12 Charts to Watch in 2026

What Data We’re Looking At:

What Data We’re Looking At:

A narrative that sits outside of crypto, but has seeped its way into all industries at an accelerating rate throughout 2025 is AI. In this chart, we focus on AI, through a specific blockchain niche that is x402. In case you’re unfamiliar, x402 is an open payments protocol, pioneered and developed by Coinbase in 2025, that enables software, including AI agents, to programmatically pay for API access, data, or digital services using stablecoins. This essentially gives AI agents freedom to transact using cryptoassets, including the use of stablecoins to automatically pay, per request, for APIs or digital services. The chart we’ve included here tracks the x402 usage throughout 2025, measured as the weekly volume generated through the x402 protocol. Of course, x402 is still early, and it’s possible some payment activity routes through alternative rails over time, so we treat the chart as a proxy for agentic, machine-native commerce rather than a single winner-takes-all standard.

Why it’s Important:

We think this one is fairly straightforward in terms of why it’s important, Ai agents now have financial autonomy and the freedom to transact, within their defined parameters. x402 acts as a clean, measurable way to test whether AI-agents are shifting to adoption, given it sits at the exact point where agents move from generating outputs to executing work on the internet. That ‘adoption phase’ framing is explicitly how VanEck describes 2026, a transition from phase 1 (build-out) to phase 2 (adoption), where the winners need a credible path to ROI. We see evidence to suggest 2026 is the year where knowledge-work automation becomes operational reality, with AI-native rewrites and AI co-workers becoming a reality. This is exactly the sort of world where machine-native payments become a bottleneck unless they’re embedded into the workflow. x402 solves this issue, and is therefore a direct proxy for this. 

What Data We’re Looking At:

Onchain vault protocols are DeFi’s answer to asset management, and the space has developed considerably this year. At their simplest, vaults are best understood as onchain fund wrappers, in which users deposit an asset into a vault, and a curator deploys that capital according to strict, pre-defined logic to generate yield, abstracting away the complexity of managing individual positions. Vault technology itself was the big winner last year, with protocols developing to mature the onchain primitive, a specific example is Morpho, launching their v2 product, which offers advanced risk management, an updated role-based governance, and customisable access controls for institutional-grade compliance. These are the feature improvements that will pave the way for mass migration of institutional capital into these onchain fund structures. This specific chart tracks the AUM of onchain vault providers across 2025, shown as a stacked bar chart split by protocol. We focus specifically on Morpho, Euler, Yearn, Veda, and Beefy.

We explore the rise of onchain vaults, and what it means for onchain asset management, in our Keyrock report, ‘Onchain Asset Management: A Guide’.

Why it’s Important:

In traditional finance, instruments are constantly packaged and included within fund structures, which are then themselves invested directly by allocators. In this sense, we see vaults as one of the clearest signs that DeFi is productising and maturing. This shift matters because it mirrors how real-world investors already consume yield, via wrappers, and in our view, it’s a prerequisite for mainstream distribution. It also forces the market to confront the real table stakes for scaling into the billions, those being institutional-grade risk management, monitoring, and transparency. We don’t see the next phase of vault growth coming from just higher APY potential, but rather enhanced process and better controls. As the vault product layer matures, it also becomes the natural interface for disclosures, reporting, suitability, and distribution partnerships, which is exactly what broker-dealers and wealth platforms need before they can offer these products confidently.

Our Prediction for 2026: 

As shown in our included chart, onchain vault AUM grew in 2025 by 73% to $11.84b at the beginning of the year. We see this as the year vaults matured, and expect 2026 will be the year that AUM growth accelerates, and real adoption takes place, as participation in, and coverage of, the fund structures expands. This would be the result of high-quality curators attracting the majority of inflows, both from proven existing players and new, reputable, entrants. Our headline prediction here is that, buoyed by the growth in RWAs, regulatory clarity and institutional participation catalysed by institutional-grade product and risk management development, onchain vault AUM will 3x in 2026 to $36b before YE. In conjunction with this, and arguably a prerequisite, we predict that at least one major broker-dealer offers and curates an onchain vault yield shelf, and multiple fintechs will embed what is commonly referred to as the ‘DeFi mullet’, with onchain vaults in the back-end, and fintech ‘earn’ UI in the front-end. Over time, we also expect a winner-takes-most dynamic across curators and rails, whereby distributions, trust and risk systems become the moat. 

