Tokenomics Fundamentals Part IV: Token supply & demand dynamics

This article is part IV of the following six-part series:

Editor’s note (August 2026): I first published this article in January 2023 and refreshed it in August 2026. The original examples have been kept, and what has happened since has been added.

In the prior parts of the series, we discussed the “Why, When, What, Where, Who, and How” of token design.

W5H framework for token design (W5H) by Charles Shen @ inWeb3.com

While many of our discussions are in the context of the initial token launch, answers to these questions directly impact the ongoing supply and demand in the ecosystem, which is crucial for the long-term sustainability of the token project. The dynamic equilibrium of supply and demand at the overall business level and the token level determine the prices of goods and services as well as the tokens. They are the remaining critical components that close our W5H loop, and are the focus of this article.

W5H framework for token design (Supply & Demand) by Charles Shen @ inWeb3.com

As usual, we approach the supply-demand dynamics at both the business level and the token level. At the business level, the supply and demand of specific goods and services are actually part of the crypto-product-market fit problem discussed in “Why token”. Therefore, let us elaborate more on the token-level supply and demand mechanisms here.

Regulating token supply and demand

Token supply regulation

Token supply may be regulated through a project’s governance, to the extent that the token’s contracts and upgrade paths actually give governance that authority. Recall from our “How” token part that token supply generation can be pre-scheduled or on-demand. If the supply is pre-scheduled, some projects’ governance chooses to modify the original token emission pattern when the situation changes, to adjust the future token supply. If the token minting is demand-driven, there are also things the project’s governance might be able to do to adjust the supply. For example, if a stablecoin is minted against collateral, then lowering the required collateralization ratio (or raising the debt ceiling) expands how much of the stablecoin can be minted - whether more actually gets minted still depends on borrowers wanting to. However, such change has consequences. Lower collateralization ratios bring increased risk to the project’s overall financial status.

While supply regulation could mean both increasing and reducing supply, the more prevalent issue projects encounter is the over-supply of tokens that cannot be sustained by their demand. Let us focus on mechanisms that reduce token supply in more detail.

Temporary supply reduction - locking and vesting

Locking tokens temporarily takes them off the market and reduces the token’s selling pressure during the lock period (strictly speaking, locked tokens may or may not be counted out of “circulating supply” depending on who is counting - the practical effect is that they cannot be sold). A vesting schedule releases the tokens gradually over time, alleviating abrupt supply shocks. The project could set up incentives to encourage voluntary locking by the token holders. The more real value these incentives bring to the users, the more effective the mechanism could be. For example, the well-known veCRV model of Curve Finance locks the CRV token for up to four years - non-transferable and with no early exit - in exchange for governance rights and for sharing ongoing protocol fees. Derivative protocols built on the veCRV mechanism, such as Convex, even turn the locking of CRV from temporary to permanent: CRV deposited into Convex is locked forever as veCRV, and the depositor gets a liquid cvxCRV token in return. That last detail matters: the underlying CRV never comes back, but the holder can still sell the cvxCRV wrapper on the open market, so the selling pressure is transformed rather than eliminated. Both mechanisms are still running in 2026; Curve has since layered “liquid locker” products on top of the same primitive, and the veCRV fee share is still the core of its model.

Note that the more tokens locked does not necessarily mean the more valuable the token becomes. For example, if a token serves as a medium of exchange for goods, and nobody can get the token because all tokens are locked, then the exchange for goods cannot happen, and the overall crypto economy would not even exist.

Permanent supply reduction - burning tokens

Burning existing tokens can permanently remove them from the circulating supply. It artificially introduces a deflationary force to counter the token’s inflationary pressure and potentially improve token value - “potentially” being the operative word, since a burn changes the supply side only; it does nothing for demand. Common token-burning mechanisms try to tie the burning process with the token demand. One approach is to burn tokens when users spend them on the protocol’s services. ETH’s EIP-1559 implementation burns the base-fee portion of every transaction fee (the priority tip still goes to the block producer). BNB has a similar real-time partial gas-fee burning, BEP-95, where a governed share of each block’s fees is destroyed. The Graph protocol that provides blockchain data query services also burns GRT, though its sinks have been refined over time - the current protocol documentation lists a delegation tax, a curation tax, and part of any slashing as burns, alongside a 3% annual issuance for indexing rewards. A burn on one side and issuance on the other is typical; you have to look at both to know the net.

