Inflation In A Hyperbitcoinized World
a cheeky little brainstorm on the methodological individualism of price inflation and capital formation
OMG GUYS WE RECORDED OUR CLUBHOUSE FOR ONCE!
What a glorious moment for culture and humanity. Thanks to the good folks at Swan Bitcoin, Aleks Svetski and I spoke to Izabella Kaminska for some ridiculous amount of time until we all fell asleep at 2.15 am or so. Once we level up our competence yet again and turn it into a podcast, you will able to able to verify the following for yourselves, but for the time being you will have to trust me: Izzy said there could be inflation in a hyperbitcoinized world.
And she’s right.
Calm down, maxis, calm down. Explaining why this is true sheds light on yet another of Bitcoin’s quasi-magical properties and paints our hyperbitcoinized future an even rosier shade of orange. This is good for Bitcoin.
I kinda thought I’d explained this well enough in The Capital Strip Mine, but then again, it’s like, suuuuuuper long and people are busy, so I’ll try to keep this under a 10-minute read.
Narrator: he failed.
Inflation — by virtue of being an economic phenomenon — has two different kinds of explanation: methodologically individualist ones and wrong ones.
The wrong one is the degenerate fiat economics trope we are witnessing in real-time as lumber go up yet number stay down. I’m not going to waste any of your time explaining the machinations in which central bank economists are willing to engage to keep the number down because I also don’t even really care. It’s Brandolini’s Law, once again, and I happen to simply not have orders of magnitude more energy than they do.
The methodologically individualist explanation requires an important distinction between price inflation and supply inflation. Central bankers are aware of this too, clearly — the entire point of their machinations is to try to convince you the two have nothing to do with one another and everything is under control, because they think you are stupid.
But you aren’t stupid, are you? You know that inflation of the money supply happens in fiat terms whenever a bank creates a loan ex-nihilo without ever writing off a previous credit asset (on a net basis) or when the central bank creates new reserves ex nihilo. You also know that this happens in Bitcoin never.
So what is Izzy talking about? Is she a useful idiot of the chattering classes, as I confessed in Clubhouse to once having called her?
No, Izzy is based AF. And she’s not stupid either. She knows she is talking about price inflation: how much stuff costs. Can the price of stuff ever go up in a hyperbitcoinized world? Yes. Here’s how.
To capture Izzy’s idea in degenerate fiatspeak we would say “an increase in velocity,” and it’s worth explaining exactly why this is so silly before describing the same thing in plain English. As Mises knew well, “velocity” is pretty stupid, and as Saylor has said of such other pretty stupid aggregates as “GDP growth” and “CPI,” it is a metaphysical abstraction. It comes from Fischer’s Equation of Exchange from the Quantity Theory of Money, usually written simply as MV=PT.
This has the curious characteristic of not really being an equation but rather a tautology, a definition, and a destroyer of useful information. It’s really quite bizarre. What it actually says is that “velocity” is defined as PQ/M, or everything bought in a year, times its price, divided by the money supply.
But were we minded to behave more like physicists (i.e. real scientists) than economists (i.e. not real scientists) we could do some cheeky dimensional analysis on this equation, and start off by leaving out V entirely. Let’s just call it “?” because it could be anything! Who knows? We might make a discovery!
so ?=PQ/M
now the dimension of “P” is $ (the average price of all the stuff), and “M” is also $ (the total amount of money), so they can cancel, but we will wait a moment before doing that.
“Q” is a super weird variable. It allegedly means “quantity” — as in, how many of the things were sold at the average price P that year? Why does that matter? It seems dimensionless so it really doesn’t (the Wikipedia page above is unintentionally hilarious on trying to explain this away, by the way) but actually what it is really there for is “that year,” meaning it gives us the dimensions of 1/time, as in, how many per year? So P*Q becomes: how much $ was spent on stuff in a year?
Dividing by M leaves us just with 1/time for “?” or “how many [of something] per year?” — how many of what?
Believe it or not there isn’t really an answer that refers to the real world, despite all the inputs being (apparently) real. The problem is we have multiplied up an average, apparently solely for the sake of dimensional consistency, then divided it by something of the same units but to which it bears no real-world relation. What is left over for “?” to “explain” is “how many times did the average dollar get spent?” Let’s call that “velocity” because why not. It sounds like it’s about speed. It’s not a vector but we are well past any of this making sense, so who cares.
