Just like stocks, cryptoassets do have some valuation metrics investors rely on. For stocks, they would be P/E, P/B, EV/EBITDA, etc.; for cryptoassets, the most popular valuation metrics for Smart Contract Platforms(L1/L2) are on-chain metrics such as Market Cap (MC) to Total Value Locked (TVL) or MC to Daily Transactions (TX). Intuitively those metrics could be nice indicators as they have a firm logical backdrop: We use Smart Contract Platforms to make transactions, so more transaction(TX) = more usage = more liquidity(TVL) = more value.
CTs, especially $FTM maxi CTs, bring those metrics to assert how $FTM is undervalued compared to other L1 competitors, and how $FTM price would make a miraculous rise if it catches up with them. Looks fine, but some might wonder: “Do the metrics really mean anything?” So here I am to provide a quick analysis on the crypto valuation metrics, namely MC/TVL and MC/TX.
Now the above is the famous graph from @TaschaLabs illustrating the strong relationship between # of daily transactions and $ETH price.
If there exists a single equilibrium MC/TX ratio that encompasses all blockchains, we might be able to say with confidence that $A token is certainly undervalued.
Research Method
TVL refers to the amount of liquidity locked up in a certain project(in this case L1 projects such as Ethereum $ETH), including Liquidity Pools on DEXes, Collaterals on lending protocols, etc. A high TVL represents that the L1 proejct offers strong DeFi services compared to its counterparts, hence it has more value.
All on-chain activities needs to push through transactions, so if an L1 project has higher daily transactions(TX), it should have more value as more people are using that blockchain.
For TVL data, I used DeFiLlama; for daily transactions, I used explorer statistics for each L1 chain, though not all L1 chains offer easily accessible statistics(ex. $SOL) so I left out some chains. For Market Cap data, I used Messari.
L1 tokens for the analysis includes: $ADA, $ALGO, $ATOM, $AVAX, $BNB, $CELO, $DOT, $EOS, $ETH, $FTM, $KLAY, $LUNA, $MATIC, $NEAR, $ONE, $SOL, $STX, $TRX for TVL analysis, and $AVAX, $BNB, $ETH, $FTM, $MATIC, $LUNA for TX analysis.
MC/TVL Ratio
The above graph shows MC/TVL for each chain over 540 days, which is the maximum period offered in DeFiLlama. Not all chains offer full dataset over the period(in DeFiLlama), so there are gaps here and there. Considering the y-axis is expressed in a log scale, the actual valuation differences should be much larger than they look in the graph.
What we can see from the graph is that, ecosystems in their infancy ($AVAX around index 200) has tremendous MC compared to their TVL, but they do decline in MC/TVL metrics as they mature.
If we scan the summary statistics, most of L1 chains have the median MC/TVL ratio of <10, but there are few projects showcasing tremendous ratio of >10000. $ADA and $DOT are ranked #6 and #11 respectively in CMC, but they are often accused of offering not that many services(at least for now) thus we expect them to be overvalued by MC/TVL metrics.
The above graph shows 10d MA MC/TVL ratio, and we scaled the y axis to (0, 20) to crowd out outliers. Except for few projects ($ADA 6000, $ALGO 58, $DOT 15000, $EOS 34, $NEAR 66), all L1 chains successfully managed to bring MC/TVL ratio down to <5 even if they started off from >80000000 ($AVAX).
It seems there are some evidences of convergence between TVL and MC. So, would a relatively undervalued $FTM (MC/TVL @ 0.7) catch up on a relatively overvalued $SOL (MC/TVL @ 4.5)? Let’s take a closer look.
Although the majority of chains are clustered around ~2.5, each chain seems to just fluctuate around its median value once it reaches a stable zone. Maybe 540 days was simply to short to price in all the “fundamental values.”
So far, what we can tell from the data is that 1) absurdly high MC/TVL does get adjusted 2) most chains have MC/TVL ~2.5. As for undervalued chains retracing back to 2.5, we cannot conclude.
MC/TX Ratio
For daily TX, the converging trend is less evident. But we can at least witness that except for $ETH, all chains are clustered at the bottom.
Now $ETH somehow skyrocketed above 180000, as its daily transaction increase did not catch up with its price appreciation. All other chains have MC/TX ratio of <75000, but still there are large discrepancies in value. $BNB, $MATIC, $FTM has MC/TX ratio of <15000, $AVAX has 30000, $LUNA is at 50000.
We are short of data to conclude anything on MC/TX ratio; but if we outstretch a bit, maybe we can say that an MC/TX ratio of >100K might imply a slight overvaluation.
I wished to see the same graph for $SOL as well, but as SolScan does not have the statistics feature yet, we will revisit the graph once again when other chains have made their statistics available.
Limitations
The limitations of this study are 1) short time frame 2) no adjustment for L1-specific attributes (idiosyncratic Chain features) 3) no implication for causation.
Perhaps if we were to look at a longer time frame, such as 3 to 5 years, maybe we should look all MC/TVL ratio converging to one golden ratio. Perhaps L1-specific attributes hampers the convergence, so we might not observe any convergence at all. Instead, each chain might have its own “steady state” at different levels. Perhaps all this was just a coincidence, and actually TVL~MC might have no relationship at all.
But to have a some sort of framework, though unreliable, is a good way to have conviction and to make investment decisions.
Conclusion
The bottom line is that, the valuation metrics we currently use in CT might be able to cherry-pick extremely overvalued projects(OR projects that have a tremendous expectation from investors), but we would not be able to tell whether an undervalued project would finally find its proper value or not.
If you still wish to use the MC/TVL as your core valuation metric, 2.5 should be the magic number for you to decide whether the project is undervalued. For other metrics, such as daily TX, unique addresses, or active addresses, factors unique to each protocol seem to play an important role in deciding the ratio, so although the daily TX growth and the MC growth be correlated, we cannot say that two different protocols would converge in MC/TX metrics.