When the burning mechanism is meant to counter token inflation, designing the right parameters to strike a real-time token supply-demand balance could be very challenging because of unpredictable factors in the economic system. One approach people use is to periodically assess the condition to decide how much supply should be reduced, then buy back and burn them. The BNB token’s quarterly auto-burn, e.g., determines the burning amount by a formula based on the BNB price and the number of blocks produced on the chain during the quarter, which means that more tokens will be burned if the price of BNB declines. That program has run every quarter since; the 36th burn took place in July 2026, still working toward the stated target of bringing total supply down to 100 million BNB.

Smoothing and reduction of future token supply

Supply smoothing and reduction may also be applied to tokens not yet released.

Tokens scheduled for future release typically adopt a relatively smooth emission pattern rather than a step function-based inflation leap, as Synthetix argued back in 2019. Synthetix’s own journey since then is worth following to its end: after several rounds of smoothing, its governance passed SIP-2043 and set weekly SNX inflation to zero in December 2023, and the old SNX staking system has since been retired altogether - a schedule that started as “smooth it” ended as “turn it off.” (In 2026 a separate, purpose-specific mint was authorized to wind down sUSD, a reminder that “zero inflation” is a governance decision, not an immutable rule.) Convex protocol’s pre-scheduled CVX token emission is an example of a smooth minting curve with a reduced inflation rate: a 100 million maximum supply, with the CVX-per-CRV mint ratio stepping down every 100,000 CVX. For tokens that are designated for specific participants, especially those at low or zero cost for investment or rewards, a norm is to enforce a mandatory lockup and vesting period. For instance, teams and VCs typically receive their allocated tokens after a cliff, followed by gradual vesting over a period of years - the exact schedule differs by project, and the specific unlock table is worth looking up rather than assuming a market norm. Similar locking and vesting can be applied to token rewards earned by yield farming, e.g., as in the DeFi Kingdoms game, where a portion of JEWEL rewards was locked and released later. That game went on to rewrite its token schedule entirely: emissions on its original Harmony chain ended, the JEWEL cap was reduced to 125 million, and the locked JEWEL was redistributed under new terms - a vesting design is only as permanent as the governance behind it.
We have also seen projects adjusting their future release schedule to explicitly cut emissions. E.g., the Sushi token started with an infinite inflation schedule. Later its governance deemed it too aggressive and approved a 250M hard cap for the token. That cap was reached: Sushi reported in November 2023 that all 250 million SUSHI had been minted and there would be no further emissions, and a proposed new tokenomics model was shelved after mixed community feedback. The Cosmos 2.0 whitepaper released in September 2022 proposed changing the ATOM token’s long-term emission schedule from an exponential expansion to a more contained linear schedule. But it also required a significant short-term increase of token emission and was rejected along with other amendments. The community came back to the question a year later in smaller steps: a proposal to halve the maximum inflation rate from 20% to 10% passed in November 2023, while a follow-up to cut the minimum to zero failed. ATOM’s experience indicates that a bundled, sweeping redesign is much harder to pass than one parameter at a time.

Regulating token demand

Handling supply and demand dynamics does not have to be limited to the supply side. It is possible to stimulate or suppress demand in response to changing supplies caused by external factors. The DAI savings rate mechanism of MakerDAO offers a good example. In a basic scenario of MakerDAO’s DAI stablecoin, people deposit their ETH tokens as collateral to mint DAI, increasing DAI supply. In a crypto bull market, people are more likely to take on ETH leverage to mint DAI, resulting in more DAI supply and relatively weaker DAI demand. At such times, the protocol may reward people for holding DAI by raising the DAI savings rate, thus increasing the DAI demand to bring the supply-demand relationship towards a more balanced position. In a bear market where people prefer the stability of holding DAI, the demand is stronger than the supply of DAI because people tend not to leverage ETH to mint DAI. In response, the protocol can reduce the DAI savings rate, practically discouraging people from holding DAI, and suppressing its demand. That is the stylized version; in practice the rate is one lever among several that governance pulls, and competing yields elsewhere in the market matter as much as the collateral cycle. The mechanism itself is alive and well: MakerDAO has since rebranded to Sky, DAI can be converted 1:1 into its newer USDS, and the savings-rate idea now has a tokenized form, sUSDS, which is itself a demand sink for USDS.

