Perps Will $DRIFT to L1s

Mark Leontev, Katarina Keller

Derivatives on decentralized assets have historically existed within centralised trading environments — and that remains true today, with many app-chains operating much like traditional CEXs, often with worse security guarantees. Still, it’s fair to classify them under the separate DEX umbrella, as they continue to push the frontier of censorship resistance and transparency.

CEX Futures are entering Q3 2025 doing around $5-7T monthly volume:

Perp monthly volume is experiencing its historical breakthrough, reaching $1T levels:

Such volumes are being monetised at ~3.7b cumulative 30d annualised revenue across the leading perp DEXes:

Combining CEX and DEX perpetuals, the market is generating around $8T in monthly volume as of Q3 2025, equating to an annualised $80-90T (depending on the period used). Applying the current average blended take rate of 0.025% from the table above yields a $20-25B addressable market today. That’s the realistic opportunity available to platforms either fully focused on perpetuals or expanding their product suite from a derivatives-first foundation.

The overall market could expand further as global crypto adoption accelerates and regulatory clarity improves. However, margins are also likely to compress as competition intensifies - with platforms like Robinhood and Coinbase launching new products, additional perp exchanges preparing to enter the market alongside CEX platforms collectively pushing more operations onchain:

As it currently stands, No ID needed for sex (yet) but when it comes to leverage - there's a very obvious latent demand for permissionless risk taking on volatile assets. An onchain derivatives venue unlocks the ability for anyone, anywhere, to trade any asset - traditional or digitally native. 2025 marks the first time in crypto history that we’re seeing truly high-quality offerings emerge across the space, led by the rise of Hyperliquid and the ensuing “perp DEX wars.”

In September 2025, spot DEX volume totaled $363B, hovering around its recent-historical average of roughly $400B. It was only in July 2025 that total derivatives volume finally started to meaningfully surpass spot, going well beyond $500B:

In other words, it took crypto years since inception - and five years since DeFi Summer - for the industry to finally reach the traditional financial equilibrium where derivatives exceed spot by orders of magnitude. This has always been the norm in finance; it just took us a long time to get here.

In a sense, we’re only now witnessing the true birth of on-chain finance.

Perpetuals as a financial primitive are the most convenient derivative instrument, hands down. They have gained widespread popularity due to several key advantages over traditional financial contracts:

• They never expire, allowing traders to maintain positions "perpetually" without worrying about rolling over contracts.
• Funding rates make perp swaps both spot-accurate and dead simple—unlike futures' contango/backwardation mess (we swear, nobody in crypto even heard these words, not once).
• Compared to spot trading and borrow/lend, they offer a much more accessible leverage functionality.
• They only require 3 core functions: a matching engine, an oracle, and a high quality low-vol asset for margin.
• They enable trading on virtually any asset - forex, commodities, equities, crypto, prediction market events, and more.

Different platforms, however, approach these advantages CEXs are becoming increasingly rigid — adopting slower, more conservative listing standards and tightening compliance across jurisdictions. Access to perpetuals is being restricted further as major exchanges like Binance continue to block users from large regions because they have to comply regionally.

At the same time, the opposite trend is unfolding onchain. Perp DEXs are accelerating — listing assets faster, iterating on product design, and, most importantly, opening permissionless access to users excluded from traditional systems. While many DeFi front ends still geoblock regions like the U.S., the underlying protocols remain accessible, creating a powerful countercurrent: as CEXs close their doors, derivatives DEXs are prying them open, filling the gap on a very big market:

There are 3 high-level things we believe with regards to perps market over the next 5 years:

We believe, the TAM for onchain derivatives is far larger than can be perceived through backward-looking calculations. By removing trust and access barriers, perps unlock trading activity that today remains suppressed or entirely excluded from traditional markets. Permissionless, global participation could connect billions to financial instruments for the first time - a scale the market continues to underestimate.

