Humans have an innate desire to first understand, and secondly reason, about what they see in the world. We go from observing things (facts), to reasoning about those facts (theories), to then applying those theories back on present facts to predict future facts. This habit pattern explains the birth of religion. It also explains how we can expect quantitative models to be formed, evolved, and applied by cryptomarket participants in the years to come.
New asset-classes, or exotic creations within asset-classes, gain mainstream attention via asset bubbles. It is only when an asset class is worth enough, for long enough, that serious work is put into quantitatively valuing it. This tendency explains why so much work was put into crypto valuations in 2017.
Relative pricing models are built relative to an asset’s past, or relative to its peers, with things like the price-earnings ratio popular in the equities space. Sometimes relative pricing can rely on a simple metric, as has been common with network values in crypto (barf), but the tool becomes more powerful when used in ratio form. In ratio form, it is often a “price metric” in the numerator, and a “utility metric” in the denominator, showing the price of an asset per unit of utility. Willy Woo and I were drawn to the NVT Ratio in the early days, and people have since iterated upon it, while new ratios like the MVRV Ratio are emerging regularly [2].
But there’s a wrinkle in here, because a theory doesn’t have to win a Nobel Prize to start guiding an asset’s price. Theories may first follow price, but then those that consistently work and achieve enough mental mass in the markets will start to influence price, potentially setting in motion a self-fulfilling prophecy for the valuation models that get to critical mass first. Values and prices are, after all, a strictly human creation.
[1] There are many “hybrid cryptoassets” for which exotic looking DCF models can be applied — for good examples, look at BNB and MKR valuation attempts.
[2] In late 2017 and early 2018 I gave a series of presentations on valuations (here’s one) and urged people to experiment with relative valuation ratios. Some examples that I gave then that still need exploring are Network Value to Growth (NVT-G, similar to a PEG Ratio, though growth in utility metrics beyond transaction value should be explored), Network Value to Trading & Transacting (NVTT, likely only suitable for reserve cryptoassets), Capital Expenditure to Network Value (CENV, this looks at supply-side strength, while most other ratios in existence today look at demand-side strength. I think assessing relative supply-side strength will get more important as similar cryptonetworks start to compete viciously for demand-siders, though we’re a long way from that).
[3] Some, like Professor Damodaran, would argue that it’s a fool’s errand to try and fundamentally value things like bitcoin, or any cryptoasset without cash flows, as he has said, “bitcoin is not an asset, but a currency, and as such, you cannot value it or invest in it. You can only price it and trade it.” We’ll see! I have lots of respect for Professor Damodaran, and so don’t take his opinion lightly.
[4] Note that hundreds of years passed between the creation of the first stock market and the application of DCFs to value equities. I expect crypto to converge on its models of choice an order of magnitude faster, as markets, data, and number-crunchers are now ubiquitous.
New asset-classes, or exotic creations within asset-classes, gain mainstream attention via asset bubbles. It is only when an asset class is worth enough, for long enough, that serious work is put into quantitatively valuing it. This tendency explains why so much work was put into crypto valuations in 2017.
Quantitative theories about asset pricing can come in the form of fundamental value models or relative pricing models.
Fundamental value models are built from the bottom up and can be used to assess an asset’s value in isolation, with DCFs the most popular technique for any cash-flow yielding investment [1]. Such models are the holy grail of valuation. The best model we’ve had thus far in crypto is the equation of exchange (MV = PQ), discounting future utility values back to the present. While we’ll need more time to understand MV=PQ’s predictive integrity, people have already started to iterate on its application. I also expect wholly different fundamental value models to be put forth and battle-tested in the bubbles to come, and Alex Evans has convinced me that options theory is where to look next.Relative pricing models are built relative to an asset’s past, or relative to its peers, with things like the price-earnings ratio popular in the equities space. Sometimes relative pricing can rely on a simple metric, as has been common with network values in crypto (barf), but the tool becomes more powerful when used in ratio form. In ratio form, it is often a “price metric” in the numerator, and a “utility metric” in the denominator, showing the price of an asset per unit of utility. Willy Woo and I were drawn to the NVT Ratio in the early days, and people have since iterated upon it, while new ratios like the MVRV Ratio are emerging regularly [2].
Any cryptoasset that wants to move beyond an asset-bubble and persist in value will need quantitative theories that convincingly explain its price over time.
In the coming years, I expect relative pricing models to be more widely used than fundamental value models, as they are easier to construct and comprehend. We will need more formal arguments and robust proof for any fundamental value models to take real root, and until they do, the market will remain highly volatile as we lack tools to collectively agree on a fundamental value [3]. Most asset-classes already have consensus on suitable quantitative models, but market participants continue to disagree on the inputs to those models. Right now in crypto we disagree on both the models and the inputs — hence the volatility [4].But there’s a wrinkle in here, because a theory doesn’t have to win a Nobel Prize to start guiding an asset’s price. Theories may first follow price, but then those that consistently work and achieve enough mental mass in the markets will start to influence price, potentially setting in motion a self-fulfilling prophecy for the valuation models that get to critical mass first. Values and prices are, after all, a strictly human creation.
[1] There are many “hybrid cryptoassets” for which exotic looking DCF models can be applied — for good examples, look at BNB and MKR valuation attempts.
[2] In late 2017 and early 2018 I gave a series of presentations on valuations (here’s one) and urged people to experiment with relative valuation ratios. Some examples that I gave then that still need exploring are Network Value to Growth (NVT-G, similar to a PEG Ratio, though growth in utility metrics beyond transaction value should be explored), Network Value to Trading & Transacting (NVTT, likely only suitable for reserve cryptoassets), Capital Expenditure to Network Value (CENV, this looks at supply-side strength, while most other ratios in existence today look at demand-side strength. I think assessing relative supply-side strength will get more important as similar cryptonetworks start to compete viciously for demand-siders, though we’re a long way from that).
[3] Some, like Professor Damodaran, would argue that it’s a fool’s errand to try and fundamentally value things like bitcoin, or any cryptoasset without cash flows, as he has said, “bitcoin is not an asset, but a currency, and as such, you cannot value it or invest in it. You can only price it and trade it.” We’ll see! I have lots of respect for Professor Damodaran, and so don’t take his opinion lightly.
[4] Note that hundreds of years passed between the creation of the first stock market and the application of DCFs to value equities. I expect crypto to converge on its models of choice an order of magnitude faster, as markets, data, and number-crunchers are now ubiquitous.