Token supply and demand analysis

We can ask three common questions when examining token supply-demand dynamics and the resulting token price. Two of them will be discussed in this section, and the third one on token value capture will be covered in the next part of this series. We also select two symbolic projects from the 2021-2022 crypto market cycle to illustrate the application of the presented questions.

What value creation drives the token demand and is it sustainable?

We temporarily ignore the supply side when addressing this question. The system should have sound fundamentals that support a consistent and increasing demand for the token. We can review our answer to the “Why token” question, diving deeper into the token project’s business model and its value creation process. Are there meaningful services and products that produce revenue? Where does the revenue actually come from? Is the revenue sustainable? How do the revenues compare with the costs?

If a system is not financially sustainable yet but provides meaningful services, it could still be successful. Just like companies in the tech world that are losing money during the growth period. These projects could receive external funding (primarily from VCs) that helps them through those early periods. Of course, there is no guarantee that such projects will indeed prevail; many of them inevitably fail.
If a system seems profitable but relies primarily on a constant influx of new user funds to sustain its business model, that should raise a flag for potential “ponzinomics”.

How will the token supply change, and can the demand match it?

If the crypto project passes the fundamental demand drive test under a given token supply, we still need to check whether the changing token demand and supply can achieve a healthy balance.

If supply decreases, it should help push the price higher in the short-term given similar demand. But if supply decreases too much, it might stifle the demand and harm the overall economy. Therefore, many token economies are designed to have moderate supply inflation. This inflationary supply can either follow a pre-determined schedule or be released “on demand.” Let us look at what could happen to the new token supply:

People may sell them on the market and cash out; this is the least wanted case since it indicates higher supply and lower demand, which pushes the price down.
People may hodl the token for the long term. This behavior is desired as it signals higher demand and keeps the new supply off the market for the time being, reducing selling pressure.
People may spend the new tokens on a service that burns them (when applicable). This outcome also represents demand for the token and is highly desirable since it essentially negates the extra supply and minimizes the dilution of the new supply.
Other user behaviors could lead to various degrees of demand impact in between the ones listed above.

If aggregate impact from user demand consistently falls short of matching up with the supply, the economy could fall into a hyperinflation mode and lead to a token price crash.

Case study 1: OlympusDAO and OHM

OlympusDAO was the poster child of what the market called “DeFi 2.0” in 2021. It aspired to “create a decentralized, censorship-resistant reserve currency for the emerging Web3 ecosystem”. Its OHM token intends to preserve purchasing power, maintain deep liquidity, and serve as a unit of account and a store of value reserve. But unlike crypto stablecoin projects, the OHM token is not pegged to the Dollar. Instead, it is backed by treasury assets such as the decentralized stablecoin DAI. The change from “pegged” to “backed” is significant. It means OHM has a free-floating value, with the treasury backing serving as a reference point for its value - not, it should be said, a redemption guarantee.

What is the primary value that draws users to the OHM token?

When OHM was launched in March 2021, its immediate demand driver was not its status as currency because it was yet to be adopted. Instead, it offered a passive earning APY as high as six figures for staking OHM - the protocol’s own first-month review reported capacity for a 100,000% APY. Even in mid-2022, the APY was still in the mid three figures (the protocol’s own metrics show roughly 468% in May 2022, falling to 266% by the end of July). The incredible APY, along with its (3,3) meme, were the most visible reasons people bought and staked OHM at the time.

Where does the yield of OHM come from, and is it sustainable?

A deeper look into the protocol reveals that the early APY was funded by minting new OHM - “rebasing” - against a treasury that grew mainly from the protocol selling OHM bonds to buyers willing to pay far more than the roughly one-DAI backing per token. But why would people be willing to pay a much higher price, which at the top reached about $1,400-$1,500 in April 2021, knowing the backing was roughly $1? Some users may think the project’s future growth, including the massive treasury assets the protocol controls through the bonding mechanism, is worth the premium. Others may FOMO into it because of the (3,3) meme. If everyone does cooperate and buys only, the price keeps going up. If the price keeps increasing, getting in early gives you a better price and a higher share of the protocol’s value accrual. The only problem is, for the price to keep going up, there need to be ever more new buyers entering the system. When the new buyer inflow stops, the system looks a lot more like Ponzinomics than Tokenomics. Jordi Alexander from Selini Capital has more insights on this topic. To its credit, Olympus’s own governance said as much at the time - an OIP in June 2021 called the high reward rate unsustainable in the long term and began cutting it.