While the vast majority of perp volume still centers around the big three - BTC, ETH, and SOL - it’s reasonable to expect growing activity in the longer tail as more users adopt Web3 wallets for non-speculative purposes: DePIN payments, stablecoin transfers, and general apps usage, among others. These users may eventually seek to hedge their on-chain exposures, often through niche contracts - for instance, taking a short position in HNT to offset the volatile value of hotspot income.

As developing economies come online, users will naturally gravitate toward DEXs that can’t be censored or reshaped by unilateral policy changes (and this does not exclusively pertain CEXes), and the most efficient venues will capture share from centralised exchanges by operating leaner — with no need for massive support, localisation, or custody infrastructure. The outcome will depend on how early and effectively each protocol balances incentives to attract both liquidity providers and takers.

Derivatives DEXs, first conceptualised in UMA’s 2019 BitDEX paper, began on Ethereum L1 before migrating to L2s such as GMX on Arbitrum, which dominated the previous cycle.

Recently, the trend has shifted toward DeFi-focused ‘appchains’ purpose-built for derivatives. However, we see appchains as structurally flawed over the long term — constrained by a set of limitations the current market seems to drastically under-index:

• they are not atomically composable with major liquidity base that's already hosted elsewhere,
• they depend on centralised bridging infrastructure,
• they suffer from absence of native stablecoin liquidity,
• by definition, they start out with literally zero third-party collateral assets,
• their founders seem to be largely ignoring the inherent anti-network effects of their venture: if they are successful they will have no choice but to transform into general purpose chain to host broader set of apps thus jumping into the red ocean of permissionless infra scaling, at which point
• they will find themselves in an investment conundrum where structural buyers will not be sure what's the right way to value such venture

Hyperliquid, the most prominent appchain-based DEX, is nowadays largely perceived as a counter example to all the points above whereas in reality it actually exemplifies these trade-offs: it operates an onchain CLOB with impressive speed as long as it relies on a small and relatively opaque validator set largely controlled by a limited set of core contributors, leaving it exposed to potential censorship despite its technical performance. Even though they have overcome initial liquidity bootstrapping difficulties, they continue relying on centralised infra providers for bridging and wrapping quality collaterals that by definition are not HL native assets. The initial success have sparked an ecosystem of founders that want to build there, yet their businesses cannot be hosted on the Core ledger as it artificially benefits a very niche type of txs. They now need to build from scratch a separate layer, introduce extra multi-layer friction, and somehow make it so happen that the next wave of quality asset issuers will be deployed here, all the while leaving potential structural buyers perplexed as to whether or not it should be valued as a chain or as an app.

That means, for any next appchain disruptor in the perps space, the founders will have to fight numerous uphill battles and they will have to execute twice as hard on each of those directions. Appchain design in derivatives dominate today largely because, a few years ago (when they were all conceived), building on general-purpose Layer1s was simply not viable. Native collateral frameworks were underdeveloped, user bases were thin, and infrastructure lacked the throughput and reliability needed for high-performance derivatives trading. In that environment, developers had no choice but to build their own stack.

But that rationale is rapidly eroding. As Layer 1s mature - with deep native collateral pools, composable DeFi ecosystems, and increasingly performant execution layers - the trade-offs of isolation are becoming clear. Appchains still introduce friction: new wallets, fragmented liquidity, and dependence on stablecoin issuers to support yet another chain.

Claims that incumbents like Hyperliquid are untouchable simply ignore the context. Market share in crypto trading has always been fluid, and liquidity reliably migrates toward superior products and ecosystems once the infrastructure and business development catch up:

Bitstamp once dominated crypto trading until Coinbase emerged, capturing nearly 80% of market share within a single year — only to later lose it to Binance after no more than six consecutive months of dominance. Binance has been the only exchange able to partially defend its position over time, yet even its dominance has waned: it now controls less than 50% of global volume, and the broader landscape is far more fragmented with a lot of players in the tail end.

