The Big Long – Valuing ETH Based on Future Cash Flows – 03/12/18

After reading John Pfeffer’s work entitled “An (Institutional) Investor’s Take on Cryptoassets” (@maarnio), I have concluded that MV=PQ should be pushed further down the list of useful metrics to value digital assets, especially base-level Proof of Stake (POS) assets.

Pfeffer makes the argument that earnings from POS chains will asymptotically approach 0 over time because of an environment creating nearly perfect competition between chains, specifically because of the ability to hard fork at will with open-source code. I think he is off-base on this approach and severely discounts the synergistic nature of these chains. If POS chains created perfect competition because of their open-source nature and forking, then ethereum classic would have near the same adoption rate as ethereum. It doesn't. I have made the case against perfect competition in previous comments.

In the paper, Pfeffer makes the case that ETH network “revenue” currently sits at about $355 million per year. He used numbers closer to peak prices, so the $355 million in revenue per year may be a bit high. He goes on to extrapolate a possible 1000-fold increase to around $363 billion in 10 years. Yet because of his assumptions of a perfectly competitive environment, he concludes that, using MV=PQ, the network would only be worth $52 billion and most value would accrue to users. If the network were a stock, this would equate to a P/E ratio of 0.143 ($52 billion/$363 billion), which would be considered absurdly cheap in traditional finance.

Using his numbers, we can easily calculate an expected price for ETH using traditional Price/Earnings (P/E). Typical P/E ratios for the S&P 500 have consistently fluctuated around 15 throughout its history, so we will go with that number. In our example, yearly revenues of $363 billion multiplied by an average P/E of 15 results in a network value roughly $5.4 trillion. Of course, all $5.4 trillion won’t accrue to stakeholders. Network operation and maintenance costs need to be paid for. If we discount $5.4 trillion by 20% to account for these costs, then we can conclude that roughly $4.3 trillion will accrue to investors over a 15-year period. This network value of $4.3 trillion equates to a value of $43,000 per ETH, assuming 100,000,000 ETH in existence in 10 years. Many assumptions are being made here, as is generally unavoidable with any type of future extrapolation.

If Pfeffer is right about utility chains tending towards a zero-fee model, then I think EOS is the likely candidate because it started out of the gate with no fees. However, I don't think that will work long term either. There are real and significant costs to redundantly processing and replicating data over many machines and storing copies of that data on many machines essentially forever. This cost must be borne by someone. If users don't pay for it, then investors will, and I don't see investors buying into that model for long. There are edge cases where indirect revenue streams can flow to investors in that model, but it may not be enough to sustain the system. Ultimately, I think a POS systems with a fee market will prevail.

Many other variables exist such as centralization issues, byzantine fault tolerance, scalability, and overall security of the system. The potential is there for such earnings streams with volumes of over $5 trillion in the forex markets alone, but earnings could be subdued due to the fact that chains will likely devolve into a core settlement layer and secondary layers for processing, typically giving investors revenue streams from the core layer only. Full scaling on the primary layer is looking less and less likely as time goes on.

I have shown how most POS coins can be valued based on revenue streams from transactions fees. In my opinion, this model is much better than attempting to value these chains using the equation of exchange, MV=PQ.