Hyperliquid primer and 5-year valuation framework

1. Executive summary

We value HYPE under a P/Fees multiple framework applied to seven revenue streams, projected across bear, base and bull scenarios over a five-year horizon. The headline outputs are as follows:

• Bear: ~$8 implied price by 2031, a -30% annualised return from current spot.

• Base: ~$147 implied price by 2031, a +25% annualised return.

• Bull: ~$456 implied price by 2031, a +57% annualised return.

HYPE is one of the few crypto assets where network usage translates almost directly into demand for the underlying token. 99% of trading fees from the underlying protocol flow to an Assistance Fund that buys back HYPE on the open market every day, with approximately 44.4m HYPE purchased to date, worth around $2.2B at current prices. The mechanism functions similarly to a corporate share buyback, but executed transparently, on-chain, and at a rate that scales with platform activity rather than discretion. Few comparable assets in the space capture this much value for holders directly. 

The methodology in this report applies a P/Fees multiple to projected revenue across perpetuals, spot, HIP-1 auctions, HIP-3 tokenised perps, HIP-4 outcome markets, HyperEVM and the AQAv2 USDC reserve yield (don’t worry if you don’t know what these are, they will be explained). 

The multiples themselves are anchored to HYPE's actual trading history. On a circulating P/Fees basis, HYPE has ranged between ~7x and 25x over the past 12 months with an average of ~12x. Bear, base and bull are set at 10x, 15x and 20x respectively. None require multiple expansion from current levels. 

2. What is Hyperliquid?

Hyperliquid, in the words of the foundation that created it, is "the blockchain to house all finance". In more direct terms, it is a high-performance Layer 1 blockchain purpose-built around an on-chain perpetual contracts exchange. Users can trade a range of assets through fully on-chain order books, including crypto, real-world assets, pre-IPO stocks and, more recently, prediction markets. 

The diagram above describes Hyperliquid’s architecture at a high level.

HyperBFT is the consensus mechanism that underpins the blockchain. BFT, short for Byzantine Fault Tolerance, means the network can continue operating even if some validators fail or act maliciously.

HyperBFT finalises the chain, ordering every transaction, secured by a validator set that earns rewards (~2.1% APY) to do so.

Sitting on top of it are two execution environments, HyperCore and HyperEVM. 

HyperCore is where the flagship perpetual contracts exchange lives. Every single order, cancel, trade and liquidation happens here, transparently and with sub second finality. It can support up to 200,000 orders per second, performance that is comparable to that of a centralized exchange but entirely on chain. HyperCore contains perpetuals, spot markets, vaults that allow users to allocate to trading strategies, as well as the governance module that decides things like fee parameters and HIPs (Hyperliquid Improvement Proposals).

HyperEVM is the general-purpose smart contract layer. EVM stands for Ethereum Virtual Machine, meaning HyperEVM can interpret and execute code written for Ethereum. Developers familiar with Ethereum tooling such as Foundry or Hardhat can deploy contracts easily, lowering the barrier to building on Hyperliquid. HyperEVM is not a separate blockchain. It is secured by the same HyperBFT consensus that HyperCore is. This is where the broader ecosystem lives: dApps like lending protocols, tokens issued via HIP-1 auctions, and bridges that connect Hyperliquid to other chains. 

Hyperliquid has come a long way since the inception of its mainnet in 2023, boasting some significant developments that have moulded it into the leading platform it is today:

2023: Mainnet launch with the perpetual contracts DEX, the flagship product that drove early adoption.

2024: HIP-1 ticker auctions go live, allowing tokens to be listed on Hyperliquid spot markets via a competitive auction mechanism.

February 2025: HyperEVM launches, opening the platform to general-purpose smart contract development and the broader DeFi ecosystem.

October 2025: HIP-3 introduces builder-deployed perpetuals, enabling anyone to launch their own perpetual contract markets (stocks, commodities, FX, indices) by staking 500k HYPE.

May 2026: HIP-4 outcome markets (prediction markets) go live, bringing fully on-chain prediction markets to the platform. Live but still very early.