Summary:

Overall the Olympus protocol has some impressive design, with an ambitious goal and a once wildly popular meme. Unfortunately, its original model lacked a sound fundamental and was not sustainable in that form. In hindsight, it should not be too surprising to see the OHM unit price fall by over 99% from its ATH by the beginning of 2023. (One caveat on that number: because staked OHM rebased - a staker’s token count grew by a cumulative index of roughly 269x over the life of the program - a chart of the unstaked unit price overstates what a long-term staker actually lost. It was still a very large loss.)

Nevertheless, the Olympus project has also produced some impactful innovations, especially the concept of protocol-owned liquidity, which is considered a key characteristic of DeFi 2.0. And the story did not end in 2023. The protocol is still operating in 2026 with a materially different design: rebasing has been switched off and staking APY is zero; the treasury now runs a buyback-and-burn facility funded by treasury yield on the downside and a premium-gated emissions manager on the upside; and holders can borrow stablecoins against their OHM at a rate tied to the backing. The protocol’s own metrics put treasury market value at roughly $185M in July 2026, with liquid backing of about $12 per OHM, and it still holds protocol-owned liquidity across several chains. Whether that is a good investment is a separate question; the point is that the 2021 “buy and stake for 100,000%” machine was replaced, by governance, with something that regulates supply on both sides. That is the supply-demand toolkit of this article put to work.

Case study 2: Axie Infinity

Axie Infinity is an NFT-based blockchain game. Players battle, collect, and trade NFT digital pets called Axies. It became the emblem of the 2021 play-to-earn boom.

What is the primary value that draws users to the Axie Infinity game and its tokens?

Many players were attracted to Axie Infinity for the prospect of earning income. A May 2021 Twitter poll conducted by the game developer, asking what people liked most about Axie, had the “Economy” well ahead of the gameplay as the top answer. (The exact poll results are no longer retrievable from Twitter/X.) That suggests a large proportion of the participants were looking for a job, not a game.

Where do the earnings from the game come from, and is it sustainable?

If people are in the game to earn, where does that money come from? Unfortunately, the early Axie Infinity economy did not generate self-sustaining revenues - and the team was candid about it: their own economics page stated that “in the beginning, to maximize growth, by design the Axie economy will be dependent on new entrants.” It required players to pay an upfront amount to buy the gaming NFTs to start playing. In mid-2021, a decent starter team of three Axies cost up to $1,000. The continuous influx of new users paying the cost to join the game was what supported a token price level that allowed existing players to cash out and earn a meaningful income. However, it’s impossible to expect an unlimited number of new users to join the system at an increasingly higher price tag. When new user growth slows down, the token price drops, and existing “worker” players cannot earn enough to make it worthwhile. They may leave the game, risking a downward spiral. In short, the demand side fundamental of the game, as it stood in 2021, did not pass the sustainability test.

How will the token supply change?

The Axie Infinity game has been acclaimed for providing low-income communities with a play-to-earn model to make a living. That is indeed a worthy cause. A guild-based scholarship mechanism was also created to help onboard more users who could not afford the game entry NFTs. The idea is to let them play with Axies the guild owns and split the earnings - Yield Guild Games, the best-known guild, used a 70/20/10 split between the scholar, the scholarship manager, and the guild. This mechanism worked and effectively brought in a large number of users to the game. YGG alone reported over 4,000 scholars by the end of July 2021, and at one time, an estimated 60%-65% of the people who own Axies were scholars.

However, so many new users entering the system also means accelerated token supply, e.g., people winning more of its SLP tokens in the game. This situation adds to the existing inflation pressure caused by bots that further push the minting of more tokens in the game.

Can token demand match the supply change?