What Data We’re Looking At:

Hyperliquid is a name you’re likely exhausted hearing about from 2025, and for good reason, they put onchain perpetuals on the map, all while epitomizing the app-chain thesis and arguably rewriting the onchain revenue playbook. This chart captures the essence behind this, in total perpetual futures Open Interest (OI), while also highlighting new entrants to the game in Aster, Lighter and edgeX. We’re deliberately using total OI, as opposed to beta OI, because the most interesting part of the onchain perps story is no longer ‘crypto majors moving onchain’, it’s whether these venues can scale into new asset classes that are making their way onchain, i.e. equities, commodities, indices, and still hold risk in size. OI is the cleanest single metric for this given it measures the stock of conviction, this being how much risk traders are willing to keep open, rather than the mere flow of activity we see with volumes data. 

Why it’s Important:

We believe onchain perpetuals could be seen as one of the strongest examples of PMF in crypto, given they deliver the simplest action traders want to undertake, clean, continuous leverage, in an onchain primitive format. It’s somewhat of a provocative shift, in that as RWAs come onchain, the default path has historically been tokenisation, but most tokenisation today is skeuomorphic, in that it replicates the old world on our new rails, without taking full advantage of our improved infrastructure. The Perpification of these assets is the more crypto-native route, providing pure exposure while abstracting away traditional finance friction. We believe that equity perps are the clearest frontier. They offer genuine novelty for both TradFi and DeFi, with 24/7 access and easy leverage benefiting the former, and RWA migration onchain benefiting the latter. The deeper implication here being that perps are a potential universal wrapper for risk expression. If you believe markets trend toward instruments that maximise capital efficiency and minimise operational friction, perps are where trading points towards.

Our Prediction for 2026: 

As for our onchain perpetual prediction, it’s important to contextualise the fact that futures trading volume is already flooding onchain at a decent rate. In 2025, the DEX to CEX futures trading ratio grew >3x from 6.34% to 21%. We expect onchain perp OI to compound materially again in 2026, but the more important prediction is what that OI represents. We see this as the year that perps become the default onchain wrapper for trading all assets, which means an expansion of perp exposure to RWAs, in line with the RWA expansion predicted above. Our headline prediction here is that weekly perp OI increases magnitudes beyond its 2025 ATH, exceeding a high of $50b in 2026. The risk to this prediction is the choice of OI as a proxy metric over volumes. As OI grows, so does reflexivity, and if we see volatile markets that result in liquidation cascades, the tell on this prediction will be whether OI rebounds faster after stress events, which would signal maturity in the instrument. 

What Data We’re Looking At:

Buybacks proved controversial in 2025, with many major teams, namely Hyperliquid, Raydium and Pump.Fun, employing multi-million-dollar programs with varying success. The chart above shows the cumulative token buyback spend of the top 10 buyback programs throughout 2025. You’ll notice clear spikes indicating that buybacks are a cyclical behaviour, driven by revenues and token price action, which we believe to be a failure in programmatic designs. We covered the mechanics, trade-offs, and best practices in our Keyrock report, ‘Designing Token Buybacks’.
 
 

Why it’s Important:

Without clear revenue share legislation, buybacks have become crypto’s most legible capital return mechanism. They’ve become a fairly blunt, but credible way to demonstrate alignment and deliver value back to holders. While tokenholder revenue has surged, in which our research shows a >5x increase since 2024 and record highs of $800m in Q3 2025, buybacks are still faulty in practice given most programmes are reflexive and pro-cyclical. By this, we mean they spend too aggressively near peaks and underspend in downturns, when buybacks would be most efficient. We also see inefficiencies in market microstructure, with many projects executing buybacks with taker orders, which remove liquidity and can amplify volatility, whereas maker-style execution calibrated to organic volume can reduce impact and improve average price. We think there’s room for change here. 

