Why AI agents can’t do without blockchain? Pantera reveals structural opportunities in the convergence of technologies

The global technology asset market in 2026 is unfolding a valuation divergence that is historically rare. Driven by leading companies such as OpenAI, Nvidia, and Microsoft, the AI sector continues to surge, and AI-related companies account for about 45% of the market-capitalization weighting in the S&P 500. However, the cryptocurrency asset market has gone through a sustained pullback, with the total market capitalization of the global crypto market falling by more than 40% from its 2025 peak. Beneath the appearance of this ebb and flow, Pantera Capital’s latest research reveals a structural problem that is even more worth deeper scrutiny: on the valuation dimension, the two asset classes have pulled apart to the greatest distance in history, yet their underlying technologies and future ecosystems are undergoing an unprecedented deep integration. This contradiction is not only a rupture in market narratives, but also a test of investors’ cognitive frameworks.

The real scale of valuation divergence: what economic logic is hidden behind the numbers

Just how severe has the valuation divergence become? According to Pantera Capital’s conclusion, derived from its own internally compiled leading AI company index and a long-term Bitcoin pricing model: as of May 2026, the premium of the leading AI company index versus its own logarithmic trendline over the past four years is about 33%, placing it clearly in a “fully priced” and even “overheated” range; another more aggressive AI index set shows that the valuation premium has reached 49%. At the same time, Bitcoin is trading at a discount of about 42% versus its own four-year logarithmic trendline. Pantera CEO Dan Morehead describes this divergence as “the largest valuation split ever recorded.”

The two sets of data are like the two ends of a scale: one end is the ongoing overvaluation of the AI sector, and the other is the structural undervaluation of crypto assets. From a macro narrative perspective, AI is unquestionably the “main character” in today’s capital markets—about 61% of global venture capital flows into the AI space, and market sentiment is almost one-directionally tilted toward tech giants and technological optimism. Meanwhile, crypto assets are enduring the tail-end pain of a bear market: as of June 5, 2026, Bitcoin’s price is $61,983.42, with a total market capitalization of approximately $1.24 trillion, and the Fear & Greed Index is at an extremely low level. But do these emotional divergences truly reflect actual value differences in the underlying technologies? The answer may not be that simple.

The deep drivers of premium and discount: how capital-flow structures tear apart these two asset classes

The first and primary driver of valuation divergence is an extreme, selective tilt in capital allocation. In the past two years, institutional capital has poured in on a large scale into leading AI companies—ranging from Nvidia’s astonishing performance after breaking through a $5.43 trillion market cap, to the funding frenzy in the private market for startups such as OpenAI, and global venture capital of about $258.7 billion being concentrated into the AI track. The rise in AI stock prices and the influx of institutional allocation funds have formed a positive feedback loop.

Crypto assets, in contrast, are in the opposite capital structure. Dan Morehead points out that most large investment institutions currently do not materially hold crypto assets. However, within the same group of institutions, about 79% plan to allocate to crypto assets within the next three years, and about 65% view crypto assets as a tool for portfolio diversification. This reality of “awareness in place but allocation missing” is precisely the source of potential demand going forward. Based on the total management volume of global institutional assets, even if only the allocation proportion rises from close to zero by 2%, it would still imply new inflows of capital on the order of hundreds of billions of dollars. Therefore, the current undervaluation of crypto assets is less about a lack of value recognition and more about a phase lag in the timing of capital allocation.

Why AI and blockchain must integrate: the complementary logic built on four foundational pillars

Putting aside short-term capital sentiment, the technological interdependence between AI and crypto assets is deepening at an unprecedented pace. Pantera partner Paul Veradittakit systematically lays out four pillars of AI and blockchain integration—payments, identity verification, open systems, and resource aggregation—and in all four areas, commercial deployment has already been achieved.

Specifically: in the dimension of trading and payments, OpenFX uses stablecoins as the underlying settlement layer, and its annual total transaction volume has already surpassed $60 billion; in the dimension of identity verification, the World project has verified more than 18 million real users and has partnered with platforms such as Tinder, Reddit, and Zoom to address the identity crises brought about by AI deepfakes; in the dimension of open systems, protocols represented by Bittensor are building a market for machine intelligence, pushing for algorithm democratization; and in the dimension of resource aggregation, DePIN networks (decentralized physical infrastructure networks) are aggregating idle GPU compute power worldwide, redefining the supply-and-demand landscape for AI infrastructure. AI agents are becoming native economic entities on-chain. To date, AI agents have completed more than 176 million transactions on-chain, with a cumulative settlement amount of $73 million, and an average transaction value of only $0.31. This data shows that the fixed-cost structure of traditional financial payment rails fundamentally cannot serve micro-payment scenarios, whereas stablecoin solutions naturally come out on top.

How much gap is implied between short-term institutional coolness and long-term demand

Although the trend of AI and crypto integration is becoming increasingly clear, the pace of institutional capital allocation appears to be slow. A Q2 2026 report jointly released by Coinbase, an exchange, and Glassnode, an on-chain data provider, shows that about 75% of institutional investors believe Bitcoin is currently undervalued; however, on the sentiment level, about 82% of institutions believe the current market is in a bear market or the late stage of a bear market. This “bullish but bearish” contradictory mindset directly maps to a long-term lack of institutional allocation exposure in the crypto space.

In its analysis, Dan Morehead also notes that Bitcoin’s own structural four-year cycle suppresses near-term price movement. The halving event in April 2024 has already been two years ago. Historical experience shows that 12 to 18 months after a halving is an active window for price discovery, yet the market is still digesting the high-pressure overhang from the previous cycle’s peak and the capital overflow effects. By overlaying institutional demand curves with market supply cycles, a basic judgment can be reached: the valuation repair of crypto assets will not unfold in a linear manner, but will be driven jointly by two key variables—an institutional allocation inflection point and a macro liquidity turning point.