Bitstamp today accounts for less than 1% of volume, Coinbase under 8%, and Binance’s share is roughly half of what it was before DeFi Summer. OKX, once approaching 40%, now holds around a quarter of that. The list of early BTC exchanges that once defined the market is long — and nearly all of them have gone into oblivion.

The likewise dynamic can be observed throughout the perps space. The “entrench-ness” of Hyperliquid is already shattering now, with its perp volume market share dropped two times since Aster and Lighter came on stage, from 60% market share in Q2 to 20-ish% now:

It’s worth noting that the grey area on the chart above was once about 70% dYdX — yet another "entrenched incumbent" of its time, now faded largely into irrelevance:

It was quite recent times when it was still reasonable to expect things from dYdX. Alas, the landscape changes drastically every quarter.

2025’s "HL-will-rule-forever" sentiment echoes what people once said about dYdX and Cosmos (that IBC interoperability would solve the friction across top-notch apps). These narratives come and go.

In our view, the rapid rise of Hyperliquid did not validate the appchain thesis while opening the new chapter of finance onchain. Rather, it demonstrated that the 3 following statements are correct:

1. It’s still very early to extrapolate in DeFi.
2. There are no true liquidity moats yet.
3. Latent demand for on-chain derivatives is far greater than what backward-looking valuations can capture.

This is what the market size looked like when the Hyperliquid founding team was first conceiving the idea for their exchange:

The “entrenched incumbent” doing ~1b daily volume. If you took this number for its face value, you'd probably end up building on some other market as an early-stage builder in 2023.

Fast forward to today, the market looks like this:

Daily volumes of the most dominant perp exchange from one year ago would barely make it to the top 10 today. Currently, dYdX is doing ~250m daily, occupying the 18th position in this ranking. The current incumbents outperform it by roughly 40x in daily trading volume — and the gap would be even wider if measured by open interest.

It’s important to recognise that permissionless access fundamentally expands participation in global financial markets — potentially reaching billions who are currently excluded. In DeFi’s ideal future, anyone, anywhere, will be able to trade any asset they choose. This shift is profoundly market-expansionary - and the true scale of that opportunity remains vastly underestimated because it’s unquantifiable. It takes a great deal of intellectual convexity to envision some platform one year from now doing 40x volumes from the current levels. Yet, if history is any indication, this is exactly the most likely outcome (and of course, the most entertaining one).

What we’re seeing in 2025 is the first innings of latent demand beginning to flow in - and there’s still no genuine liquidity moat established by anyone, including Hyperliquid. A key differentiator for Hyperliquid will be how it responds to emerging competitors as they chip into the next 40x volume expansion.

Due to all the above mentioned, we believe it's not the most risk-reward prudent choice for capital allocators to take the momentous status quo at face value, betting on the seeming incumbents for the long term as a high-conviction bet. In the markets like this, it's often rewarding to revisit the overlooked design choices that have been disregarded and left for dead due to lack of traction, or poor UX, or even for no reason because it's seldom in this space for people to need one.

In our view, The most asymmetric opportunity in this entire vertical lies with perpetual DEXs natively deployed on Layer 1 chains.

We can categorically conclude that current market consensus is fixated on the app-chain model - fast, self-contained, and seemingly dominant. Yet, this consensus itself has become a form of complacency we have seen time and time again in the space. For years, on-L1 perp architectures have been written off as inefficient or impractical, ever since dYdX’s high-profile departure from Ethereum. What was once considered an engineering constraint has quietly turned into a psychological one, not being challenged by enough people from first principles.

The reality is that the foundations of Layer 1 ecosystems have evolved dramatically since then. With scalable execution (Firedancer, DoubleZero), efficient liquidity routing (Jito BAM), and native collateral bases deep enough to sustain institutional activity, L1s are now very close to being able to host high-performance derivative systems natively. We expect tx ordering innovation happening on Solana via initiatives like Jito BAM come to fruition before EOY 2025, which will in turn open a customisability broadway for apps and market makers.