Hyperliquid has amassed strong traction due to a strong user experience: low spreads and fees, high liquidity, transparency, and most importantly, access to a wide variety of markets with varying levels of leverage. It has become an important venue for large scale hedging, directional trading, and anything in between. 

• 1.20 million lifetime users

• US$4.33T+ lifetime volume

• US$1.1B+ in cumulative revenue

3. Token economics & the value accrual flywheel

Oftentimes in crypto, a protocol might be an amazing product, providing genuine utility to its users, or even generating large amounts of revenue. However, more often than not, there is only a tenuous link between the underlying economics of the application or protocol to the token that represents it. 

For Hyperliquid, a novel approach to revenue was introduced. Through the Assistance Fund (AF) mechanism, approximately 99% of the fees generated on the platform are used to buyback HYPE on the open market. As previously mentioned, this has been ~44.4M tokens to date, with a current value of over US$2 billion. 

Different from a direct burn (i.e. Ethereum), the assistance fund sets limit orders below the current market price, using USDC to bid up the token, then sending the purchased HYPE to a null address that can never be accessed (formally recognised as a burn by 85% validator vote in December 2025). This has the dual effect of reducing circulating supply while providing genuine downside support at any given moment. Ethereum's EIP-1559 burn, by comparison, removes supply but generates no equivalent demand. Hyperliquid does both. To be clear, this compares the specific value-accrual mechanisms. Ethereum has other demand drivers; the point is that its burn alone does not include a buying component, whereas Hyperliquid's mechanism does both. 

This is what makes HYPE genuinely different. Over time, the token should perform akin to the underlying business, meaning it can be valued like equity in a way most crypto assets cannot. If revenues grow, buybacks grow, creating direct pressure on circulating supply and ongoing demand that fundamentally backs the token. A unique mechanism that has, in our view, propelled HYPE into a deserved spotlight.

Trading fees are not the only thing that accrue value to HYPE token holders. Recently, the Aligned Quote Asset (AQA) framework was formalised, which routes a portion of the yield earned (repos, short term treasuries) on USDC reserves held on Hyperliquid back to the Assistance Fund.

When users deposit USDC to trade perpetuals on Hyperliquid, that USDC sits as collateral against open positions. With roughly $5 billion of USDC currently on the platform, those reserves earn a meaningful amount of yield (3.5-4% at current rates), and a share of that yield (90%) flows to the AF, where it is converted into HYPE buybacks via the same mechanism described above.

USDH, a stablecoin issued by Native Markets, was the original AQA deployer when the framework launched in September 2025. It never gained meaningful traction. By May 2026, USDH supply had stalled at around $100m while USDC on Hyperliquid had grown to over $5B. Coinbase has since taken over as the official AQA treasury deployer for USDC, with USDH being gradually wound down.

At current rates, the AQA layer contributes approximately $140m-160m per year to the buyback. This revenue line is dependent on interest rates and the outstanding USDC supply on Hyperliquid, both of which we model in detail later. 

A last mechanism that affects the supply side is staking tiers. Users that stake HYPE on the platform can earn discounts on fees depending on the amount staked, ranging from 5% at the lowest tier to 40% at the highest. 

The mechanism creates natural alignment. The traders generating the most fees on the platform are the same traders with the most reason to stake the largest amounts of HYPE, which locks supply away from the tradable float. Stakers also earn ~2.1% APY in HYPE rewards on top of the fee discount. Unlike the AF burn, staked HYPE is recoverable (subject to a 7-day unstaking period), so this does not technically remove tokens from circulating supply, but the reduction in tradable float still amplifies price sensitivity to any incremental buying. 

