Understanding New Capital Forms: Using a Three-Layer Model to Analyze the DeFi Protocol Token Value Flow
Editor's Note: This article comes fromBabbitt Information (ID: bitcoin8btc)Babbitt Information (ID: bitcoin8btc)
Babbitt Information (ID: bitcoin8btc)
, by Jon Itzler, a researcher at venture capital firm Accomplice, translated by: Overnight's Porridge, published with permission.
image description
Figure: Trying to infer the composition of the protocol token
Over the past 12 years, we have become accustomed to the term protocol tokens, but there is still a great deal of confusion surrounding what constitutes it fundamentally.
My hope is that the vast majority of protocol tokens will be productive assets. It simply means that, like bonds, businesses, farms, etc., protocol tokens confer cash flow income rights.
As it stands, the agreed cash flows are still relatively small, but the intrinsic value of a productive asset is the sum of all future cash it can generate, discounted at the current price.
Tokens can also grant access to different forms of utility related to non-cash flows, and can even be socially accepted as commodity money. Sometimes these additional demand pressures are quantifiable, but often, we can only reason about them qualitatively. For widely integrated DeFi protocols, how much can governance be worth during their migration to Layer 2 projects, or during fork selection?
In combination, the protocol token has intrinsic value as a claim or potential claim on future cash flow, as well as a structural demand premium due to grant utility, currency attributes, etc.
first level title
A simple way to think about the value stream of an agreement is to use a three-tier model.
secondary title
forbitcoinLayer 1: The protocol facilitates a certain amount of total amount of money
bitcoin
Or a layer network like Ethereum, this is the total value of transactions transferred, for a DEX this is transaction volume, and for a lending protocol this is the total amount of loans outstanding.
It is simply a measure of the total dollar value of demand for a particular service.
secondary title
Layer 2: Protocols capture a portion of the total facilitation volume as a revenue stream by charging users a cost, which I refer to as protocol revenue.
Uniswap's 0.3% protocol fee is proportional to the transaction volume, while the fee generated by 0x is determined by a function of transaction frequency and Ethereum gas price, which is basically independent of transaction volume.
There can be this disconnect between total transaction volume and fee revenue, which is why we cannot rely on protocol revenue alone to see the full picture. For example, a DEX with a monthly trading volume of $100 million may not charge users transaction fees. Obviously, its potential to generate cash flow in the future will be greater than zero.
secondary title
Layer 3: The protocol divides the total revenue stream and distributes it to various groups of participants in the market.
Cooperatives are often compared to protocols because the share of protocol revenue is allocated directly to market participants, rather than to a single profit-maximizing entity. This is why there is no easy way to define a blanket "benefit" metric - the protocol's stakeholders are more diverse than the business's shareholders.
first level title
Agreement income distribution
It makes sense to classify four types of participants who distribute protocol revenue:
Any supply-side participant (LP, lender, miner, keeper/liquidator);
Any demand-side participant (DSR depositor, Nexus Mutual claimant);
Supply-side participants (PoS validators, 0x MM, Keep signers) who own tokens;
A protocol token is a productive asset if the revenue share is distributed to one of the last two types of participants. That is, the token itself, or the token combined with some service, grants the holder a claim on future cash flows.
It is more common today that the protocol will distribute 100% of the revenue to "arbitrary" supply-side participants, where no native token is required to provide the service and the revenue share is not held/staked against the service provider Native tokens are distributed. In DeFi protocols, this service is usually some form of liquidity provision. For example, LPs in Balancer and Uniswap V2 distribute all protocol revenue.
Most PoW chains have a similar dynamic, where all fee revenue is distributed to a supply side that does not own the tokens: the miners. We can think of Bitcoin miners as “staking” coupons for future BTC in the form of mining hardware, which is why BTC would not be classified as a productive asset. The mining hardware gets 100% of the protocol revenue, not the tokens themselves.
image description
For tokens that are productive assets, the most common model is to pay a portion of protocol revenue as dividends to token owners. A buy & burn mechanism, like a stock buyback, is the economic equivalent of taking a dividend and using it to buy more stock. And when tokens are burned along with protocol revenue, effective dividends will be paid out to all token holders.
Staking rewards are another form of dividends. Unlike the buy&burn model, this dividend is only paid to token holders, and participants also provide an additional supply-side service: locking capital by staking native assets.
The implied rate of return for buy & burn dividend capital is protocol revenue paid / total supply * price, but in the case of staking rewards, the calculation is converted to protocol revenue paid / pledged supply * price.
Dynamic reallocation of future cash flow
The Ethereum 2.0 economy, which does not yet exist, can help us illustrate the situation where both dividends exist at the same time. It is important to note that a subsidy such as a block reward is not a productive cash flow, rather it represents a dilution of future cash flow from existing holders to recipients of the subsidy.
In the specific case of the first layer, the original assets are both productive and maintain a large currency premium, so the issuance is between seigniorage and dilution.EIP-1559image description
Figure: Ethereum 2.0 cash flow distribution
Suppose we successfully implement
and Ethereum 2.0, then Ether (ETH) will transition to a productive asset (this more broadly applies to all native assets of PoS chains). Validators, and stakers delegating validators, are members of the aforementioned set of participants, where token ownership is necessary to perform supply-side services.