The current bottleneck for solving liquidity on Solana is a function of the "adverse selection" problem which you can know more about here. We evaluate this issue as an engineering problem whereas the current consensus largely treats this problem as inherently fundamental to any general purpose design. This assessment is wrong and it is destined to reflexive panic-repricing.

The trade-offs that once justified building isolated stacks are rapidly eroding — composability, network effects, and shared user liquidity are increasingly outweighing the performance edge of appchains, which is factually marginal while the market treat is as the paramount.

If the market’s assumption that “perps must live on appchains” proves just a little bit less certain than it appears in 2025, it opens one of the widest asymmetries yet. The next leg of this cycle could favor teams that have quietly chosen to build within the L1 environment — where asset issuance, collateral flows, and liquidity naturally converge. And this is where projects like Drift begin to stand out — not as an underdog, but as a logical beneficiary of the market’s structural mispricing. It will become one of the things that make perfect sense in the hindsight.

History of DeFi shows that the industry repeatedly overestimates incumbents and underestimates architectural shifts. Because the former is visible and sexy, and the latter is quiet and boring. The next phase of derivatives growth may not come from those adding more incentives and features, but from those positioned where the assets - and the users - already live. On-L1 perp DEXs represent not a relic of the past, but one the most underappreciated opportunities of the future.

As discussed earlier, we are likely to see another winner-takes-most dynamic in crypto trading both in derivatives and spot, similar to what Binance achieved in 2018. Launched in mid-2017, Binance became the largest crypto exchange within a year, capturing roughly 50% of global market share. Its rise showed how a compelling product, combining diverse trading options with a native token incentive model, can rapidly reshape the market landscape.

These shifts tend to create a dominance structure in which the leading platform rapidly grows to hold a market share roughly five times larger than any of its top-10 competitors:

Binance became the fist crypto superapp around 2018, and they achieved this through one key principle: integration. It unified critical financial services under a single platform, tied together by one native token. The same formula applies to DeFi: multi-product ecosystems where one token captures the value across all services are the winning model.

We haven’t yet seen a truly successful example of a multi-product DeFi platform. The Blockworks team predicts that it will change soon — and that the first breakthrough will likely happen on Solana.

Historically, perpetuals appchains have struggled to be perceived as anything beyond their core product. We believe the core reason has been the absence of a powerful base chain capable of hosting all user cohorts across all product types. Solana in 2025 appears uniquely positioned to fill that gap — potentially becoming, for the first time in crypto history, a genuinely integrated financial layer.

If this plays out, multi-product DeFi is poised to succeed. Drift already stands as the leading incumbent at this frontier: originally a perps DEX, it has since expanded into balance sheet management, curated vaults for market makers - all built on a single, composable L1 under the umbrella of one platform and one token, which is now announced to capture 100% of Drift's revenue.

Such one-stop-shop platforms gain stickiness through convenience — one of the strongest moats in crypto. Every additional day of Tron’s proliferation is proof of that:

USDT on Tron has become over the years of previous crypto cycles the second crypto super-app. By owning the user relationship end to end, they can extract far higher margins than competitors, because their service, even if imperfect, remains “good enough” in the eyes of a large, entrenched user base.

Looking back at the history of super-app development, Binance provides the clearest example. It rapidly built a full suite of products for its existing users — Binance Chain, derivatives, spot margin, global fiat gateways, options, and more. This integration allowed Binance to grow from virtually nothing to the world’s largest crypto exchange in just six months.

Replicating this model onchain is challenging precisely because everything is on chain. Each component must be natively either composable or otherwise interoperable and achieve near-instant transaction finality. That level of integration and efficiency simply isn’t possible across hundreds of isolated appchain environments.

Which is why, the key differentiator between Drift and the dominant perpetuals exchanges of 2025 is its strategic decision to remain integrated with the Solana L1, unlike Hyperliquid, Lighter, and other newer entrants.