Acknowledging the other side of the supply story, the tokenomic structure is as follows:

Total Supply: 1,000,000,000 HYPE (1 billion, fixed)

Future Emissions and Community Rewards: 388.88m (38.8%)

Reserved for future community incentives

Initial Genesis Airdrop: 310m tokens (31%)

Fully unlocked

Current and Future Core Contributors: 238m tokens (23.8%)

Team allocation, 1-year cliff post-TGE, then vesting through 2027-2028, some beyond

Hyper Foundation Budget: 60m tokens (6%)

Fully unlocked

Community Grants: 3m tokens (0.3%)

Fully unlocked

HIP-2 (a marginal allocation tied to the HIP-2 spot AMM mechanism ): 120k tokens (0.012%)

Fully unlocked

Another rarity in the midst of the crypto-sphere is that there was absolutely zero allocation to VC, private investors, centralized exchanges or market makers. This is strong not only from a narrative perspective, but it meant the launch of the token was fair, rewarding early users of the platform and allowing for natural price discovery. This also matters analytically. With no VC, no private investors and no exchange listings to pay for, the only insider supply that can hit the market is the team allocation discussed below. 

The team allocation, which began vesting after the one year cliff post-TGE, is the most obvious source of supply overhang. On paper, roughly 9.9 million HYPE per month was expected to enter circulation through linear vesting over 24 months. In reality, the actual distributions to team members have come in materially lower than the on-paper schedule:

• December 2025: 1.75m HYPE

• January 2026: 1.2m HYPE

• February 2026: 140k HYPE

• March 2026: 173k HYPE

• April 2026: 333k HYPE

The vesting contract releases its scheduled amount, but the team has consistently chosen to re-stake, hold or otherwise not distribute the bulk of what becomes technically available. Co-founder Iliensinc confirmed in late December that the team formalised a "6th of each month, if any" cadence, with the implicit caveat that the size of each distribution is discretionary rather than mechanical.

Combined with the burn mechanism on the demand side, this means HYPE has consistently been a net deflationary asset since the cliff expired. The market initially priced in significant supply overhang as the team cliff approached. The data so far suggests that overhang has been considerably smaller than anticipated.

This does not eliminate future risk. The unvested portion of the team allocation still represents a meaningful share of total supply, and behaviour can change. But the early signal is that incentives are aligned around long-term value capture rather than short-term distribution.

For the 5-year framework, we assume the contractual vesting schedule plays out in full, with ending circulating supply at 707M HYPE across bear, base, and bull cases. There are two choices embedded here. First, we do not give the model credit for the team continuing to under-distribute relative to the schedule. The recent pattern is encouraging but not guaranteed, and modelling it forward as continued restraint would be assuming a particular behaviour that has not yet been tested through a full cycle. Second, we do not retire HYPE from circulating supply via the buyback. As discussed in Section 6, the buyback effect is captured within the multiple itself, and reducing supply on top of that would double-count the same value-accrual mechanism. This feels like the right choice as it would likely, in most scenarios, overstate the ending supply and therefore give our cases higher upside.

Having established how value accrues to the token and how supply behaves, we turn to the size of the market Hyperliquid operates in. 

4. Market sizing & competitive landscape

Hyperliquid primarily operates in the highly competitive perps market. According to CoinGecko's 2025 Annual Crypto Report, perp trading volume was a whopping US$92.9T. Of that, volumes on decentralized exchanges hit an all time high of US$6.7T (346% y/y), a clear signal that the on-chain perp category is in the steepest part of its growth curve. Although there is a segment of users that would only trade on the former rather than the latter, the majority have the ability to select either. We can therefore say that Hyperliquid's competition is not just decentralized platforms (dYdX, Drift, Jupiter, Aster, Phoenix etc), but also the centralized behemoths like Binance, Bybit and OKX.

Whilst modelling the growth of the overall market size is important, and is covered in our model, it is also equally important to model how much of the pie Hyperliquid has taken historically and where that could go.

A year ago, Hyperliquid was trading at roughly 5% of the volume of all centralized exchanges combined. That number now sits at around 6-7%. Modest in absolute terms, but it is important to remember that they started from scratch in 2023 (data does not go that far back to display). Hyperliquid has been steadily gaining share on a market dominated by venues that have been operational for the better part of a decade and have significantly more resources at their disposal.

The picture is more pronounced when broken down by individual CEX. Against Bybit, Hyperliquid has gone from roughly 30% to over 50% of trading volume in the space of twelve months. Against OKX, the ratio has moved from ~25% to ~31% on volume, and on open interest from ~45% to ~57%. Binance remains the most resilient of the three, but even there Hyperliquid has gained some ground.

