November 20, 2018 . 12 min read
Crypto Chat #63
As always, the following is for educational purposes only. This is not investment advice.
Fundamental Value in a Sea of Speculation
The continued, aggressive market recession has many investors questioning the fundamental value underpinning the crypto asset class.
Several groups and individuals have sought to look beyond market activity, providing alternative, crypto asset-specific metrics, which aim to more accurately measure the fundamental strength of these underlying networks.
The latest offering comes from blockchain media behemoth, CoinDesk, via their ‘Crypto-Economics Explorer’, which, unfortunately, leaves much to be desired.
In addition to the unnecessary appropriation of the term ‘ crypto-economics’, which has a previously well-established relationship with the intersection of cryptography and economic incentives, the CoinDesk product a) features notoriously spoofable metrics — Transaction Count, Exchange Volume, Twitter Followers etc.— b) does not allow for comparisons over time, and c) employs murky methodology across various verticals.
And yet, despite the rather amateur quality of the ‘explorer’, we should applaud CoinDesk for their willingness to look beyond the ever-hegemonic ‘network value’ indicator and contribute to the emerging field of alternative metrics. Indeed, by their very nature, public blockchain networks provide a wealth of transparent data for analysts to play with: by leveraging this access, we can look beyond the wholly speculative market dynamics and begin to measure the performance and utility afforded to users by these emerging networks.
Several weeks ago I introduced the Fee Ratio Multiple (FRM), which seeks to measure the transaction fee revenue growth required for disinflationary Proof of Work chains to maintain existing security levels once block rewards are fully phased out. A high FRM suggests that a network will have to maintain inflationary block rewards in order to sustain security, whilst a low FRM suggests that a network can feasibly maintain existing security levels in the future while simultaneously achieving low inflation. The latter outcome is arguably more desirable for protection against inflationary government-controlled currencies. Check out the Bloxy site for a live representation of FRM for the Ethereum network.
FRM is useful as a means of predicting future network security and the appropriateness of an asset as a Store of Value. For today’s newsletter, I want to focus on present network demand and its relationship with network value.
There are numerous ways of measuring demand in these networks: to my knowledge analysts have yet to converge around a single metric and, in all likelihood, will forever contest the validity of a single data point as a source of demand. The most obvious indicator would be spot price, which has fallen considerably over the last 10 months. But, like CoinDesk’s ‘Social Media’ indicators, spot price can often be misleading, with price activity and network usage producing disparate images of the network’s health.
Fees as Indicator of Demand
After some consideration, I believe that aggregate transaction fees in dollar terms across different time periods can provide a semi-accurate representation of present day network demand.
Fees are more resistant to spoofing than alternative metrics like ‘Transaction Count’ and ‘Transaction Volume’, the former susceptible to low fee transaction spamming, the latter susceptible to ‘wash transactions’ by wealthy investors. Conversely, the only way to significantly boost a network’s fee revenue is to spend significant capital on fees.
Fees also act as near-direct proxy for demand to use the crypto asset as a Medium of Exchange, ‘gas’ for Decentralized Applications, or a hybrid of the two.
Network Value to Aggregate Fee Ratio (NVAFR)
By measuring aggregate fees over varying time periods and comparing it to network value, we can ascertain the extent to which a network is over or undervalued relative to demand for the network as an alternative payment system and/or decentralized application execution system. This might be somewhat analogous to the Price/Sales ratio used in the stock market.
The Network Value to Aggregate Fee Ratio (NVAFR) is calculated as Price * Circulating Supply ($) / Sum of Daily Fees ($) Over T.
T can be adjusted depending on time frame preference – a higher time frame T, such as 365 days, will provide a less volatile image than a low timeframe T, like 7 days. Each time frame comes with trade offs — I personally tend to lean towards higher time frames as they allow you to observe the ‘bigger picture’ and closely reflect the time frames of metrics employed in capital markets.