“Velocity” has been tautologically defined (yippee!) but it tells us absolutely nothing because there is no “average price” and hence there is no “average dollar.” The average dollar is a metaphyiscal abstraction, or, to quote Mises, “lol, what? velocity? you serious, brah?”
All this averaging and abstracting obscures that all of this happens across time. What is being captured here is dynamic, but the equation of exchange implies, without ever being too explicit, that these quantities just exist as features of the universe. The degenerate fiat economist would say “velocity went up” when what he really meant was, “persons x, y, and z made decisions to spend more for reasons a, b, and c.”
Or, crucially, if asked to explain “inflation” any attempt will be next to incomprehensible. Let’s evaluate the options:
Velocity went up: meaningless.
Quantity went down: huh? No seriously, what does this mean? Here we finally pay for the mistake of including a variable that does nothing beyond throw “1/time” into the mix to balance the dimensions but otherwise has no clear way to be understood in real terms.
Money supply went up: sounds appealing at first glance, but, for example, does this mean state money increased or bank money? Credit or equity? If unbacked credit has been created ex nihilo, as fiat bankers are wont to do, how do we distinguish this from velocity? Is it even possible to hold velocity constant and point purely to the money supply? Doesn’t that require some even more confusing offset somewhere else? The mind, it boggles …
Price level went up: wait, what? Isn’t that exactly what we are asking? Yes, but it gets even more confusing because “P” is “the average price,” which tells us absolutely nothing if we want to know about a specific price.
Let’s scrap this nonsense and go with the Farrington Theory of Price instead, which I just made up right this moment:
Price = Money spent on stuff / Stuff available to be bought
This is consciously descriptive of the aggregate level of “stuff,” such that ∂-price might give some useful measure of “inflation,” but, unlike the Fischer waddayacallit above, we can easily translate this to the price of an individual good or service:
Price of good = Money spent on that good / Amount of that good available to be bought
This has the benefit of being both simpler and more accurate in terms of what it captures. It’s a gloriously elegant model. It’s basically science. It may even be The Science™.
This now ties back very nicely to my (much longer) explanation in The Capital Stripe Mine. Given (price) inflation is not a metaphysical abstraction but comes from individuals making decisions and acting, it is now pretty clear what decisions and actions lead to inflation: when the money being spent on a good goes up or when the amount of that good available goes down.
Notice this is saying very little more than: either demand went up or supply went down. This is a very good sign given that’s literally all of economics in one sentence. The tiny bit more it is saying helpfully hints at why either or both of these has happened.
The merchant will have a window of time in which she expects to sell the amount in her inventory before then expecting to restock. If she gets through her stock faster than this, and still has more inquiries she cannot meet — or if she has the same number of inquiries but gets through her stock quicker because there was less to begin with, then she will decide — she will act — to increase prices to boost her margins, boost her returns, and better clear the market. And there you have it, price inflation. Customers decided, for reasons, to buy more stuff, and the merchant decided, for reasons, to raise prices. Methodological individualism 1, metaphysical abstractions 0.
I mentioned above that the Fischer waddayacallit gets very confused about the different kinds of money supply growth. We can now be a little more specific because the Farrington Theory of Price doesn’t address this either. Is this a flaw in my Nobel-(memorial)-worthy theory?
Lol. Of course not. The difference is the Fischer equation pretends to capture this, and then when you try to work through the implications of varieties of monetary inflation, you just end up confused. Mine doesn’t even try because … drumroll … to the individual, it doesn’t matter!
The merchant does not say to herself: “Gee whizz, I sure would like to raise prices, boost margins, and increase returns, but I can’t for the life of me tell if the money being spent on stuff faster than before has come from unbacked credit creation by banks, from central bank reserve creation, from customers dipping into their savings, from my competitors going out of business, or some combination of all four or possibly more! What a pickle!”
She just sees the change and reacts. This is a perfect example, incidentally, of the one sensible way in which markets can be considered “efficient”: with respect to information. She doesn’t need to know the answer because she has been given the pure, condensed signal of customer behavior, into which all the relevant information has been compressed. She knows how to act in response without needing to know why.
The Fischer equation has special difficulty with the final two options considered— dipping into savings and the competitive landscape shifting — because these are distinctly local events that cannot be explained within an aggregate or an average.
This all gets nicely to Izzy’s point. You absolutely can have inflation in a hyperbitcoinized world, and the Farrington Theory of Price makes it totally obvious why. There are two possible reasons:
Stuff bought goes down: not that interesting as likely indicative of a natural disaster, but for the sake of argument, let’s suppose we are in the wake of a credit crunch in which previously misallocated capital is being liquidated and less is being produced in the interim. This also won’t happen in a hyperbitcoinized world, but just humor me, will ya?