It is also worth pointing out the regime that this has happened in. Crypto markets have been relatively muted for much of this period.

The chart above shows Hyperliquid's market share among perp DEXs over the past year. A few caveats are worth flagging upfront. The dataset does not capture every competitor (Drift, Jupiter and Phoenix being notable omissions), so Hyperliquid's share is somewhat overstated here. Our internal estimate of true market share among perp DEXs sits closer to the 30-40% range.

The chart still surfaces something worth discussing. The black line is Aster, which briefly captured significant share in August and September 2025 before collapsing back down. As mentioned earlier, a meaningful portion of Aster's volume during this period was widely believed to be wash trading and incentive farming rather than genuine demand. The chart is a reasonable visual representation of how distortive that kind of activity can be to share metrics, and a useful reminder that not all reported volume is created equal.

The bottom line is that the perp DEX category is competitive and dynamic. Hyperliquid is the clear leader, but the competitive set is real and continues to evolve. The question for valuation purposes is not whether competition exists, but how much share Hyperliquid can defend over a 5-year horizon. We address this in detail in our scenario assumptions later in the report.

The perp category itself is also evolving. For most of crypto's history, perpetual contracts have meant pretty much just crypto exposure (long BTC, short ETH, and so on). HIP-3 has opened the door to perpetual contracts on tokenised real-world assets such as commodities, equity indices, FX and pre-IPO names. These are markets where the traditional counterparts trade trillions of dollars a year. This is a fundamentally different addressable market. Hyperliquid does not need a meaningful share of these markets to materially expand its revenue base. The bull case is that on-chain perps become the natural venue for accessing these markets 24/7, with global access, and without the friction of traditional brokerage rails. The bear case is that regulated venues retain the institutional flow and on-chain stays niche. Both scenarios are reflected in the model.

A similar logic applies, on a smaller scale, to prediction markets through HIP-4. The category is still nascent on Hyperliquid (live since May 2026) and on-chain more broadly, but the underlying market exists and is growing fast. Polymarket cleared roughly US$26B+ in trading volume in Q1 2026 alone, up 90% from the prior quarter, with March 2026 the first month to break US$10B in a single month. Hyperliquid is unlikely to displace Polymarket in the near term, but does not need to. Prediction markets are an optionality line in our model rather than a core driver of the thesis.

Pulling it together, Hyperliquid has won meaningful share in its core market against well-established competition, in a macro environment that has not been particularly forgiving. Several adjacent markets, notably tokenised real-world assets and prediction markets, are meaningfully expanding the addressable opportunity. Hyperliquid is attempting to be “the blockchain to house all finance”, and will do its best to take share from incumbents.

5. Revenue streams

5a. Perpetuals

Hyperliquid is currently running at an annualized volume of ~US$2.45T (~US$938B YTD). The fee structure is of course, as mentioned previously, dependent on whether the user is staking the underlying token for discounts. At certain volume sizes, makers can earn rebates from the protocol in exchange for providing liquidity and tighter spreads.

The current observed protocol take rate is ~3.14 bps after rebates, builder codes and other allocations. This would mean, ceteris paribus, the protocol is on track to earn ~US$770M for 2026. Worth noting that the gross fees paid by users sit closer to ~7 bps. The ~2x gap reflects the rebates, builder code routing (a fee-sharing mechanism that incentivises third-party front-ends to route flow to Hyperliquid), and HLP allocation (the protocol's own market-making vault, which receives a portion of fees as compensation for providing liquidity) .

From here, the question is how much volume Hyperliquid can capture over the next 5 years. The model approaches this in three layers: how big the total perpetuals market becomes, how much of that volume moves on-chain, and how much of the on-chain volume Hyperliquid holds. 

Note that the bear case take rate is slightly higher than the base and bull cases. In a low-volume regime, fewer makers reach the top rebate tiers and the protocol faces less competitive pressure to discount, so more of the gross fee stays with the protocol. In the base and bull cases, the opposite applies, and we model take rate compressing to ~2.5 bps as institutional and builder-routed flow grows as a share of volume.