Our Prediction for 2026: 

The preface to this prediction is that we believe in 2026 more protocols will reach the stage in which they have consistent revenues, clearer compliance footing, and a growing need for a credible tokenholder value-return mechanism. As such, we expect token buybacks to broaden and mature simultaneously. The number of protocols running meaningful open-market repurchase programmes should increase by roughly 50-100% versus 2025. Adding more nuance, we predict that the total weekly buyback spend reaches at least 2x that of 2025. An unquantifiable, but more important shift, is how buybacks are actually executed. We expect leading teams to move away from naive ‘fixed percentage of fees’ approach, and towards value-aware approaches, such as smoothed treasury policies, e.g. pacing bands, trigger frameworks, and more disciplined accumulation.

What Data We’re Looking At:

The chart we’re looking at here shows a fairly bleak story of how Solana-based MEV, measured as both validator and Jito tips, has declined significantly over the course of the year, from a peak of 100k SOL, equivalent to $25m at the time, to a low of 1k SOL, equivalent to $139k. This chart isolates the most visible MEV payment rail on Solana today, that being the searchers paying tips to secure priority and favourable execution, with those payments ultimately accruing to validators. The split is important because it lets us see not just the level of MEV, but how MEV is being monetised and routed through the stack, directly to validators versus through Jito’s auction pipeline. 
 
 

Why it’s Important:

This will be an important chart this year for one reason only, the fact that the Block Assembly Marketplace (BAM) was announced earlier this year. You see, today, a  large share of MEV is expressed as tip-based payments, which are highly reflexive, spiking during periods of heavy trading, memecoin mania, and liquidation cascades, then compressing as activity normalises. BAM introduces a block-building architecture that brings, above all else, customisability to how transactions are processed on the Solana network. In layman’s terms, transactions routed through BAM are privately sequenced inside trusted enclaves and can be ordered using application-defined sequencing rules before being passed back to Solana’s validator leader for execution with public attestations. By making ordering private until execution, BAM should reduce malicious MEV like front-running and sandwiching, while still preserving auditability of the final ordering. More importantly, the programmable sequencing plugin layer unlocks new market structures, think onchain CLOBs, priority workflows, JIT oracle updates, and dark-pool style execution. We believe that BAM will be adopted heavily at the protocol level this year on Solana, and will result in material shifts in how Solana MEV is captured. 
 
 

Our Prediction for 2026: 

We anticipate that, driven by the adoption of BAM on Solana at the protocol level, a larger share of the MEV budget should move upstream into explicit execution pricing set by applications and venues. That shift changes who gets paid, how predictable those revenues are, and whether MEV feels like a hidden tax or a transparent fee for execution quality. At the center of our outlook, we believe the easiest forms of extractive MEV will be materially suppressed, and the reflexive spikes in validator and Jito tips that have historically appeared during mania periods will become less frequent and less extreme. So, we anticipate a new category appearing on this chart, and dominating, that being the application-based MEV, while tip-based payments become a smaller share of total Solana economic activity. Crucially, if BAM meaningfully enables customised transaction ordering, we should see application-level revenues rise as protocols monetise execution quality directly.

What Data We’re Looking At:

One of the best performing altcoins of the year was ZEC, riding on the curtails of the privacy narrative driven by a broader push for financial confidentiality. This chart measures privacy through actual onchain behaviour by depicting the amount of ZEC deposited to Zcash’s shielded pools, broken out by Sprout, Sapling, and Orchard, all iterations of shielded Zcash pools. The key idea here is straightforward, in that the larger the shielded pool is, the larger the anonymity set, and, in practice, the stronger the privacy guarantees for users who transact privately. Orchard is the latest pool enabled by NU5, built on Halo, which in layman’s terms means it has removed trusted setup assumptions. Orchard is paired with Unified Addresses that make shielding easier to use in real wallets such as the famed Zashi.
 
