Valuing Crypto Assets using a DCF Model

Currently there exist various models that attempt to capture a valuation framework for crypto-assets. In both traditional finance and crypto-assets, these models can be divided into relative and absolute valuation models. Relative models assume some baseline starting point which can be used as a comparison. In equities, the most commonly used relative valuation method is the price-to-earnings ratio (P/E). Investors use this to determine whether an asset is overpriced or underpriced by looking at the ratio compared to peer companies. A high P/E ratio (relative to peer companies) implies the market is paying a premium for the same earnings as a similar company. In various cases this premium may be warranted, such as earnings that are growing at a higher rate, or an innovative CEO who can push earnings higher in the long run. Many times, a high P/E is seen as the market overpaying for a company. As such, those shares appear less attractive relative to a peer company with a lower P/E multiple.

Relative Valuation Models

In the crypto-asset world, experts say the most similar ratio to the P/E multiple is the Network Value to Transactions multiple (NVT). Willy Woo and Chris Burniske first pioneered this metric in September of 2017 (Woo’s Forbes article: https://www.forbes.com/sites/wwoo/2017/09/29/is-bitcoin-in-a-bubble-check-the-nvt-ratio/). When a crypto-asset has a high NVT, it indicates that its network valuation is outstripping the value being transmitted on its payment network. Similar to equities, this can happen when the network is in high growth and the market is pricing this in, or when the asset price is in a fever pitch and is irrationally overvalued. Willy Woo described in his September 2017 Forbes article how Bitcoin had two major bubbles as shown by elevated NVT ratios (see graphic below).

As Willy describes, this metric cannot predict a bubble before it happens but after the asset sells off, the NVT ratio then spikes indicating that the asset was previously in a bubble. This is what we are currently witnessing (June 2018) as the NVT ratio is reaching a several year high as the Bitcoin price continues its free fall (see chart below). The reason being for this, is that a crypto-asset’s payment network is usually highly used during periods of hysteria as speculators move these tokens to different addresses and through other avenues for trading. This increase in on-chain transaction volume for Bitcoin is compounded as Bitcoin serves as the most used trading pair for the exchange of all other crypto-assets as well. These moves are then recorded as use of the payment network when in reality they are just being used by speculators to move assets around as trading volumes pick up during moments of elevated hysteria. Thus, during a fever pitch, while market prices are elevated, the number of transactions are likely more elevated. As such, a low NVT ratio does not necessarily mean the asset is cheap, it could be that the payment network is simply being overused by speculators (in December of 2017 the NVT ratio actually reached a localized low).

This, however, is not how a P/E multiple works. If we strictly interpreted the NVT ratio similar to a P/E, we would be inclined to say Bitcoin was undervalued during its market price peak in December of 2017 as its NVT ratio reached a localized low. Similarly, as the NVT ratio reaches a high during this market correction, we would be inclined to say Bitcoin is overvalued. Thus, it is clear that the NVT ratio is not the most attractive valuation tool as it fails to act similarly to the P/E ratio and also only becomes useful after a crash has already occured. I.e. the NVT ratio is telling us now that Bitcoin was overvalued in December/Jan of 2017.

Additionally, using relative valuation models (either across assets or relative in time within the same asset) in such a novel space as cryptocurrency is a difficult case to make. If I argue that Litecoin is a higher growth network than Bitcoin, and hence should trade at say a multiple of 2x what Bitcoin’s NVT ratio is, by definition we must first accept that Bitcoin is fairly priced. In a novel space in which there are experts who insist Bitcoin should trade at $1mm a token while others insist it is worthless, it becomes hard to base the valuation of other assets on the uncertain ratio of this asset. As such, relative valuation models are less useful in novel spaces with limited past historical multiples and prices. For that reason, cryptoasset valuations may be better served by utilizing absolute valuation metrics.

Absolute Valuation Models

The most commonly used absolute valuation metric currently for cryptocurrencies is the popular Equation of Exchange that is taught in any introductory economics course. Chris Burniske most popularly re-imagined the formula to fit a cryptocurrency model. While this method is the most appropriate that I can think of for currency tokens, it fails when applied to crypto-assets that are not used as a currency or medium of exchange. Many crypto-assets function more as equity instruments or bonds. Many crypto-assets now record a transaction fee for usage on their platform. This fee is then paid out to token holders who perform “work” on the network. The cost of this work is varied, but in many instances the work often requires as little as a high speed internet connection to run a simple process. For these fee incentivized networks, I believe it is most appropriate to use a DCF method to reach a valuation. Most crypto experts do not use DCF analyses to value crypto-assets, as they argue such assets do not generate a traditional cash flow. While true, these crypto-assets often pay fees out to token holders in Ether, or an ERC-20 token or Bitcoin. So while not a “cash” flow, we can call this an Ether flow.