The bear case is not "Hyperliquid fails". The bear case is a world in which the total perp market does not grow much from current levels, DEX share grows modestly to 12%, and Hyperliquid loses meaningful share to competition (from 35% to 22%). Even under those conditions, the platform is still doing US$1.58T a year in volume. Revenue declines from US$770M to ~US$475M, which is a real drawdown, but the platform is far from irrelevant. Note that the bear case does not apply to possible risks that lead to Hyperliquid failing as a whole. These are discussed separately. Bear case is a scenario in which it still operates properly, just does not succeed.

The base case assumes a more constructive view of crypto adoption. The total perp market roughly quadruples to US$225T (in line with 3 and 5y CAGRs) as crypto continues to mature, DEX share grows to 25%, and Hyperliquid holds 40% of that. We have already covered the share dynamics in Section 4, and projected forwards a moderately conservative growth rate. The output is US$22.5T in annual volume and US$5.6B in revenue, a 48% CAGR.

The bull case is where on-chain perps go mainstream. Total perp market hits US$300T, DEX share reaches 38%, and Hyperliquid deepens its moat to 45% of DEX volume. This produces US$51T in annual volume and US$12.8B in revenue, a 76% CAGR. Aggressive, but not absurd, considering perp volumes have grown at well over 100% CAGR in some recent years. Importantly, this section on perps does not include HIP-3 perps, which are tradeable stocks, indices, FX pairs and commodities. 

Perpetuals account for over 75% of Hyperliquid revenue today. The remaining streams are smaller individually but can become significant contributors over time.

5b. Spot

Spot trading on Hyperliquid is currently running at an annualized volume of ~US$56B, generating ~US$28M in revenue. Small in absolute terms next to perps, but worth taking into account for two reasons. First, spot take rates are structurally higher than perps (~5 bps vs 3.14 bps), meaning every dollar of spot volume contributes more to the buyback than every dollar of perp volume. Second, the HIP-1 ticker auction (a Dutch auction where projects bid HYPE for the right to list a new token on Hyperliquid spot) creates a natural pipeline of new listings, giving the platform recurring volume (as explained in the next section).

The drivers are similar to perps but with one additional dynamic. HYPE-USDC spot pairs are not routed through the AF buyback. The HYPE portion is burned directly.

Note also that the take rate is modelled to compress over time, as fees likely become more competitive across spot markets generally. 

5c. HIP-1 Auction fees 

HIP-1 ticker auctions currently generate a small revenue stream for Hyperliquid but could meaningfully contribute to the bottom line in the future. Each auction settles in HYPE at the winning bid price, with 100% of proceeds flowing to the AF for buyback. The HIP-1 system is currently producing ~US$7M in annualized revenue (~500 HYPE per ticker). 

In its early days, it peaked at ~US$75M annualized, and our bull case would be a return to those levels.

Ultimately, HIP-1 will not be a massive driver for revenues, but needs to be mentioned for completeness.

5d. HIP-3 (tokenised RWA perps)

HIP-3 went live in October 2025 and is the layer of the Hyperliquid stack that genuinely opens up new addressable markets. The premise is straightforward: Any builder with 500k HYPE staked can deploy a perpetual contract market on Hyperliquid for whatever asset they want, subject to validator review. The result is a permissionless marketplace for tokenised commodities, equity indices, FX pairs, pre-IPO stocks and anything else with a credible oracle feed.

Adoption has been rapid, catalysed by two key events: Precious metals outperformance in early 2026 led to many crypto traders turning their attention from crypto pairs to GOLD and SILVER contracts, and geopolitical escalation in Iran in late February leading to a massive boom in OIL contract volume, particularly over weekends, when Hyperliquid offered some of the only liquid pricing while traditional commodity markets were closed.

HIP-3 has cleared US$60B+ in cumulative volume since launch, with over 75,000 unique traders. trade.xyz, which is currently around 90% of HIP-3 open interest, secured an official license from S&P Dow Jones Indices in March 2026 for a perpetual contract referencing the S&P 500. The first time a major TradFi index has been officially licensed for a decentralised perpetual product. That kind of institutional validation does not happen for niche infrastructure.

