How to value transaction medium tokens?
TechFlow Selected TechFlow Selected
How to value transaction medium tokens?
In the long run, having a foundational framework model is helpful for understanding whether each token is at a high or low position.
Navigating Web3 tides with focused insightsAuthor: Vitalik Buterin
Translation: Sanjin, BlueFox Notes
Preface: Currently, there is no particularly good model for token valuation due to the numerous influencing factors and the manipulable nature of markets. However, in the long run, having a foundational framework helps understand whether each token is at a high or low level.
In many token sale projects, a model known as the "network medium-of-exchange token" has become popular. Let's describe the typical scenario for this type of token. We developers build a network that allows you to do some new and cool things.
This network is a shared economy system: it consists purely of a series of sellers and buyers, where sellers provide resources under a certain protocol, and buyers purchase services—all participants come from the community. However, buying and selling within this network must be conducted using the token we are selling, which is why the token holds value.
If the developers themselves act as sellers, this arrangement is reasonable and normal, essentially similar to a "product-launch-style (Kickstarter-style)" sale. Economically speaking, the token is actually backed by the services provided by the developers.
We can see more details through a simple economic model. Suppose N people anticipate a future product worth $x, believing the developers can deliver it. The developers conduct a sale, raising N contributions, each worth $w < x, thus raising a total of $N * w. The developers build the product and distribute it to each buyer.
In the end, buyers are happy, and developers are satisfied. No one feels they made an avoidable mistake during the process, and everyone’s expectations are met. This economic model is clearly stable.
Now, let’s look at the case of a "medium-of-exchange" token. N people anticipate a future product worth $x on a decentralized network; the product will be offered at a price of $w < x. They purchase tokens worth $w during the sale.
The developers build the network. Some sellers join, offering products on the network at a price of $w. Buyers use their tokens to purchase these products, spending tokens worth $w and receiving $x in value. Sellers spend $v < w in resources and effort to produce the product and now receive tokens worth $w.
Note here that the process does not end—it never truly ends; it requires continuous inflows of buyers and sellers to maintain its value. Strictly speaking, these flows don't need to be endless; if in each round there's a chance that v/w < 1 proceeds to the next round, the model can continue functioning because even though some individuals may eventually be defrauded, any single participant’s risk of being the unlucky one is lower than the benefit gained from participation.
It's entirely possible for the token to depreciate in each round, multiplied by a factor f where v/w < f < 1, until eventually reaching zero, yet it remains in circulation allowing anyone interested to participate.
Thus, this model is theoretically viable, but you’ll find it significantly more complex and fragile than the simple "developer-as-seller" model.
Traditional macroeconomics offers a simple formula for valuing a medium of exchange:
MV = PT
Where:
• M is the total money supply, i.e., the total token supply
• V is the "velocity of money," i.e., the average number of times a token changes hands per day
• P is the "price level." This refers to the price of goods and services measured in tokens, so it's actually the reciprocal of the currency's price
• T is the transaction volume: the daily economic value transacted
This can be proven with a simple equation: If there are N tokens, and each token changes hands M times per day, then the daily economic value transacted is M*N. If this represents $T in economic value, then the price per token is T/(M*N), so the "price level" is its reciprocal, M*N/T.
For easier analysis, we modify two variables:
• Use "H" to represent 1/V, indicating how long users hold tokens before using them in transactions
• Use "C" to represent 1/P, indicating the price of the currency (think C = cost)
Now we get:
M/H = T/C
MC = TH
The left-hand term is simply market capitalization. The right-hand term is the daily economic value transacted multiplied by the average time users hold tokens before spending them.
This is a steady-state model assuming a constant number of users. In reality, however, user numbers fluctuate, prices change, and holding times vary, all affecting price.
Let’s reconsider the economic impact on users. What do users lose when using an appcoin instead of regular ether (or bitcoin, dollars)?