Utilizing DCF analysis on Crypto Assets

A DCF is an absolute valuation metric that requires no relative multiples, but relies strictly on the cash flow paid back to holders of an asset (stock, bond, or token). As such, we can value these assets irrespective of other crypto-assets trading in the market. I would expect over the long term more crypto-assets to adopt a fee based incentivization model as it creates a much more attractive value capture for token holders than many other crypto models such as strictly governance or medium of exchange.

In our flagship report on Republic Protocol (REN), my partners and I at Blocktown Capital built out a DCF model to value the REN token. Republic Protocol is a decentralized dark pool platform in which nodes run a simple matching process and in turn for their work receive trading fees from the network. The primary incentive to own REN tokens, is to be able to stake these tokens to be eligible to run a node, and hence receive fees. We can treat these fees as a cash flow paid out to token holders. Below, we run through how a DCF model is then used to find a fair token price today for REN. Note that this type of model can be used for any fee paying network in which fees are paid out to token holders.

First, we need to assume certain growth rates of the platform which lead us to expected future cash flows. See our full report for why we used the below assumptions. Note that the assumptions below were for our bull case scenario for REN. In our full report, we also run through a base case scenario. Link here: Blocktown Capital REN Report & Valuation

The key aspect in the above table, is the estimate of cash to be paid to nodes projected over the next five years. Using the formula for DCF, PV = [CF1 / (1+r)1] + [CF2 / (1+r)2] + … + [CFn / (1+r)n] + TV, we can discount our cash flows back to present day to arrive at a Present Value (PV) for the REN network.

PV = Present cash flow value

CF1 = Cash flow at the end of year 1

CF2 = Cash flow at the end of year 2

CFn = Cash flow at n specified year

r = Discount or required rate of return

TV = Terminal Value

In any DCF model, we must discount our cash flows back to the present day as a dollar next year is not worth the same as a dollar today. One key point to note, is that all expected cash flows are not treated equally. We must build in a risk measure. The more uncertain an expected cash flow is, the more we should discount it. Presently, there is no appropriate risk rate in the digital asset marketplace that can be used as a benchmark. As such, in our report we borrowed from other financial markets. In Series A equity funding rounds for venture capital, the generally accepted rate is between 30–50% per annum. As a digital asset platform with certain technological components that remain unproven, Republic Protocol has a relatively high risk profile so we settled on a similar discount rate of 40%.

Lastly, in our DCF model, we need to include those fees paid after the five year period. We modelled the REN platform to last into perpetuity, as is appropriate in the equity markets. To achieve this, we used the Gordon Growth Method to find a Terminal Value, TV. We conservatively projected out a sustained growth rate of 2% (g = 0.02), inline with mature company estimates and developed countries’ GDP estimates. Using the Gordon Growth formula below:

TV = [year 5 cash flow * (1+g)/(r-g)]

TV = 1,506,093,750*(1.02)/(0.4–0.02) = 4,042,672,697

Adding up our yearly cash flows and terminal value, we arrive at the below summation:

0 + 42,857,142 + 117,091,836 + 214,012,390 + 280,034,686 + 4,042,672,697 = 4,696,668,753

Finally, by dividing our total network cash value (4,696,668,753) by the number of circulating tokens (519,094,022), we arrive at a dollar price per REN of 9.05 USD. That is, with our assumptions and best case scenario forecasts, REN should be valued at 9.05 USD per token.

In conclusion, I expect absolute valuation metrics will become more commonly used when valuing crypto-assets and I also expect DCF analysis to become more utilized in the space. Currently, very few experts have looked to use a DCF to value a crypto network, in part because the majority of crypto assets still do not have fee incentivized networks and hence no cash flows. Given the attractiveness of such fee based models, I anticipate more new tokens will be structured to incorporate a stream of cash flow back to the token holders in order to create value for the token asset. My hope is that this post will shed light on how to value those crypto-assets which incorporate a cash flow.

Thanks to Joseph Todaro and James Todaro for contributing to this piece.