HIP-3 markets charge 2x the standard perp fee, split 50/50 between the deployer and the protocol. The net effect is that the protocol collects roughly the same per-trade fee as a standard perp. The catch is "Growth Mode", an optional setting deployers can activate that cuts all-in fees by 90% for 30 days to bootstrap liquidity. Most HIP-3 markets today are running in Growth Mode, which is why the realised protocol take rate sits at ~0.55 bps rather than the ~3 bps it would be at standard terms.

The take rate is held flat across all three scenarios at ~0.55 bps. The reason is that Growth Mode can be renewed indefinitely at the deployer's discretion, and we have no visibility into when (or whether) the dominant deployers will move to standard pricing. We treat eventual maturation to standard terms as upside not modelled, rather than a base case assumption.

On volumes, the spread between bear and bull is wider here than for any other revenue stream in the model. This reflects the genuine uncertainty about HIP-3's eventual size. Things are happening right now very quickly. Being able to ascertain that Hyperliquid will be the dominant venue for on-chain RWA perps would be disingenuous. It's certainly possible, proving so far to be a useful venue for trading markets 24/7 and access to pre-IPO stocks (see Cerebras recently - Wall Street execs pictured with Hyperliquid on their screen before the stock went live).

Somewhere between all 3 cases is a material risk that regulated venues such as CME or ICE launch their own tokenised markets and absorb future flow that would otherwise have gone to Hyperliquid or competing chains. There is also a more immediate angle worth flagging: CME and ICE have already pushed US regulators to scrutinise Hyperliquid over alleged manipulation risks. The concerns themselves are debatable. The motivation for raising them is less so: incumbents have a clear interest in slowing on-chain competition before it scales. It is worth noting that Hyperliquid is geoblocked in the US, so this is not about US regulators stopping it in any literal sense. It is about the bull case requiring regulatory acceptance, which as we know can be a long and winding road.

5e. HIP-4 (outcome markets/prediction markets)

HIP-4 went live on the 2nd May, 2026. At time of writing, this is the newest addition to the Hyperliquid stack and the least proven. We treat it as a $0 contribution in 2026E for that reason. The mechanism allows for fully on-chain prediction markets, binary outcome contracts on real-world events, settled in USDC.

The category exists and is growing. Polymarket cleared US$26B+ in volume in Q1 2026 alone, with March 2026 the first month to break US$10B. Kalshi has scaled similarly on the regulated side. Hyperliquid enters with structural advantages: zero fees on opening positions, full collateralisation, and the ability to combine prediction market exposure with derivatives positions on the same platform. If HIP-3 is any indication on whether the execution of this market will be successful, there is reason to be at least moderately optimistic on how it plays out.

The scenarios are deliberately wide. The bear case is HIP-4 launching but never getting material traction. $5B annualised by 2031 would mean Hyperliquid captures less than 5% of Polymarket's current monthly run-rate over five years of growth. The base case is Hyperliquid reaching ~4x current Polymarket volume by 2031 (we would expect Polymarket to grow also in this scenario). The bull case is ~8x. None of these would make HIP-4 the headline driver of the valuation, but in the base case it adds ~$50M to annual buybacks, and in the bull case ~$100M.

This section likely contains the most forecasting error. Given its nascency, the underlying mechanics are not well understood.

5f. HyperEVM

HyperEVM is the general-purpose smart contract layer described earlier in the report. Total Value Locked sits at roughly US$1.5B today, spread across lending protocols, vaults, and various dApps. The revenue contribution to the protocol comes from gas fees and ecosystem activity rather than trading fees directly.

The honest view is that we are unconvinced HyperEVM displaces Ethereum or Solana for general DeFi. That said, the ecosystem has scaled faster than most challenger L1s. HyperLend and Morpho together hold roughly half of all HyperEVM TVL, with lending the dominant category. There is room for HyperEVM to surprise to the upside, but we model that even in our bull case, the burn contribution from fees on the chain will be modest.

For context, post-Dencun Ethereum L1 captures only ~0.1-0.25% of its TVL in fees, while Solana captures closer to 1% during active periods. Our bull case at 1.25% for HyperEVM sits roughly in line with Solana during active periods.