 

Why it’s Important:

In traditional finance, opaqueness is built-in by design. We believe that radical transparency has a cost, and that cost rises as more real financial workflows move onchain. In onchain finance, as greater volumes of traditional finance capital enters the ecosystem, it starts to become table-stakes market structure to offer privacy. That’s what makes shielded supply such a useful proxy, it tells us whether privacy is just an ideology, or whether users are actively choosing confidentiality as a feature of onchain finance. Perhaps the most important aspect here is that there’s a flywheel in that a larger anonymity set generally makes privacy stronger, which can encourage further shielding and private usage. Of course, this can cut both ways, with privacy being a regulatory sticking point of late, particularly with recent moves like Dubai’s financial regulator banning privacy tokens on regulated venues. 

Our Prediction for 2026: 

We anticipate that privacy demand will become more structural in 2026 as onchain finances continue to grow and the ‘radical transparency tax’ becomes harder to ignore. We see this sticking point forcing regulatory clarity, and proxies like Zcash that allow for transparent transactions for certain compliant purposes prevailing. We predict that the total shielded ZEC amount rises from 4.9m today to >7m ZEC by YE 2026, driven primarily by Orchard-led shielding and the anonymity-set flywheel we described earlier, in which more shielded value strengthens the privacy value proposition, thus furthering the willingness to use the product. The key risk to this prediction is that the regulatory pressure swings out of favour, suppressing access via compliant venues, but we see selective-disclosure primitives as the path that lets privacy remain viable as market structure.

What Data We’re Looking At:

Zero’ing in on Ethereum, we’ll turn our attention to its scalability roadmap in L2s, which came under scrutiny this year due to questionable economic compensation to the L1. The roadmap made blobspace the default lane for L2 data availability after Dencun, but in line with the compensation debate, the economics are still evolving. This chart compares the cost of posting the same amount of data using blobs vs calldata, and shows how blob pricing behaves in practice. It shows long stretches where blobs are extremely cheap, in fact often close to zero, punctuated by episodic spikes when demand surges.

Why it’s Important:

As mentioned, the chart highlights a core tension in Ethereum’s scaling roadmap, in that if L2 activity grows but blob pricing spends long periods near negligible levels, then Ethereum’s data availability product can be heavily utilised without translating into consistent protocol revenue. This is why we’ve seen proposals like an execution-linked blob fee floor gain attention this year, acting as a mechanism to ensure blobspace carries a meaningful minimum price so that rising L2 throughput is reflected in Ethereum mainnet’s economics. Our thesis is that Ethereum’s rollup-centric roadmap only works as an investment and a protocol if Ethereum can monetise being the data-availability layer for L2s. Monetisation improves when blobspace carries a meaningful price floor and demand becomes steady, whether that’s through protocol changes that reduce prolonged near-zero pricing, e.g., an execution-linked blob fee floor, or simply through sustained growth in rollup demand that keeps blobspace consistently utilised.

Our Prediction for 2026: 

In 2026, we expect blob pricing to shift such that we see a higher and more persistent baseline, consistent with our thesis that Ethereum increasingly monetises DA as a core product. We predict that the median hourly blob cost rises to ≥$0.05 per blob on a full-year basis, and that the 30-day rolling blob vs calldata cost ratio averages ≥10% for at least one sustained 60-day stretch. If this plays out, it will mean that blobs remain cheaper than calldata, but are no longer a financial burden for Ethereum most of the time.

What Data We’re Looking At:

This chart from @obchakevich tracks the monthly spend volume processed through crypto-linked payment cards, segmented by card provider. The story here is clear and intriguing, since the beginning of 2025, at only $17m in monthly spending volume, crypto-linked cards have exploded in volume by 6.2x to $106m in weekly volume, transitioning into a meaningful consumer payments segment. The vertical is currently dominated by Ether.fi Cash and Cypher, though is still up for contention, with several mid-sized players commanding high-single-digit percentage of market share.  
 
