KEYTAKEAWAYS
- Messari argues most Layer 1 blockchains face a valuation reset as high inflation, low fee revenue, and excess blockspace expose the limits of TPS-driven narratives.
- The report predicts AI agents will dominate on-chain activity by 2026, shifting crypto’s focus from human-facing interfaces toward APIs, automation, and machine-native financial infrastructure.
- DePIN and equity perpetuals emerge as core growth engines, with real revenue, derivatives liquidity, and system integration replacing speculative token demand as primary value drivers.
CONTENT
Messari’s 2026 Crypto Theses reveal a shift from speculation to system-level integration, highlighting L1 valuation risks, AI agents, DePIN revenue, and equity perpetuals.
Every year, the crypto market produces countless outlooks and predictions. Most are quickly forgotten. A few, however, shape how capital, talent, and attention actually move. Among these, Messari’s annual “Crypto Theses” report occupies a unique position.
For many institutions, this report functions less like a forecast and more like a strategic memo. It does not merely ask where the market might go; it implicitly defines what matters in the coming cycle. Historically, sectors emphasized in Messari’s theses often become the very areas VCs and builders pursue over the following year.
The 2026 edition is framed around a clear transition: crypto is moving away from pure speculation and toward system-level integration. In this first part of a three-part series, I focus on several core arguments from the report, combined with my own interpretation and skepticism, especially from a market participant’s perspective rather than a purely first-order research lens.
THE VALUATION TRAP OF L1 BLOCKCHAINS
Between 2024 and 2025, a wave of VC-backed Layer 1 blockchains launched with fully diluted valuations frequently reaching tens of billions of dollars. Names like Monad, Berachain, and Sei arrived with familiar promises: higher TPS, stronger teams, and superior execution environments. The implicit assumption was that each new L1 had a credible chance of becoming “the next ETH.”
That assumption, Messari argues, has now collapsed under the weight of real data.
The early valuation logic treated L1s as potential money. If a chain could plausibly become a settlement layer with enough activity, then a massive monetary premium seemed justified. In practice, however, most of these networks have turned out to be highly inflationary systems with minimal organic revenue. Token emissions vastly exceed on-chain fee generation, turning many L1s into structurally loss-making entities.
At the same time, the market environment has shifted. Ethereum’s Layer 2 ecosystem has matured significantly, while Solana has consolidated a dominant position in high-performance consumer crypto. Under these conditions, new chains struggle to attract loyal, long-term token holders. What they attract instead are airdrop farmers and short-term liquidity tourists.
Messari’s conclusion is blunt: outside of BTC and a small handful of ecosystems with genuine gravitational pull—Solana and Base being the most frequently cited examples—most L1 valuations have become completely detached from fundamentals.
Looking toward 2026, the report expects the market to actively strip away the so-called “monetary premium” from L1 tokens. Simply advertising high throughput will no longer justify multi-billion-dollar FDVs. At a minimum, a chain should generate more in daily gas fees than it distributes via inflationary rewards. If not, the math is unforgiving.
In extreme cases, some newly launched parallelized EVM chains still trade at FDVs between $5–10 billion, while producing less than $10,000 per day in gas revenue. At that rate, it would take thousands of years of fee generation to offset token issuance. This is not a temporary mismatch; it is a structural one.
While Messari appears optimistic on Solana—possibly influenced by its own exposure—the broader takeaway is more important: a viable L1 must possess either genuine “minting power” or some form of application-level monopoly. Speed alone is no longer a moat.
CHAIN ABSTRACTION AS A SURVIVAL STRATEGY
One area the report highlights, and which I believe deserves closer scrutiny, is chain abstraction.
The goal of chain abstraction is simple in theory but radical in impact. Users should not need to know which chain they are using. With a single wallet and a balance denominated in stablecoins, a user should be able to initiate an action while the system automatically handles bridging, gas conversion, routing, and signing behind the scenes.
If this vision succeeds, blockchains cease to be consumer-facing products and instead become backend infrastructure.
Projects such as Near and Berachain are experimenting with different versions of this idea. Near is positioning itself as an AI-centric distribution layer, while Berachain enforces capital stickiness through its “liquidity consensus” model, effectively compelling liquidity to remain within its ecosystem.
