The Future of NFTs
When I first started digging into on-chain data for NFT markets back in 2021, I'll be honest — I thought I was looking at a bubble with a very fancy coat of paint. Floor prices were doubling weekly, gas fees were absurd, and half the "utility" I saw amounted to a JPEG and a Discord invite. But three years of watching wallet flows, building valuation models, and integrating NFT price feeds into our risk engines at JOYFUL CAPITAL have changed my mind. Not about the hype — the hype was mostly noise — but about the underlying infrastructure. NFTs are evolving from speculative collectibles into a genuine financial primitive, and the next five years will determine whether they become boring, essential plumbing or fade into a cautionary tale.
For readers who need a quick refresher: a non-fungible token is a unique digital asset recorded on a blockchain, proving ownership and provenance of anything from digital art to real estate deeds to carbon credits. The market peaked dramatically in 2021–2022, with global NFT trading volume exceeding $25 billion in 2021 according to DappRadar, then collapsed by more than 90% into 2023. Most people stopped paying attention right there. That was a mistake. The interesting part — the part where NFTs stop being a casino and start being infrastructure — is happening right now, in quiet corners: tokenized treasuries, on-chain identity, fractional real-world assets, and loyalty programs that actually retain users.
In this article, I'll walk through the aspects of NFT evolution that I believe matter most from a financial-data and AI perspective — not because I'm a maximalist, but because our models keep flagging these trends before the headlines do. I'll share a few things we got right, one thing we got badly wrong, and where I think this all lands by 2030.
Beyond the JPEG Era
The single biggest shift I've observed is the decoupling of NFT technology from NFT art. In 2021, roughly 90% of trading volume flowed through PFP (profile picture) collections. By 2024, that share had fallen below 40%, while utility-driven categories — tokenized assets, membership passes, gaming items, and identity credentials — absorbed the difference. This isn't a narrative I'm spinning; it's what the wallet-level data shows. The token standard survived the crash of its first killer app, and that's actually the strongest signal of technological durability you can get.
What does this mean practically? It means the ERC-721 standard (and its flexible cousin ERC-1155) is doing what good technical standards do: becoming invisible. When we built our NFT valuation pipeline at JOYFUL CAPITAL, we stopped segmenting by "collection" and started segmenting by cash-flow characteristics. A tokenized T-bill NFT and a Bored Ape have almost nothing in common analytically, even though they share a smart contract interface. Treating them as one asset class was our first big modeling error — more on that later.
The shift also changes who's paying attention. Art collectors dominated 2021 discourse; today the phone calls I get are from family offices asking about tokenized private credit and from gaming studios asking about interoperable item economies. Different audience, different due diligence, different everything. The NFT conversation has migrated from crypto Twitter to wealth management, and honestly, it's a lot less fun but a lot more consequential.
Tokenization of Real Assets
Real-world asset (RWA) tokenization is where I think the largest long-term value sits. The premise is straightforward: take an asset that's traditionally illiquid — a rental property, a piece of commercial real estate, a fine wine collection, an invoice receivable — represent fractional ownership on-chain, and suddenly you have 24/7 settlement, programmable distributions, and global access. Boston Consulting Group has projected the RWA tokenization market could reach $16 trillion by 2030, and while I take headline projections with a healthy grain of salt, the direction is credible.
We've seen real traction in specific niches. Tokenized U.S. Treasury products, for instance, grew from under $100 million in early 2023 to several billion by 2024, according to data from 21.co and RWA.xyz. These aren't speculative instruments — they're yield-bearing tokens backed by actual government securities, and institutional treasuries are using them for collateral management and cash sweeps. When BlackRock launches a tokenized money market fund, you know the plumbing phase has begun. That was a watershed moment I didn't expect to see so quickly.
Real estate tokenization has been slower and messier — regulatory fragmentation across jurisdictions is a genuine nightmare — but the deals are happening. I recall a mid-2024 conversation with a client who had structured a tokenized commercial property in a European jurisdiction; the paperwork for the token wrapper alone took seven months. That's not a technology problem. That's a legal and compliance reality that any honest RWA projection has to account for. Expect gradual, jurisdiction-by-jurisdiction progress rather than a big bang.
From a data-strategy standpoint, RWAs actually solve one of our biggest headaches: valuation. A PFP floor price is a social construct; a tokenized T-bill has an unambiguous NAV. The more of the NFT market migrates toward assets with underlying cash flows, the more our models can behave like traditional finance models — discounted cash flow, yield curves, duration analysis. That's a much more comfortable world for those of us who came up through quantitative finance.