 

Why it’s Important:

Crypto cards are one of the few places where onchain value becomes real-world utility without asking users to change their existing behaviour, thus making them a relatively frictionless onboarding route. The trojan horse, as it were, is the back-end of these cards, underpinned by stablecoins, crypto wallets and blockchain settlement rails. We see this vertical as a meaningful catalyst for stablecoin growth, and believe that despite the heavy growth in 2025, this vertical is in its infant stages. 

Our Prediction for 2026: 

We expect crypto card spend volume to compound again in 2026, with total monthly spend reaching $500m at least once during the year, a 5x on current highs, and ending 2026 at a $0.2-$0.6b monthly run-rate. The catalyst here simply being UX improvements, with cards embedded inside wallets, more stablecoin-native balances, increasingly competitive utility-based value propositions, and protocol incentive schemes bolstering onchain yield potential. We believe the market structure will reflect the existing payments market, with a winner-takes-most dynamic, though given the competitiveness of the space, will not back a horse this early in the year. 

What Data We’re Looking At:

Institutional flows, specifically spot BTC ETF flows, have been incredibly volatile this year, largely reflecting market sentiment. Nevertheless, BTC held within these vehicles increased 25% in 2025, as can be seen on the above chart, which tracks the amount of BTC held by US spot Bitcoin ETFs, using onchain holdings as the ground-truth measure of AUM, broken out by issuer. The key signal is the steady rise in aggregate holdings, despite sporadic pullbacks, alongside a two-horsed race between BlackRock and Fidelity to capture the bulk of new BTC accumulation. We see this chart as effectively as a live scoreboard of BTC holdings in traditional portfolios. 

Why it’s Important:

In 2025, we believe we saw inflows into US spot BTC ETFs transform from a launch-year phenomenon to a structural bid with consistent, gradual growth in AUM. This trend has continued into the early days of 2026, with the vehicles pulling in $1.2m in just a matter of days, running at $150b+ projected accumulation rate. Of course, we don’t expect steady accumulation all year. The point here is that ETFs have the potential to change market microstructure, with the marginal buyer becoming steady and rules-based, which makes Bitcoin feel less like a speculative trade and more like a standard sleeve in multi-asset allocation.

Our Prediction for 2026: 

We don’t expect it to be linear, but we predict that BTC ETF AUM will continue its growth to surpass 2.5m BTC this year, with net inflows being positive in at least 8 of 12 months. Structurally, however, we expect concentration to diversify as new brands enter the race, this has already begun with Morgan Stanley filing for a BTC, and SOL, spot ETF in the early days of 2026.

What Data We’re Looking At:

Throughout 2025, Aave USDC borrow APY on Ethererum, a proxy for stablecoin lending markets, fluctuated between 2.4% and 9.8%. The above chart tracks this USDC variable borrow rate on Aave, alongside a 30-day rolling volatility measure. The purpose of this volatility measure is to tell us whether that cost is becoming predictable enough to underwrite longer-duration strategies. 

Why it’s Important:

Within DeFi, a result of the lack of an existence of a money market rate means that stablecoin borrow rates sit underneath everything, from basis trades, vault strategies, structured products and the broader onchain credit stack. As a result, a volatile funding regime makes leverage fragile, in which we’d see rates spike, with either rapid de-risking or liquidations. What DeFi needs to mature is a stable funding regime, which would enable more institutional-style portfolio construction as financing becomes a reliable input. This is why rate stability is meaningful to watch this year, it’s downstream of the development we’re eyeing and building towards as an industry. 

Our Prediction for 2026: 

Our core prediction, having tracked and commented on these rates throughout 2025 in our weekly market commentary, is that onchain stablecoin funding becomes materially more stable in 2026. Concretely, we expect the 30-day rolling volatility for the USDC borrow APY to compress relative to its 2025 results, falling to an average of <0.25 in 2026, vs. ~0.40 in 2025, and remaining below 0.50 for at least 9 months of the year.This could only be the result of deeper stablecoin liquidity and arbitrage between onchain and offchain rates becoming more efficient.