This choice introduces a distinct set of trade-offs. Drift and other Solana-native perp DEXes consciously trade temporary architectural efficiency for access to Solana’s broader capital base and user network, placing much of the resulting operational complexity and optimisation burden on market makers rather than the protocol itself.

As the base layer continues to improve and they platforms continue to build out new products, they will look and feel more and more like a CEX. The possibilities derived from composability are effectively boundless and underappreciated while the technical bottlenecks are surmountable and over-indexed. We are just starting to see how they are going to evolve when BAM plugins are in, builder codes and Drift APIs are well connected and the cohort of strategists and JIT Keepers have further grown. We are also yet to see the innovation in prop AMMs expanding their services onto the derivatives.

With all that in mind, we are confident the following assumptions are going to be validated in the near future:

• Derivatives-focused platforms are optimally (if not best) positioned to capture the super-app status as global finance moves onchain.

Perpetuals are the most convenient derivative instrument yet devised for capital managers to express hedge intentions - a capability unattainable years ago due to immature infrastructure.

Robust financial strategies typically rely on derivatives because derivatives are the way to express hedge intention. No money manager can forgo hedging, and most TradFi strategies normally pursue delta-neutrality, and nothing unlocks delta-neutrality at scale like perps.

Contrary to DeFi’s current fragmentation, driven by mercenary capital and retail prosumers with very low switch cost, perp-driven liquidity will become the liquidity of superior kind because it will be fuelled by well-capitalized institutionals who need to express the hedge intention in size and for a long time. That is literally the definition of “lock-in” - something that can't be achieved by a spot-first business.

With mature infra coming in place, perps will empower delta-neutral strategies that underpin most institutional approaches, naturally gravitating capital toward perp-heavy products and solidifying their moats through entrenched hedging liquidity. These players will account for >70% of trading volumes, prioritising stable, low-risk strategies that endure over long periods. Their sticky capital - unlike transient retail flows - will form the backbone of liquidity moats for derivatives-first superapps. Those moats will allow such platforms to become “good enough” one-stop shops, retaining users at a higher margin even when slightly better/cheaper/faster alternatives arise.

• Capital efficiency for professional managers is going to be derived from a much broader set of assets than today.

Unlike today’s perp platforms, which mostly accept only USDC as collateral, future systems will support a wide range of YBAs, RWAs, synthetic assets, and major native tokens — giving capital managers unprecedented flexibility to differentiate as they compete.

The ability to cross-collateralise diverse assets through an integrated risk engine will define the next generation of derivatives platforms. While relevant today, its impact remains constrained by the limited availability of quality collateral — mostly a few stablecoins and major cryptocurrencies. In five years, a broad mix of yield bearing, tokenised assets (real estate, reinsurance, gold, equities), and more stable, widely adopted altcoins will vastly expand collateral options, multiplying capital efficiency and deepening the liquidity moats of derivatives-first applications.

Multi-product platforms will operate under a cross-collateral regime, netting risks across perps, spot, and lending markets. This will allow allocators to attract capital from diverse investors holding different assets. It sounds quite easy but two key challenges emerge:

1. Technical Complexity. Building a true cross-collateral margin engine is extremely difficult. It took Drift over a year to bring one live on Solana mainnet. Such a system demands exceptional engineering depth and persistence to ensure assets compose as fungible risk units on the backend, while still allowing for isolated margin and interoperability across products like spot and borrow/lend. To our knowledge, no other platform in the perp space has managed to build it so far.
2. Native Onchain Assets. This kind of engine is nearly impossible to build in isolated environments like app-chains - simply because there are few, if any, native assets to begin with. Just as liquidity begets liquidity, issuance begets issuance. Quality asset issuers naturally gravitate toward ecosystems where others already exist - namely, L1s. Dex-chains face a structural uphill battle: either overcome cold-start issuance or accept the friction of bridging, wrapping, and fragmentation. In most cases, the resulting complexity and risk management overhead will be pushed onto users and capital allocators via conservative collateral haircuts — reducing capital efficiency and competitiveness.