Fee income from staking rewards (“tips”) is paid out as dividends, and due to delegation fees, delegators may receive slightly less than validators. In EIP-1559, because of the existence of BASEFEE, a portion of ETH will be burned, which is another form of dividend, and it is paid to all token holders. While validators and delegated stakers effectively "stack" these two forms of bonuses, they receive:[1]
Burn fee income, where the return on capital is proportional to the ownership percentage of the total token supply;
Fee income rewarded to stakers, where capital gains are proportional to ownership of the total tokens staked;
Verifier;
commissioned pledger;
token holders;
first level title[2]
Dividend Model Problems
There are two problems with deterministically making protocol revenue a dividend to token owners or supply-side participants who own tokens. Buffett explains the first one, which applies to the buy & burn model:
Simply put, buy & burn makes sense when assets are cheap, but not when assets are expensive. Using protocol revenue to buy back a token that just skyrocketed 10,000% in the last month may not be the most efficient way to distribute surplus cash flow.
Historically, Maker has been short convex, diluting owners (MKR flops) when its tokens are cheap, and increasing their share of future cash flows (MKR flips) when they are more expensive.
Another problem with paying dividends is that you lose the natural compounding effect that a business exhibits in terms of its ability to reinvest retained earnings.[3]
first level title
more optimized model
“A properly designed and efficient distribution of fee streams can further entrench network effects by giving users direct economic incentives to generate more defensibility, which in turn strengthens the viability of fee streams.”
The optimal token model is:
Incentivizes all protocol participants to fulfill their stated roles (i.e. supply-side revenue) while minimizing user costs;
Incentivizes all protocol participants to own a single capital instrument as a unified upside incentive;
Maximize the effectiveness of retained cash flow allocations after completion of a, b, and c;
As can be seen from the Ethereum 2.0 chart above, most protocols have not yet addressed the issue of sustainable developer funding (which is a crucial set of supply-side players), but iterations on the last two points could serve as a solution .
We see another model — Compound’s reserves, where the protocol retains surplus income as a manageable on-chain balance sheet.
image description
Chart: Compound Reserve USD by Assets
This forms Compound’s moat, allowing a feedback loop to propagate forward, ultimately resulting in more reserves being created. Through governance, COMP has granted an effective claim to the reserve fund, so the question becomes: why pay dividends when you can let retained cash flow accumulate to an on-chain balance sheet and continue to compound?
The reserve is first and foremost an in-protocol insurance buffer, but in theory it could be used for other things. Governance can determine some margin of safety to grab excess reserves and auction them off as buy & burn dividends when prices are at favorable lows.
The reserve also acts as a protocol treasury, computationally rebalanced in order of importance to active contributors, voters, and delegators.
first level title
Dilution as funding & cash flow expansion[4]Protocols need to incentivize broader protocol offerings, including development, governance roles, and insurance. Allocation of these needs can well promote long-term returns on capital compared to paying dividends.
Issuing MKR via flop auction is a model where dilution is used as effectively limitless insurance backstop
. A Tezos-style issuance that rewards protocol upgrade contributors is another potential case for dilution as funding. It is conceivable that Compound governance would add something similar if they thought it would be beneficial.
image description[5]。
Figure: Compound Daily Protocol Revenue
Liquidity mining as a growth strategy reminds me of the story of Teledyne CEO Henry Singleton in the 60s and 70s. During the period of 8 years, when Teledyne stock price was high, he bought 130 companies, all but 2 of which were done with his own stock, and here is what he got during that period s return
Figure: Singleton's returns during the release frenzy
Then, Singleton abruptly changed direction. Over the next 12 years, he bought back more than 90% of his outstanding shares at cheap multiples. Similar to expanding sales through acquisitions, liquidity mining uses dilution to buy volume on the demand side.
In the long run, protocols can be issued through subsidies, bringing future growth into the present, and then ultimately capturing a portion of the expanded value stream.
first level title
Reflections on the present and the future
Bullish valuations can indeed only be attributed to speculative premiums, but cash flows have also historically increased by orders of magnitude during this time. The last cycle was about Bitcoin and Ethereum fee income, and there is evidence that this will repeat itself, but we have also seen massive increases in DEX transaction volume, lending volume, etc.
One thing to be wary of is forward-looking "P/E" or "P/S" ratios.
They serve as useful indicators of price versus current cash flow, but not as a signal that historical prices are cheap. According to the timeline, the Ether price/forward fee-to-income ratio bottomed out when ETH was most expensive. Price/Trailing 365-day fee income is often a more reasonable cheap signal.
image description
Chart: Ethereum Price/Fee Income Ratio - Trailing 365 Days
We also found that due to low liquidity and the need to chase these yields, real-time "collateral premiums" are reflected in assets, examples include all YAM mineable assets, etc. This is yet another example of tokens that are very different from traditional forms of capital such as shares.
Relevant information:
1.Charlie Noyes on computational equity
2.Buybackquotes
3. Crypto’s Business Model is Familiar. What Isn’t is Who Benefits
4.Tom Schmidt on the two token schools
5.The Outsiders: Eight Unconventional CEOs and Their Radically Rational Blueprint for Success