NFTs as Digital Identity
The identity use case deserves more attention than it gets. Soulbound tokens — non-transferable NFTs that represent credentials, memberships, or attestations — are quietly becoming infrastructure for on-chain reputation. Vitalik Buterin's original 2022 paper on soulbound tokens laid out the vision: a wallet that carries your educational credentials, employment history, and community standing, all verifiable without a central authority. The implementations since then have been uneven, but the direction is clear.
Why does this matter for finance specifically? Because credit is fundamentally an identity problem. Traditional credit scoring relies on centralized bureaus; on-chain lending currently relies on overcollateralization because there's no reliable way to assess an anonymous borrower's history. Soulbound credential NFTs could, in principle, bridge that gap — a verified employment attestation here, a repayment history NFT there, building toward undercollateralized on-chain credit. We're years away from that being robust, but the building blocks are being laid in public.
The consumer-facing version is already here, albeit under different branding. Loyalty programs built on NFTs — Starbucks Odyssey was the famous, if short-lived, example; various airlines and retailers have experimented since — use non-transferable tokens to track tier status and rewards. The interesting design property is composability: a membership NFT from one brand can, if the issuer allows, unlock perks across partner ecosystems. That's something a traditional loyalty database simply can't do. Whether consumers care is an open question, but the technical capability is real and increasingly cheap to deploy.
AI and NFT Convergence
This is the aspect closest to my day job, so forgive the enthusiasm. The convergence of AI agents and NFTs is, I believe, one of the most underrated developments in the space. The basic idea: AI agents need identities and payment rails, and NFTs provide both. An autonomous agent that transacts on behalf of a user needs a verifiable identity (an NFT credential), a way to hold and transfer assets (a wallet), and a way to prove it acted within its mandate (on-chain attestations). The pieces fit together remarkably well.
We're already seeing early implementations. Several startups have launched "agent NFTs" — transferable tokens that bundle an AI model's configuration, permissions, and memory state, so that an agent can be sold or delegated like an asset. Whether that particular pattern survives is unclear, but the underlying need is genuine. In our own work at JOYFUL CAPITAL, we've been experimenting with AI-driven portfolio monitoring where an agent's access rights to client data are encoded as expiring NFT credentials — auditable, revocable, and granular. It's early, but the audit trails alone are worth it.
There's a flip side that deserves honest mention: AI-generated content has flooded NFT marketplaces, and provenance verification — supposedly NFTs' superpower — has struggled to keep up. Deepfake art, wash trading by bot networks, and synthetic collection activity are real problems. The same AI that powers useful agents also powers market manipulation at scale. Our fraud-detection models now dedicate roughly a third of their feature set to distinguishing organic wallet behavior from agent-driven activity, and that share is growing. It's an arms race, and I don't expect it to end.
Gaming and the Metaverse
Gaming remains the largest real-world testing ground for NFT utility, and the results are instructive. Axie Infinity's 2021 boom and 2022 collapse is the canonical case study: a play-to-earn economy that worked beautifully until the token emissions outran the underlying demand, at which point the whole thing spiraled. The lesson wasn't "NFT gaming is dead" — it was tokenomic design matters more than the technology. Games that treat NFTs as a feature (interoperable items, player-owned economies) rather than a business model have fared better.
The data supports a quieter resurgence. Despite the broader market downturn, gaming-related NFT transactions have held up better than collectibles, and several major studios — including traditional publishers, not just crypto natives — have shipped or announced blockchain-integrated titles. The pattern that seems to work: NFTs as optional ownership of in-game assets, with the game fully playable without them. The pattern that fails: requiring purchases to participate. That's an obvious insight in retrospect, but the industry took years to learn it.
Interoperability remains the great unfulfilled promise. The dream of carrying a sword from one game into another requires rival studios to agree on standards, and commercial incentives rarely align that way. I'm skeptical of full cross-game interoperability within this decade. What I do expect is intra-publisher ecosystems — one company letting items move across its own titles — which captures most of the player value without requiring industry-wide cooperation. Less romantic, more likely.
Regulation and Market Structure
Regulation is the aspect where I've had to revise my views most. In 2021 I assumed regulatory clarity would arrive quickly and settle things. Instead, we've had a patchwork: the EU's MiCA framework treats certain NFTs as regulated financial instruments depending on characteristics; the U.S. SEC has pursued enforcement actions against specific projects while offering limited general guidance; several Asian jurisdictions have moved faster in both directions. The practical result is that compliance has become a competitive moat — well-funded, well-advised platforms can navigate the maze, while smaller players get crushed by legal costs.