Drift is one example that embodies these hypotheses as Solana’s leading cross-collateral perp DEX at the moment - purpose-built for institutional capital managers and designed to leverage the technical strengths and emerging capabilities of the L1 best positioned to host the collateral issuers of tomorrow.

Drift builds on Solana’s speed, composability, and concentration of asset issuers and users to maximize its capital base and its collateral base - currently held back by technical bottlenecks that Jito BAM and Alpenglow aim to unlock within the next 6 months.

The team is fully dedicated to product, unburdened by infrastructure maintenance: Solana scales autonomously through Firedancer (targeting 1M+ TPS) and Agave, both of which enhance perp matching without requiring Drift’s intervention.

Powerful tailwinds are now aligning. Drift’s dominance in perps coincides with major Solana upgrades - Jito BAM enabling custom order flow and Alpenglow reducing finality times - alongside a wave of TradFi inflows: ETFs channeling billions onchain, tokenised funds emerging as native buyers, and DATs unlocking institutional perp strategies.

With strategic backers like Forward Technologies and Kyle Samani, substantial new TVL is likely to accelerate this momentum. The overall setup positions Drift for a meaningful repricing as it evolves into a derivatives-powered super-protocol.

Historically, DEXes have been among the most profitable business models in crypto, with operating margins of roughly 70–95%, as they avoid many of the fixed costs faced by CEXes—customer support, compliance overhead, custody management, etc. For example, Coinbase employs over 4,000 people to generate around $6bn in annual revenue, while Hyperliquid operates with only 11 employees and generates roughly $1bn—equivalent to $90m of earnings per employee, versus $1.5m per employee for Coinbase.

Within an increasingly competitive “red ocean” of trading protocols, we believe the ability to offer a comprehensive trading suite—spot, perps, vaults, and structured products—will enable super-apps like Drift to extract superior margins over time. Users are typically don’t mind to pay a modest premium for convenience of one-stop shop.

Drift is currently traded in a midst position, on the half-way between the sweetheart incumbent and highly discounted platforms with little traction such as GMX, Contango and Aster (which is discounted for a different reason alongside its spot brother Pancake):

In summary, our base case assumptions are:

• Total crypto derivatives volume grows 5x by 2030 to approximately $395 per year.
• DeFi derivatives reach 50% market share by 2030, up from 15% in September 2025.
• Drift captures 5% of the DeFi derivatives market share by 2030, up from approximately 1.15% today. Drift’s market share stood at 0.22% in August 2023 (5x growth in 2 years).
• Net take rate assumed at 3 bps, which is 1 bps lower than Multicoin’s assumption (page 25) because we take into account that margins are going to compress as we go forward due to competition and decreasing monetisation rate, with platforms like Lighter already introducing RH playbook and not charging anything on the retail side.
• MEV contribution in Solana ecosystem grows 5x by 2030, with 40% app-layer capture, reflecting Drift’s integration with Jito-BAM and broader Solana ecosystem upgrades.

We stress-test our base case by adjusting key assumptions — DEX to total futures ratio and Drift’s market share within perps and MEV capture — to very conservative levels. Even under pessimistic scenario, with DeFi derivatives representing only 5% of total crypto derivatives (below the current level) and Drift reaching just 2% share (2x its current market share), the model still implies approximately 4.8x upside ($4.1 token price) to the current token price $0.9 (Sep 2025).

In summary, based on the assumptions laid out above and using our valuation model, we think DRIFT is undervalued at its current price of $0.9. Our base case price target is $14 implying 16.5x growth over current token price.

If you would like to play with our assumptions, we have attached our calculations here.

P.S. Following the market disaster 10-11 Oct 2025, Drift’s token price decreased to $0.45 which roughly doubles the implied upside, increasing the base case upside from ~16.9x to ~32-35x.