A token's value is driven by whatever it captures from the platform it runs on. The three canonical structures are currency-style (value follows total transaction volume), coupon-style (value follows platform revenue), and equity-style (value follows platform profit). Each one trades differently against platform risk, investor risk, and volatility, and the right choice depends on how much revenue or upside the platform is willing to forfeit to back the token.
The value of a token is not an abstraction. It follows whatever cash flow the token captures from the platform it runs on, and the capture mechanism is decided by the token's structure, not by the marketing around it. Three canonical structures are worth knowing: currency-style, coupon-style, and equity-style. Each one maps to a different value driver, and each one trades differently against platform risk, investor risk, and volatility.
The setup: BreadCoin on a bakery platform
To keep the three models comparable, I'll use the same fictitious example throughout. A platform connects local bakery stores with customers, taking a 5% fee on every transaction plus a $100 annual listing fee per bakery. It doesn't own any bread, so all the economic activity on the platform belongs to third parties. Annually, 1 million loaves sell on the platform, across 1,000 bakeries, at an average price of $2, with operating cost of $100,000 per year. Before launch, the platform raises $100,000 by selling 500,000 BreadCoins at $0.20 each, and the same numbers run under all three models below.
Model 1: Currency-style tokens, value driven by turnover
This is the structure most ICO-era token sales defaulted to, and the one still most common on retail-facing launchpads. Everything on the platform is priced and settled in BreadCoin. The total transacted volume the token has to facilitate is $2.2 million: $2 million in bread sales, $100,000 in fees, and $100,000 in listing subscriptions. There are 1,001 actors who want to cash out BreadCoin for fiat (the 1,000 bakeries, plus the platform), which means the token's market price is governed by total turnover divided by velocity (see the quantity theory of money for the mechanics).
The value driver is turnover, not revenue. As long as sales on the platform are growing, everyone profits; a bakery that receives BreadCoin from a customer can immediately resell it to another customer buying different bread, and each round-trip nets the bakery its asking price. Risk is spread across everyone in proportion to how much they have at stake, so the platform has roughly $200,000 of exposure (fees plus subscriptions) while the bakeries collectively have $2 million.
The structural problem is that the platform raised money to build itself but is selling someone else's product. In a down year, the bakeries are the ones holding BreadCoin they can't offload at face value. If enough of them start mass selling, the price spirals down, and recovery from that is very hard because the token's value driver, turnover, has already collapsed. This is what most L1 gas tokens and closed-economy game tokens look like in practice; when transaction volume stalls, the token has nothing else to fall back on.
The platform raised money to build itself, but is selling someone else's product.
Model 2: Coupon-style tokens, value driven by revenue
Only platform fees are payable in BreadCoin. Everything else (the bread itself, the listing subscriptions) settles in fiat. The transacted volume the token has to facilitate drops to $200,000, which is 9% of the currency model's $2.2 million. There is also only one actor with a natural interest in cashing out the token: the platform itself, because it's the only party that receives BreadCoin as income.
This changes the economics completely. The platform sets the par value at which BreadCoin is redeemed for fee payment, a variant of the soft price floor technique, and in return it forfeits real revenue to back the token. At par value $0.50, the total circulating BreadCoin is worth $250,000. If every holder eventually spends their token on fees, the platform forgoes all of its Y1 revenue and 25% of its Y2 revenue to "buy back" all the BreadCoin in circulation. It raised $100,000 and pays $250,000 back in forfeited revenue over two years. If instead the par value is set at $0.10 (so the sale price was a 100% premium), half the token sale was raised at investor expense, which is a reputation disaster and probably a legal one.
The coupon pattern shows up in practice in most exchange tokens. Binance's BNB is the best-known: a 25% discount on spot trading fees and 10% on futures when fees are paid in BNB, plus a quarterly auto-burn that retires supply using Binance's operating activity. The coupon model is stable because the value is tied to platform revenue, which is usually less volatile than turnover. The trade-off is that the platform is forfeiting revenue, not profit, which can create liquidity problems even for a business that's otherwise solvent.