In our bear case, TVL falls and is largely idle.

5g. AQAv2 USDC reserve yield

The AQA framework was explained earlier in Section 3. Today, with roughly US$5B of USDC on the platform earning 3.5-4% in short-term yield, the contribution sits at ~US$140-160M annualised. That is approximately 15% of total 2026E revenue, a meaningful share. 

Forward projections are based on two things: USDC supply on Hyperliquid, and interest rates. We are not in the game of projecting where rates will be 5 years from now, as apart from a fool's errand, no matter what the projection is, it would be wrong.

Admittedly, the projections here are more “finger in the air” than other sections. We would expect that in the base and bull case, USDC supply grows in tandem with the protocol, at different paces reflecting the level of growth. In the bear case, USDC contracts due to Hyperliquid losing out to competing chains, and to be more punitive, terms are renegotiated and yield contribution is lowered. 

In all cases, we have projected 2.5%, for no reason other than it is lower than today and so within the rate path possibilities, could err on the conservative side. Upside here would be higher rates, with downside being ZIRP (with secondary effects likely benefitting crypto as a whole, however).

6. Valuation framework & methodology

We have covered Hyperliquid fairly extensively. It should be clear what it is, and how the underlying revenue drivers work. What makes our framework fairly straightforward is the buyback mechanism from HyperCore and burn mechanism from HyperEVM.

We choose to apply a P/Fees multiple to total revenue to derive market cap, dividing by circulating supply to derive an implied terminal price for the token under each scenario.  Before diving in, see below the historical P/Fees for circulating supply. 

The average P/Fees ratio the market has assigned to Hyperliquid is 11.78x. There are, obviously, exogenous factors that drive volatility in this multiple: broader market downturns, large buyers, large sellers, speculative frenzies, and so on.

Instead of trying to model the projected amount of tokens that would be removed from the supply from the buyback (this would be a circular exercise), the buyback is captured within the multiple itself.

We use 10x, 15x and 20x for the bear, base and bull cases respectively. Bear at 10x sits roughly at the historical average. Base at 15x sits modestly above, reflecting our view that Hyperliquid's revenue base will be materially larger and more diversified by 2031 than today, which justifies a moderate re-rating. Bull at 20x reaches the historical high range but does not exceed it.

The P/Fees multiple can also be expressed differently. Every dollar of revenue routed through the buyback represents a return to HYPE holders, in the same way a dividend or share buyback represents a return to equity holders. Dividing annual buyback dollars by market cap gives us a "buyback yield".

This yield is useful as a “point in time” metric, as it shows, through a different lens, opportunities where HYPE could be stretched or undervalued in the short term. It is tautological to P/Fees multiple, but the framing allows you to compare it to capital opportunities in well established equity names.

Our higher multiple in the base and bull cases reflects a future willingness to accept a lower buyback yield as HYPE de-risks over time, with the protocol maturing and diversifying its revenue base.

Looking at the historical buyback yield chart above, HYPE has traded between ~4% and ~14% over the past 12 months. The current yield sits at the lower end of that range, consistent with the higher current multiple. Given HYPE’s recent rally, it could indicate it is overheated in the very short term.

A DCF approach is theoretically applicable here, given the programmatic nature of the AF buyback as a recurring cash flow. We have chosen multiples as our primary framework for the reasons set out above, but it is worth showing what a DCF produces also.

The method is a standard two-stage approach. Project the buyback stream over the 5-year horizon at our scenario CAGRs, then apply a terminal value at the end of year 5 using a 5% perpetual growth rate. We run this at three discount rates (12%, 15%, 18%) to bracket the typical range for crypto cost of equity.

The two approaches land in roughly the same place in the base case, regardless of which discount rate you pick. They diverge at the tails, in opposite directions. In the bear case, the multiples approach is harsher than the DCF, because a 10x multiple on shrunken revenue is more punishing than a cash flow stream that still has some terminal value attached. In the bull case, multiples give a higher number than DCF, because multiples capture things a pure cash flow model does not, namely optionality, network effects, and the dynamics of the buyback removing supply while creating demand.

