🧠Grip: DataAlways's Flows Analysis
Reading through dataalways’s newsletter can be daunting sometimes considering, their subject matter is not mine. This is my attempt at understanding the work done on the Flows Analysis series. Flow analysis is not my expertise and I was merely curious about it considering I had some ideas about analyzing on-chain liquidity flows. This is just a hash of what I’ve learned reading the articles on it written by dataalways.
What is Flow-Based Analysis?
As stated by dataalways, the Flows-based model is meant to be an attempt to define long-term demand based on the thesis that flows analysis is dependent on the net monetary demand inflows and net coin supply inflows - the latter being issuance in a certain time period.
Price/Coin ~ (Net Monetary Demand Inflows) / (Net Coin Supply Inflows)
There seems to be two core concepts behind the flows-based approach:
Coin Supply Inflows
Net Demand Inflows
Net Demand inflows can be further divided into: (a) Absorbed Supply and (b) Implied Demand.
What is Supply Inflows?
Coin Supply Inflows is essentially the issuance to miners - in the case of Bitcoin and Ethereum. These are the two main public blockchains on the Proof of Work (PoW) consensus mechanism to ensure security of the blockchain. In other words, whatever the issuance method is to miners or in the case of Proof of Stake - validators, coin supply inflows can be calculated on-chain.
What is Absorbed Supply?
Absorbed supply (AS) is the net inflows into cold storage in a particular time period. It implies that an owner moving into cold storage has no intention on selling their crypto asset in that particular time frame. In the case of Ethereum, this includes assets locked in smart contracts.
At high time frames (HTFs) this could imply owners prefer to hold rather than spend but the absorbed supply metric coupled with miner revenue (MR) gives a better understanding of the macro market. AS tends toward negative during price bubbles and when their in capitulation events in the market.
In the case of Ethereum:
When AS > Supply Inflows it may imply a decrease in holders and ETH may be underpriced.
When AS is positive but lower than supply inflows, it may imply a greater proportion of short-term traders/speculators possibly indicating ETH being overpriced.
When AS is negative, it may indicate a possible capitualation event and holders wanting to exit.
What is Implied Demand?
Implied Demand (ID) is multiplying absorbed supply by the daily closing U.S. dollar price.
Implied Demand = (Absorbed Supply) * (BTC/USD)Flows-based Arguments
The argument being made is that a flows-based approach accounts for the changing issuance schemas of blockchains in the future as we focus on flows and can define changing issuances within the coin supply inflows aspects of the model. For example, Bitcoin’s issuance currently this is both using subsidy dependent (mint on creating a block) and security spend dependent (transaction fees) schemes but is to change to just a transaction fee based (security spend model) in the future after all 21 million are issued. This change can still be captured in the supply inflow aspect of the flows-based model. And according to dataalways, the model will tilt bearish in the future as transaction fees become the dominant source of flows.
The flows-based analysis is only designed to look at the long term behaviour of the market. For more shorter-term analysis, methods such as sentiment analysis and other on-chain metrics may be useful. Datalways states that the Flows-based analysis fails like every pricing models due to autocorrelation errors where errors in one time period (in a time series) are transferred into the next time period leading to a snowballing effect in errors.
Thoughts
Every cycle brings in new models. It is only a facet of learning. I remember PlanB’s about five years ago now and the Rainbow model. My only experience doing this was for a client in 2016 where they wanted a 1% through to 10% allocation of Bitcoin to their portfolio using Markowitz portfolio theory and also providing a correlation analysis of Bitcoin to all other asset classes.
I think the Flow-Based model is interesting but I also think there could be more granularity in explaining demand. Just like the models by PlanB, SquishChaos and others, the assumption made is that there is always going to be demand - and not just any kind of demand but very high demand. In a world where anyone can create an economic system (granted the challenge is adoption to create network effects) - and newer capital wanting the same appreciation as those that were early to Bitcoin and Ethereum - have the option of a newer and younger (in terms of time in crypto) community. You see this today in Terra, Solana, Avalanche and Fantom just like we saw this in Ethereum five years ago but this time with the learnings of Ethereum and Bitcoin. In other words - why would I buy your bags when I can create my own.
And this doesn’t account for the demand based on the NFT communities arising. Demand may not come from what layer 1s bring low fees and a good user experience but rather what communities exist via their NFT pfps. Not viewing demand by accounting for this perspective may pin point to an early adopter/holder’s fallacy among the Ethereum and Bitcoin community.