On the demand side, users can sell the earned SLPs and cash out; that does not create new tokens, but it puts the freshly minted ones straight onto the market. The users can also spend those SLPs to breed Axies. That will burn and deflate SLP tokens. Therefore, the outcome will be determined by the comparable speed of SLP inflation vs. deflation. An analysis of SLP minted vs. burned by NAAVIK showed that through the second half of 2021, the SLP minting speed had been faster than its burning speed, and the gap kept widening; the Axie team’s own economic balancing post in January 2022 described “a growing chasm between SLP minted vs. SLP burned” and called the inflation unsustainable. Apparently, the deflation power was not strong enough to offset the increasing supply. This phenomenon is probably not surprising if we recall that most of the scholars and low-income players treat the game as a job to win the SLP tokens and sell them for cash earning. They are mainly adding to the sell side, not the burn side.

Chart of SLP minted versus SLP burned through 2021, showing minting outpacing burning Source: CoinGecko and NAAVIK

If we look at the price of the SLP token and its market cap, the summer of 2021 marked a period when the SLP token price started a quick descent from near $0.34 in mid-July to around $0.11 by the end of August and $0.06 by the end of September. This pattern suggests that the game suffered a severe token-inflation problem.

Chart of SLP price and market cap in 2021, showing the decline from July onward Source: CoinGecko and NAAVIK

Balancing a game economy has never been an easy task. Solving the in-game SLP inflation issue requires the designers to reduce the token inflation source and/or increase the deflation sink. But these decisions often bring dilemmas. Reducing the number of SLP token rewards also changes the incentives for players and could affect their engagement. On the other hand, raising breeding costs burns more SLP per Axie but slows the growth of the Axie population, and a scarcer Axie supply means higher Axie prices that further increase the barrier for new players to join.

To their credit, the team publicly acknowledged these problems and has been working in various directions to improve them. They addressed the inflation problem by making it more difficult to earn SLP tokens and by raising breeding costs. For the fundamental question of actual revenue generation, they listed several options on the Axie Economy & long term sustainability page. They also released a free-to-play version of the game as part of those efforts, potentially onboarding more people from the broader player community and earning revenue from in-game items.

The years since have been a live experiment in exactly the source-and-sink levers discussed above. The free-to-play version, Axie Infinity: Origins, launched in 2022 with three free starter Axies, removing the upfront NFT barrier; SLP rewards were stripped out of the old game modes and moved to Origins’ ranked play; new SLP sinks were added, such as crafting in-game runes. In January 2024 the team announced a 44 billion SLP supply cap - enforced, notably, by “social contract and in-game emission mechanisms” rather than by the token contract. Bot farming never fully went away; after banning bot accounts in 2025, the team went to the root and stopped Origins SLP emissions altogether in January 2026, citing automated farming. Game rewards have since shifted to bAXS, a non-transferable token backed by AXS that can be spent in-game or converted to AXS at a rate that depends on the player’s standing. The game is still running - Origins was in its 18th season in July 2026 - and SLP still has its breeding and crafting uses. So: the 2021 economy did not survive, but the game did, and it got there by turning off the inflation source entirely and redesigning the reward token so it cannot be dumped on the market. That is about as direct an answer to “can demand match the supply?” as a project can give.

Conclusion

This article follows our prior coverage of the “Why, When, What, Where, Who, and How” of token design, and explored the dynamic equilibrium of supply and demand that closes the W5H loop. Since supply and demand dynamics at the business level is part of the prior “Why token” discussion, we looked at various ways to regulate supply and demand at the token level in this article. Using two symbolic crypto projects as case studies, we also elaborated on two questions covering both the business and token levels, that help examine the supply and demand relationship of a given crypto project.

Across the examples in this article the pattern is the same: the supply schedule a project launches with is rarely the one it ends up with. Synthetix and Sushi took inflation to zero, Cosmos cut it in steps after a sweeping redesign failed, DeFi Kingdoms rewrote its vesting, Olympus replaced rebasing with a two-sided policy engine, and Axie turned its reward emissions off. Token supply-demand design is an operating system that governance keeps tuning, not a chart that gets published once.

One remaining question we have not delved into is the connection between the two levels. Specifically, how the value created at the business level accrues to the token level. This is a key aspect of tokenomics design that impacts the valuation and pricing of the tokens. It will be our topic in the following part of this series.

This article is part of the six-part W5H framework for token design series, as listed below:

Tokenomics Fundamentals Series: W5H framework for token design, Part IV - Token supply and demand dynamics

Table of Contents