Money spent goes up: this is the trap.
The terribly, terribly silly Bitcoiner answer here is to scream “21 MILLION!” until somebody calls the police for breaching the peace.
The sophisticated Bitcoiner will realize there are all sorts of reasons money spent could go up that, given they do not result from an expansion of the base money supply BECAUSE 21 MILLION YOU IDIOT, are worth thinking about more carefully to see what, if anything, they reveal.
Izzy’s idea was straightforward enough and I alluded to it above: people start to dip into their savings to spend. For any confused not-quite-bitcoiner readers, “savings” is an antiquated concept whereby you earn money and then don’t spend it. No seriously, you just don’t — you leave it alone and then you have the option to spend it in the future. No seriously, because in a hyperbitcoinized world there won’t be a bottomless pit of consumer credit so if you want the option to spend in the future you will have to “save.” It basically means “not-spend-right-now,” but it has fewer letters and syllables.
But actually, patterns of investment could affect this dynamic in a more complicated way. Bitcoin TINA and I (and implicitly Saifedean Ammous too but we’ve never all talked about it together) have a fascinating disagreement about to what extent, if any, credit will exist on a Bitcoin standard. TINA thinks none, I think some, but neither of us can prove it, obviously, so we just have to yell at each other in Clubhouse while we wait and see. We did so in Izzy’s episode. It was 🔥.
In any case, credit would be the easiest-to-understand version of my following point but is by no means the only one that works. Imagine literally any temporally discordant boost in capital investment. Perhaps this could also just be a more roundabout version of “dipping into savings.” People snap out of their scroogey miserdom and realize that if we just rely on deflationary money to provide for us we will literally all starve and die within a week or two, and they shift their liquid capital into illiquid capital accumulation.
You don’t even really need to think about this too hard. That investment could go towards all kinds of things: likely mostly novel capital goods, but also real estate, wages, financing services (bankers need a cut too, obvs), which will quickly filter to regular people looking to consume (or “save” [spits on ground]). By any number of channels which, as good methodological individualists we don’t need to pretend to understand, money spent will go up. Merchants will react in order to maximize their returns. Inflation. Boo! hiss! Bitcoin is deflationary! 21 million! 21 MILLLLIIIIIIIIOOOOOOOON!!! 😡
Just wait! What happens next is key — and unfortunately spoils what I thought was the most interesting bit of the entire podcast, besides Izzy’s musings on the future of biowarfare. Don’t quit before you get to that part, fo sho.
What happens next invites — arguably demands — clarity on the typical capital structure post-hyperbitcoinization. And this is where TINA’s and my disagreement really becomes clear as rather more aesthetic than practically consequential.
If there is debt in such a world, there is a very, very small amount, at least relative to now. So with either zero debt or very little debt, entrepreneurs will be in a position to react to these price signals with maximal flexibility. Contrast this to degenerate fiat finance in which the answer to every question is, more unbacked credit.
Stock market volatility? More unbacked credit. Repo market liquidity crunch? More unbacked credit. Too much unbacked credit? More unbacked credit. Tea or coffee? More unbacked credit.
But in a hyperbitcoinized world, the answer will be … well … I frankly have no idea because it will emerge from methodologically individualist entrepreneurs acting from flexible balance sheets and no-longer-politically-misaligned incentives.
But one thing I can say relatively confidently is that any “inflation” will be short-lived. It will only serve to reflect some unexpected change in economic reality that it requires a reallocation of capital to satisfy. But unlike in degenerate fiat finance, it will not either be the product of, or itself be, permanentized inflation from unbacked credit that will never be redeemed. It will actually be healthy inflation 😱 because the underlying money/stuff situation will be real and will demand it.
In the moment, that is. In anything longer than the very, very short (basically instantaneous) run, entrepreneurs will deflate it away again with effective capital allocation that identifies, responds to, and improves upon this new reality.
This all provides yet another, far more roundabout and conceptual route to explain the idiocy of “velocity.” To the extent the variable means anything at all, it is entirely endogenous and has no causal import. So, it emerges from a range of other behavior, and doesn’t then affect any behavior. To take it seriously is to be deeply confused.
But what else is new? The end. Listen to the podcast when it comes out. Peace and Love.
follow me on Twitter @allenf32