From an investment perspective, this reframes how L1s should be evaluated. A chain that merely processes transactions faster is replaceable. A chain that controls user flow, liquidity routing, or application distribution retains leverage. In a post-abstraction world, power shifts away from execution speed and toward coordination and control.
THE RISE OF THE AGENT ECONOMY
Perhaps the most controversial claim in Messari’s report is its projection that by 2026, the majority of on-chain activity will no longer be driven by humans.
Instead, AI agents will dominate transaction volume.
The logic is straightforward. Traditional banking systems cannot open accounts for autonomous software agents. Yet AI systems increasingly require continuous, 24/7 access to payments, hedging instruments, and yield optimization. Crypto-native assets, particularly stablecoins, are uniquely suited to this role.
If AI agents become economically autonomous, they will pay one another, rebalance portfolios, and seek optimal execution paths without human intervention. Messari estimates that as much as 80% of on-chain transactions could be machine-generated within the next cycle.
This shift has profound implications. User interfaces, dashboards, and retail-friendly design lose importance. APIs, smart contract composability, and machine-readable financial primitives become the real battleground.
Protocols like Virtuals Protocol and Wayfinder are early attempts to build infrastructure for this world. Virtuals positions itself as a launchpad where AI agents can possess both autonomous identity and treasury control. Wayfinder focuses on navigating complex on-chain operations on behalf of agents.
However, the real opportunity may lie elsewhere. Rather than speculating on abstract “AI tokens,” Messari implicitly encourages attention toward AI consumables: gas optimization layers, identity verification systems for agents, and infrastructure that AI must pay to use. These are closer to picks-and-shovels than narrative plays.
EQUITY PERPETUALS AND THE NEW DERIVATIVES FRONTIER
Another major theme is the emergence of equity perpetual contracts.
Following the success of protocols like Hyperliquid, DeFi is expanding beyond crypto-native assets and toward global equity price exposure. Equity perpetuals are synthetic derivatives whose prices are anchored via oracles and funding rates, not ownership.
This distinction matters. Trading NVIDIA equity perps does not grant dividends or shareholder rights. It is purely a wager on price movement, enforced through funding rate mechanics. If the underlying stock rises, shorts pay longs; if it falls, the reverse occurs.
Messari contrasts this model with tokenized stocks, which in theory represent ownership but in practice suffer from thin liquidity, opaque custody, and platform risk. For now, equity perps appear far more scalable.
If this market truly expands in 2026, Messari believes Hyperliquid’s surrounding ecosystem could see exponential upside. Competing models also exist. Aster, often described as closely aligned with Binance’s ecosystem, takes a liquidity aggregation approach across multiple chains rather than building a vertically integrated L1.
Hyperliquid prioritizes on-chain transparency and performance by owning the entire stack. Aster prioritizes capital efficiency and convenience, allowing users to deploy leverage across chains with minimal friction. In bull markets, the latter’s appeal is obvious. Yet its architectural complexity also introduces higher systemic risk.
DEPIN AND THE TURN TOWARD REAL REVENUE
Finally, Messari identifies DePIN as the only sector it expects to generate hundreds of millions of dollars in verifiable revenue by 2026.
This is not a claim without controversy. DePIN narratives surged before and largely failed due to one-sided supply growth without real demand. Many early projects incentivized hardware deployment without solving the question of who would pay for the service.
The report’s argument is that this equation is changing. Infrastructure is already deployed. What arrives next is demand—driven primarily by AI compute shortages.
Projects like Render and io.net focus on GPU aggregation rather than speculative data harvesting. The key metric is external revenue: are token buybacks funded by real enterprise contracts, or merely by new participants buying equipment?
Messari points to cases like Aethir, whose reported annual recurring revenue exceeded $160 million in Q3, largely from AI compute renters priced out of owning high-end hardware. Whether such numbers persist remains to be seen, but the framework is clear.
For DePIN, revenue quality—not network size—will determine survivability.
A FINAL CAVEAT
Despite its depth, Messari’s report reflects a predominantly first-market perspective. For retail participants, fundamentals alone rarely drive price action. Liquidity and narrative still dominate returns.
Usage does not equal upside. Marginal capital flows do.
This tension—between what should matter and what actually moves markets—will define the next cycle. And it is precisely where critical reading of reports like Messari’s becomes most valuable.
To be continued in Part 2.
The above viewpoints are referenced from @Web3___Ace
Read More:
