an independent roundup of interesting trends in cryptocurrency, never financial advice. Author is a $CRV maximalist.