A couple things to note:
1. Aggregate Fee figures should not have any direct impact on the price of Proof of Work-based crypto assets as asset ownership does not confer any claim to transaction fee revenue. However, in Proof of Stake systems, asset holders do have a claim to transaction fee revenue, and so I imagine that this metric will likely play a significant role in the future evaluation of these assets. Nevertheless, even for PoW networks, NVAFR should provide meaningful insight into the health and activity of these ecosystems.
2. NVAFR is still highly experimental: there are likely numerous ways to improve methodology, and numerous opinions as to which time frames should be appreciated and depreciated. I look forward to further experimentation and improvement of this metric.
As for the results:
Perhaps tellingly, a logarithmic scale is required in order to represent the various assets – Bitcoin (BTC), Ether (ETH), Litecoin (LTC), Monero (XMR), Decred (DCR), Bitcoin Cash (BCH), Ripple (XRP) – all on the same chart.
As of 11/13/18, LTC, XMR, DCR, BCH, and XRP had 365 Day NVAFR’s of 688x, 285x, 2308x, 7845x, and 17942x.
This means that the network value of LTC is a 688x multiple of the dollar demand for use of the network as a settlement layer. And that is one of the least egregious examples: XRP, which has somehow appreciated over 80% vs. USD since mid-September 2018, has 365 Day NVAFR or 17,942x. That XRP is overvalued is by no means a controversial statement, but the sheer extent of the disparity between valuation and underlying demand should surprise even the most ardent critics and irk even the most zealous supporters.
BTC & ETH NVAFR
Where does NVAFR leave Bitcoin and Ether, both of which are the only assets measured to have sub-200 365 Day NVAF ratios?
The positive news is that both networks have seen reductions in NVAFR in the past year, suggesting that their valuations are now closer in line with their fundamental performance. Ether’s 365 Day NVAFR reduction is more pronounced, falling by close to 93% year to date, although Bitcoin has still made material improvements, with a 70% reduction over the same period. Compare this to Decred, which has seen its 365 NVAFR 10x over the same period!
The bad news? In absolute terms these NVAFRs are still very high. To put these ratios into (an admittedly imperfect) context, a Price/Sales Ratio above 4 is commonly considered unfavourable. In order to justify today’s valuations, cumulative fees have to increase dramatically or network value has to take a significant haircut to more accurately fall in line with today’s usage.
The alternative, and likely more accurate, news? Very few, if any, investors are using these metrics to value crypto asset networks. If we were to take Bitcoin’s NVAFR as a benchmark for the wider market, XRP would be valued at a just over $200m, or 1% of its present valuation, DCR at $26m, or 7% of its present valuation etc. etc. Similarly, taking BTC as a benchmark, ETH should be valued at over $32bn, or 200% its present valuation. This may be because investors are pricing these networks based on future expected utility, in which case the data would suggest they are very optimistic regarding future gains, or because their investment strategies are weighted towards qualitative data over quantitative data.
Nevertheless, in my overly-biased opinion, NVAFR provides more immediate insight into the existing state of affairs than most metrics. Unfortunately, I have not yet had the time to settle upon the most optimal time frame, nor to try and surmise what levels of NVAFR coincide with overbought/oversold territory or the various real-world factors that lead to lower or higher ratios. However, it should still be overwhelmingly clear that outside of Bitcoin and Ethereum — both of which are also likely overvalued — the state of the leading crypto assets by network value is precarious.
I look forward to hearing your thoughts and feedback, and encourage anyone interested in NVAFR to take a look at the data, available for download here, and improve upon my methodology. In particular, I am interested to see further exploration of:
a) the relationship between an asset's value and network activity where asset ownership does not confer claims to transaction fee revenue
b) the aggregate ETH fees associated with individual ERC20 transfers
c) the effect of using Simple Moving Average of Exponential Moving Average for calculating Aggregate Fees
d) analogous metrics for 'fee-less' networks like EOS
For more alternative metrics, I recommend reviewing Nathaniel Whittemore and Clay Collin’s thorough survey of emerging indicators.
Until next time,