7. Valuation summary 

Bringing the seven revenue streams, the multiples framework, and supply mechanics together, the full picture is below.

At current spot price, the base case implies ~25% annualised returns over five years, the bull case ~57%, and the bear case ~-30%. 

Optionality Premium - This is applied on top of the multiples-based market cap to capture what our framework does not explicitly model: the value of HyperCore being able to deploy new products (HIP-5 or beyond is not in our model, but the developers have demonstrated a high ability to launch things successfully), the network effects of being the dominant on-chain trading venue, and the reflexivity of the buyback creating buy-side pressure that compounds with platform growth. It also captures speculative hype, which is most pronounced in the short term and is a real driver of crypto prices in either direction.

This premium is the least precise component of our model. Sentiment shifts quickly and can swing prices meaningfully in either direction over short windows. The dislocations to extreme highs and lows in the historical P/Fees chart shown earlier can largely be thought of as this same unexplained residual. We size the premium at 30% in 2026E to reflect both the visibility of optionality today and the short-term speculative dynamic, stepping down to 5% in the base case and 10% in the bull case where much of that optionality is by 2031 already reflected in the revenue numbers themselves. The bear case has no premium because the scenario assumes the platform thesis does not materialise.

Our 2026E implied price of $66.52 sits modestly above current spot price. We read this as a sign that our model is conservative. The market is meanwhile pricing in a mix of forward growth and short-term frenzy. "In the short run, the market is a voting machine, but in the long run, it is a weighing machine". 

Circulating supply is also explicitly difficult to project. Higher/lower emissions can vary results, and the buyback cannot be projected due to its circularity with price, so our estimations are likely to be significantly above or below the ending circulating supply.

8. Risks

Investing in the HYPE token carries meaningful risk. Below is a non-exhaustive list of risks we currently identify:

• Competition - Hyperliquid is the clear leader today, but the perp DEX category is dynamic. Aster captured meaningful share in mid-2025 before collapsing (albeit on dubious legitimate traction), which is a useful reminder that share is not permanent. Solana-based perp venues, Jupiter, Phoenix, or new entrants could materially impact Hyperliquid’s ability to retain or expand market share. 

• Regulatory - Hyperliquid is geoblocked in the US, but that does not eliminate regulatory risk. The CFTC has shown growing interest in on-chain derivatives. CME and ICE have already pushed US regulators to scrutinise Hyperliquid, and incumbents have a clear interest in slowing on-chain competition. 

• Team supply overhang - The team allocation continues vesting through 2027 and 2028, with a meaningful tranche scheduled to unlock in November 2027. Although distributions to date have come in well below the on-paper schedule (as we covered in Section 3), there is no guarantee that this will continue, and in fact, we have modelled essentially full unlocks in all of our cases to account for this.

• Community supply overhang - Part of the locked supply is for the community. An airdrop in the future could lead to high selling pressure which could not only materially impact prices, but also skew some of the assumptions we set earlier.

• Protocol and technical risk - Hyperliquid's validator set remains relatively concentrated. A successful exploit, consensus failure, or USDC bridge incident could lead to a total loss of funds for users, which would materially impact prices, potentially making the token essentially worthless.

• Key man risk - Jeff Yan and the core team have driven the platform's trajectory from inception. This is a strength today, but is a double-edged sword. Dependence on a few key people to execute a long-standing vision can impact execution if not diversified.

• The contrarian short case - HIP-3 take rates stay compressed indefinitely as Growth Mode renewals continue. HIP-4 fails to find product-market fit. The AQA gets renegotiated as Coinbase exerts leverage. Perps revenue grows but at meaningfully slower rates than our base assumes, while the multiple compresses back toward the historical average, or far below. 

• Builder concentration risk - trade.xyz being ~90% of HIP-3 open interest. If trade.xyz fails, runs into legal trouble, or migrates elsewhere, the HIP-3 projections would be meaningfully impacted (barring a new, successful builder entering).

By the time you finished reading this article, the Assistance Fund bought back 960 HYPE.

If you jumped straight to the end… 3 HYPE.