How macro narratives reshape crypto asset pricing logic

When analyzing the medium- to long-term pricing logic of crypto assets, it is not enough to focus only on capital flows between sectors and valuation gaps. Dan Morehead describes crypto assets as a hedge against “the erosion of paper currency value.” In a macro environment where inflation pressures persist and money supply expands, Bitcoin’s fixed total supply constructs scarce-value characteristics that are structurally comparable to gold. In the first half of 2026, factors such as rising geopolitical risks and increased volatility in the bond market further reinforce investors’ focus on scarce assets.

On a longer time scale, the technological fusion of AI and blockchain is giving rise to an “agent economy,” which is redefining the relationship between the subjects and objects of economic transactions from the ground up. Research institutions predict that the agent economy could bring structural demand for SOL on the scale of hundreds of billions of dollars. This means that the valuation framework for crypto assets is shifting from “asset speculation” to “infrastructure value capture.” Macro narratives and technical implementation are forming a coordinated resonance along two tracks: AI is spawning new economic entities (autonomous agents), and the economic behaviors of these entities inherently require blockchain as the underlying financial infrastructure and value-settlement layer.

Where do the structural opportunities in the fusion cycle go next?

Taken together, the core contradiction in the current valuation divergence between AI and crypto assets is not which sector has brighter technological prospects, but rather the market’s core capability to set prices.

From historical experience, every major economic inflection point brought by technological change often originates at the intersection of two disruptive technologies—steam engines and railways, electricity and manufacturing, the internet and smart phones, and so forth. Today, AI creates nearly unlimited supply (content, agents, computing power), while blockchain anchors a fixed, verifiable scarcity and ownership—together forming a naturally complementary structure. Global capital’s attention to AI is understandable, but AI is only half of the picture of industrial transformation. AI agents do not open bank accounts, do not use Fedwire or ACH for cross-bank clearing, yet they must still complete large-scale economic coordination and value transfer. This gap is currently the only one that blockchain has the capacity to fill.

Although short-term market sentiment remains defensive, given the historical extremes of valuation divergence, the recognition gap in institutional allocation, and the accelerating real-world deployment of AI×blockchain integration, the medium- to long-term value of crypto assets has been systematically underestimated. The large-scale explosion of the agent economy will drive structural demand for crypto infrastructure, and this is being validated by an increasing number of empirical data points. When the market gradually recognizes that AI and blockchain are two sides of the same industrial transformation, the current valuation divergence may be headed toward a directional rebalancing.

Summary

Pantera Capital’s research reveals one of the most core contradictions in today’s market: the AI sector’s valuation is at a historic high, with premiums ranging from 33% to 49% over leading indices, while Bitcoin is trading at a discount of 42% versus its long-term trend, leaving crypto assets in an undervalued range. Behind this valuation divergence is not a difference in the technological outlook of the two major sectors, but a misalignment between short-term capital allocation tendencies and the logic of long-term technological integration. Deep integration between AI and blockchain is accelerating around four pillars—payments, identity verification, open systems, and resource aggregation—where AI agents, as a new type of economic entity, are becoming native on-chain users. This trend is validating their complementarity: AI addresses supply abundance, and blockchain anchors scarcity and verified ownership. From multiple dimensions—including the recognition gap in institutional allocation, empirically supported real-world deployment of the technology integration, and macro narratives of scarcity—the logic for a medium- to long-term valuation repair in crypto assets has solid foundations.

FAQ

How are the “AI index premium of 49%” and “BTC retracement of 42%” calculated in the Pantera report?

Pantera’s method is based on pricing evaluation using each asset’s own logarithmic trendline over the past four years. The current price of the AI index is about 33% above its four-year historical trendline (some more aggressive AI index calculations reach 49%), indicating that the AI sector is in an overbought/overheated range. Bitcoin is trading at a discount of about 42% versus its own four-year logarithmic trendline, placing it in a significantly undervalued range. This comparison reflects a historical divergence in valuation levels between the two sectors.

What exactly are the “four pillars of AI and blockchain integration”?

The four pillars are: payment settlement (microtransactions by AI agents and stablecoin applications), identity verification (solving issues of AI forgery and machine-identity verification), open systems (building machine intelligence markets and algorithm democratization platforms), and resource aggregation (DePIN decentralized compute networks aggregating global idle GPU resources). All four areas have achieved commercial deployment to date.

What is the attitude of institutional investors toward allocating to crypto assets?

The data shows that about 75% of institutional investors believe Bitcoin is currently undervalued, and about 79% plan to allocate to crypto assets within the next three years. However, actual allocation exposure remains low; most institutions are in a state of “recognition with action delayed.” This contrasts with the rapid influx of institutional funds into the AI sector, and also represents room for growth in future demand for crypto assets.

Will the integration of AI and crypto assets bring new regulatory challenges?

The integration of AI and crypto assets involves multiple overlapping areas such as cross-border transactions, digital identity management, data privacy protection, and regulation of compute markets. At present, the global regulatory framework is still evolving. Because there are no unified international regulatory standards, different jurisdictions may adopt differentiated regulatory paths—this is an uncertainty that needs ongoing attention as this field develops.

How much impact will the agent economy have on the long-term value of crypto assets?

At present, there is already empirical evidence: AI agents completed more than 176 million on-chain transactions in the past 12 months, with a settlement amount exceeding $73 million. Independent research varies significantly in its predictions of the agent economy’s scale, ranging from structural demand on the order of hundreds of billions of dollars to larger tail effects at even higher magnitudes. The general consensus is that as the number of autonomous agents grows exponentially, their demand for on-chain financial infrastructure will rise in a systematic manner.