The simplest way to express this is: such systems impose a "hidden fee" on users—the time they hold these tokens instead of equivalent currencies they’d prefer to hold.
This fee includes many components: cognitive costs, transaction fees, service charges, and other minor items. A particularly important component is expected returns. If users expect the appcoin to appreciate only 1% annually while alternative assets rise 3%, and they hold $20 worth of tokens for 5 days, the expected loss is roughly $20 * 2% * 5 / 365 = $0.0054.
A direct consequence of this insight is that appcoins are a multi-equilibrium game. If the appcoin appreciates 2% annually, the loss drops to $0.0027, effectively making the app’s (or most of it) "de facto fee" twice as cheap, attracting more users and increasing its appreciation. If the appcoin declines 10% annually, the "de facto fee" rises to $0.035, driving away users and accelerating its decline.
This increases opportunities for market manipulation, because manipulators aren’t just spending money trying to break a single equilibrium—they could successfully nudge the currency from one equilibrium to another and profit from correctly "predicting" (or causing) this shift.
It also implies significant path dependence with lasting effects. The epic battle over which Bitcoin fork gets to be called Bitcoin is a particularly high-profile example.
Even more importantly, the conclusion is that an appcoin’s market cap critically depends on holding time H. If someone builds a highly efficient exchange enabling users to instantly buy appcoins and immediately use them in-app, and sellers to instantly cash out, the market cap would plummet.
If the currency is stable or prospects are optimistic, this might not matter—users see little downside to holding such tokens versus others (in other words, near-zero "de facto fee"). But if prospects worsen, such high-performance exchanges accelerate its death.
You might assume exchanges are naturally inefficient—requiring account creation, login, deposits, waiting for 36 confirmations, trading, then withdrawal—but in fact, highly efficient exchanges are coming. There’s already a discussion group designing fully autonomous, synchronous on-chain trading mechanisms that can convert token A into token B in a single transaction, possibly even doing further actions with token B. Many other platforms are also in development.
All this shows that relying solely on medium-of-exchange competition to support a token’s value is extremely fragile—its appeal lies merely in the illusion of creating money from nothing. Protocol tokens using this model may remain stable for a time due to temporary equilibria driven by irrationality and negligible hidden fees, but they always carry an inevitable risk of collapse.
So what are the alternatives? A simple alternative is EtherDelta’s approach: applications simply collect fees directly from the interface. A common criticism is: Can’t others just fork the interface and take the fees? Rebuttal: Others could also fork the interface and use ETH, BTC, DOGE, or any other user-preferred currency instead of your protocol token.
This leads to more complex arguments since "clone" versions must compete with the official version’s network effects, but one can easily create an official paid client that refuses interaction with unpaid clients—this kind of network-effect-based enforcement resembles how VAT is enforced in Europe and elsewhere.
Official-client buyers won’t interact with unofficial-client sellers, and official-client sellers won’t interact with unofficial-client buyers, so a large number of users must simultaneously switch to the "clone" client to successfully pay with Dogecoin. It’s not perfectly robust, but it’s certainly as good as creating a new protocol token.
If developers need to raise funds upfront to kickstart project development, they can issue a token and use all collected fees to buy back and burn some tokens—making the token backed by its expected future value and the fees soon to be consumed in the system.
This design can also evolve into a more direct utility token, requiring users to pay with the utility token, automatically purchasing it from an exchange via the interface if users don’t already hold it.
Critically, for tokens with stable value, it’s highly beneficial if the token supply has a sink—a mechanism where tokens are actually destroyed, so the total supply gradually decreases over time. This way, user payments become more transparent, and the protocol token’s valuation becomes clearer and more straightforward, rather than relying on calculating highly variable and difficult-to-estimate "de facto fees."
Join TechFlow official community to stay tuned
Telegram:https://t.me/TechFlowDaily
X (Twitter):https://x.com/TechFlowPost
X (Twitter) EN:https://x.com/BlockFlow_News
@VitalikButerin