The market structure implications are significant. Institutional participation requires custody solutions, audited valuation, and legal certainty about asset classification — all of which are arriving, slowly. When we onboarded our first NFT-collateralized lending pilot, the single biggest bottleneck wasn't technology or data; it was the legal opinion on how to treat the collateral in a liquidation scenario. The boring infrastructure is being built, and that's exactly what mass adoption requires.
Tax treatment deserves a special mention because it's a mess. Different jurisdictions classify NFT gains as capital gains, ordinary income, or something in between, and the record-keeping burden on traders is substantial. This is one area where I'd genuinely welcome harmonization — not because I love regulation, but because ambiguity suppresses legitimate activity while doing little to stop bad actors. Clarity, even strict clarity, would be an improvement.
Market Maturation and Cycles
Let me share a personal failure, because it shaped how I think about this market. In early 2022, our models flagged rising wash-trading signatures in a mid-tier collection, and I dismissed it as noise — the collection's fundamentals looked fine on paper. Six months later it was down 95% and several of our data feeds had been contaminated by the fake volume. We had the signal and ignored it because the narrative was louder. I now insist that any NFT data pipeline we build includes wash-trade filtering as a first-class feature, not an afterthought. It's a small thing, but the lesson generalizes: in immature markets, data quality matters more than model sophistication.
The broader pattern I see is classic market maturation. The 2021 cycle was driven by retail speculation and reflexive hype; the current phase is driven by infrastructure, institutional capital, and utility. Liquidity has concentrated into fewer, higher-quality assets — which is painful for speculators but healthy for the ecosystem. Volatility is compressing, bid-ask spreads are narrowing on blue-chip assets, and lending markets are developing term structures. These are the vital signs of an asset class growing up.
I don't expect the next bull cycle to look like the last one. If NFTs return to headlines in a big way, it will likely be through a use case most people haven't fixated on yet — tokenized fund shares, on-chain credentials, or something in the AI-agent space. That's the nature of this market: the narrative that drives the next wave is usually not the one that drove the last. Our job in data strategy is to build frameworks flexible enough to catch it, rather than betting the house on any single story.
Conclusion
The future of NFTs is, in my view, less glamorous and more important than the 2021 mania suggested. The technology is maturing into financial and identity infrastructure: tokenized real-world assets with genuine cash flows, soulbound credentials enabling on-chain reputation, AI agents requiring verifiable identity rails, and gaming economies built on sustainable tokenomics. The speculative frenzy has given way to a slower, more durable build-out — one where compliance is a moat, data quality is the battleground, and institutional participation is the tell that something real is happening.
My recommendation for anyone working in this space, whether at a fund, a data provider, or a startup: invest in wash-trade-resistant data infrastructure now, treat NFT categories by cash-flow characteristics rather than collection branding, and watch the RWA and identity segments for the leading indicators of the next cycle. The research directions I'd prioritize are cross-jurisdictional valuation standards and robust on-chain credit scoring — both are bottleneck problems whose solutions would unlock significant value.
Looking forward, I suspect the most important NFT developments of the next five years will be almost invisible to retail — embedded in settlement layers, compliance tooling, and agent infrastructure. The token standard won't make headlines; it will just quietly do its job. And that, honestly, is the best outcome the technology could hope for.
JOYFUL CAPITAL Insights
At JOYFUL CAPITAL, our work integrating NFT data into financial models has taught us that the asset class's future hinges less on speculation and more on infrastructure quality. Our internal research indicates that tokenized real-world assets and identity credentials will drive the majority of institutional adoption through 2030, while speculative collectibles consolidate into a niche market. We've found that wash-trade filtering and multi-source valuation feeds are non-negotiable prerequisites for any serious NFT data pipeline — our own early mistakes made that painfully clear. For our clients, we recommend treating NFTs as a heterogeneous category requiring category-specific models: cash-flow-based valuation for RWAs, reputation-weighted scoring for identity tokens, and liquidity-adjusted marks for collectibles. We also see AI-agent identity as a sleeper use case worth monitoring closely. As the market matures, JOYFUL CAPITAL remains committed to building the data and analytical frameworks that let institutions engage with this space responsibly — grounded in evidence rather than narrative.