Model 3: Equity-style tokens, value driven by profit
The token represents a pro-rata share of the platform's profit, either directly (dividend) or via a buyback-and-burn that retires supply when revenue accrues. Every sale and fee on the platform settles in fiat; BreadCoin exists only to define who gets a cut. The numbers on our platform: $200,000 in fees, minus $100,000 in expenses, minus 20% corporate tax, leaves $80,000 of net profit. Divided across 500,000 tokens, that's $0.16 per token per year, so an investor who paid $0.20 breaks even in about 15 months and everything after that is profit.
The clearest live example of this structure in late 2025 is UNI, the Uniswap governance token. For five years UNI was a "governance only" token with no direct claim on the roughly $1 billion in annual fees the protocol generates. In December 2025, governance approved the UNIfication proposal with 99.9% support: all protocol-level fees now route to a burn contract, and 100 million UNI were burned retroactively from the treasury to approximate what would have been retired if the fee switch had been on from launch. UNI holders effectively moved from governance-only to profit-share overnight.
The equity model has the cleanest value driver of the three, because profit is what the business actually produces. The structural cost is legal. In most jurisdictions, a token that entitles holders to profit is a security, which triggers registration, disclosure, and investor protection obligations. The classification is not uniform: in the EU, MiCA classification treats asset-referenced tokens, e-money tokens, and utility tokens distinctly, and "security token" is still a country-specific legal concept. This is not a legal opinion, and if you're designing for this model you need a lawyer in the jurisdiction you're selling into. If you're weighing these three structures for a live project, the next step is running the valuation math for each, which I've walked through in how to value a token.
Side by side: platform risk, investor risk, volatility
Stacked against each other, the three structures trade off cleanly across the four dimensions that matter for design: platform risk, investor risk, price potential, and volatility. Reading the table below from left to right, each model has one column where it wins and two or three where it loses. There is no universally best structure, only a best fit for a given platform and a given risk appetite.
| Token is | Token type* | Value driven by | Platform risk | Investor risk | Price potential | Volatility |
|---|---|---|---|---|---|---|
| Currency | Utility/Security | Turnover | Lowest | Highest | Highest | Highest |
| Coupon | Utility | Revenue | Average | Lowest | Lowest | Lowest |
| Equity | Security | Profit | Highest | Average | Average | Average |
*Not a legal opinion. Classification depends on country-specific laws and regulations. Consult a lawyer.
The currency model pushes most of the upside (and most of the risk) onto the investor. Platform exposure is low, because the bakeries take the transaction risk, but the token's price swings with platform turnover, which is the most volatile of the three drivers. The equity model flips the positions: the platform carries the most structural risk (legal and operational), but the token is anchored to profit, which is the most stable driver of the three. The coupon model is the lowest-risk option for investors by construction, because the platform deliberately dampens the upside in exchange for stability.
What to actually pick
The three models are not interchangeable, even though they often get mixed together in practice. The right choice depends on what the platform expects to produce, how much regulatory exposure it can absorb, and how much of its revenue it can afford to forfeit. In most cases, the design decision is made early and is expensive to reverse.
Pick currency-style if the platform's core pitch is "we process a lot of volume" and the token holder's upside comes from that volume growing. Most L1 gas tokens land here because they have to: the token is the unit of account for the network. Be honest that a flat year breaks the model.
Pick coupon-style if the platform has steady fee revenue and wants a low-volatility token whose price is credibly pegged to the business. The cost is forfeited revenue, which is not the same as forfeited profit, so the finance team needs to model the cash-flow impact before committing. Most exchange tokens and marketplace tokens land here.
Pick equity-style if the platform is willing to take on the legal classification as a security token, and profit is large enough that a per-token dividend or buyback-and-burn is meaningful. This is the most powerful structure when it works, and the most dangerous when the legal work is done poorly. The UNI fee switch is the clearest recent example of a token graduating into this model, and it took five years and a governance vote to get there.
Most live projects end up hybrid in practice, combining coupon features (pay fees in the token at a discount) with buyback-and-burn (profit accrual through supply reduction), all wrapped in governance rights. See token economy design components for a fuller inventory of what's available. Hybrid is fine, as long as each feature is load-bearing and the design is not a grab bag of incentives nobody actually uses.
Frequently asked questions
01Can a token combine more than one value model, such as hybrid designs?
+
02How does token velocity affect the value of a currency-style token?
+
03Is "utility token" the same as a coupon-style token?
+
04How does a buyback-and-burn change which value model applies?
+
