In the jungle of the fiat world, there are crafty creatures known as scammers who have a distinct trait: they never admit they're wrong. You could proclaim, "Hey, this isn't working!" They'd shake their heads, waggle a finger, and claim you simply misunderstood. Suggest that Bitcoin would be a more useful route? They'd passionately argue for a convoluted token system. Even if you stated the obvious - that the sky is blue - they'd spin doubts about your statement by talking about its purple hues.
You see, scammers, like poker players, know that a single misstep can shatter the illusion of trust they've carefully built up with their victims. They've been so deeply conditioned to lie that they'll parry any perceived attack, all to maintain their delicate house of cards.
Now, let me make it clear, I'm aiming to be brutally honest with myself. I've previously analyzed the stock-to-flow model and, with conviction, declared it the only price model that made sense with me. While that hasn't changed entirely (though that says more about price models than this particular one), I've realized that something making sense doesn't necessarily make it true.
The past 18 months have been a philosophical sabbatical from the price-driven narrative of Bitcoin, mainly because this particular set of models have let us down. It was wrong, and there's nothing like a major error to trigger some deep self-reflection. So, armed with introspection and hopefully, some humility, I pen this article.
Let's revisit the source - the stock-to-flow model. My understanding of it, and I suspect, many others, hails from Saifedean Ammous's enlightening work, The Bitcoin Standard. It suggests that the existing stock of an asset, coupled with its annual increase, reveals how challenging it is for us humans to accumulate or create more of it. This is the core concept, and it's a handy tool for understanding relative scarcity and the effort needed to obtain a certain resource.
However, the model is a very different thing. Initially, the S2F model was a clever method to link an asset's stock-to-flow ratio with its market cap. At first glance, it seemed plausible; scarcity does, after all, influence price. But, as I reflected, I realized that scarcity is just one cog in the vast machinery of price determination.
Scarcity can nudge supply and possibly demand, but there's so much more involved. When it comes to supply, think about the influence of knowledge, technology, labor, and time. On the demand side, factors multiply - potential uses of the asset, liquidity, portability, and so on. So, the price, and thus, the market cap of an asset, isn't a straightforward equation of a single variable like scarcity.
In retrospect, I see now that many of us, myself included, yearned for this model to hold true. We were seduced by the possibility that since the supply of Bitcoin is so predictable and known, we could perhaps also do the same with price. The S2F/S2FX models promised the allure of future knowledge in the rollercoaster ride that is Bitcoin's global adoption.
Could there still be a model out there that accurately mirrors price? Possibly. But as with any hypothesis, it must stand the test of accuracy. If it doesn't hit the mark, it's time to swallow our pride and admit we got it wrong. It didn't and I was wrong.
Great article. I am generally suspicious of price models, but I also liked the s2f. Before we dismiss it, is it worth considering how any model can be affected by manipulation? I remember when Jimmy Wu "threw in the towel", albeit temporarily, out of frustration about what he perceived as paper bitcoin manipulation in the futures market. Bitcoin TINA expressed similar concerns, and sold a portion of his holdings.
Is it time we ignore the price and just focus on adoption?
reply
Adoption will be one of the drivers of price -- perhaps the main one. It would be a more fruitful place to direct attention, but probably not a very psychologically satisfying one.
Though maybe these days it would count as therapy.
reply
Well written Jimmy. This reflection challenged me to think about how I immediately accepted S2F as fact and then, when it did not hold true, completely forgot about it. One final thought, maybe the first two paragraphs are not super necessary?
reply
Price determination is a process and fundamentally rooted in subjective valuations. While certain models may give insights about price dynamics or hint at misvalued assets, there can never be totally accurate pricing models.
All models are wrong. Some models are useful.
reply
To develop a model that mirrors price, wouldn't we need an accurate model of demand, and therefore human behavior? I am often guilty of being too optimistic on the inevitable happening quickly, so S2F looked reasonable at first.
reply
"All models are wrong, some are useful".
I quote this from @Undisciplined as this is something engineers must hold in their minds during their profession.
PlanB mentions that in this model it is assumed that scarcity drives value. And that scarcity is modelled via SF = stock / flow
I personally believe this model has not yet proven wrong and should be evaluated when the external factors influencing the price are dormant, for example, I think some blocks before each halving (unless something extraordinary happened like the COVID shock). I have personally taken a few datapoints (in the s2f vs. market cap space) before each halving and fitted a logarithmic expression and the fit so far was quite good. i can give more info if anyone interested.
If this expression holds by halving 2024 we should be at approx. 120k USD per bitcoin. Looking at the spiral charts from the twitter user @therationalroot this is not too crazy, since the spiral has never broken and we're heading to the next market cap ring (or in price 100k USD per bitcoin). I would also not be surprised if we're even at half of that by halving affected by the diminished ATH from the previous cycle due to all the shinanigans with paper bitcoins... who knows, no one can predict the future.
At the same time, the on-chain data analytics suggest that bitcoin on exchanges is drying up, while at the same time BlackRock and others are adding to potential future demand if the spot ETF's are approved.. both of which support a tendency towards a potential higher price due to increase in demand while decreasing supply.
I don't know and honestly I don't want to rely on price models for my individual life plan because at the end they are models, but I just want to say many people trash about this one being useless while I think the model still has not been proven wrong and in my opinion this model does not take into account hype at different periods (for that you need to add some statistics from previous cycles to the base case model).
PS: At the end, price is driven by supply and demand forces, and those can be modelled to some extent, since they are dependent on user adoption (sometimes unpredictable) due to acknowledgment of the use case for example from the demand side and also external factors (sometimes unpredictable) on the supply side.