There is a list of real-world asset classes everyone agrees can be tokenized. Most of that list is wrong, or at least misleading at the ticket sizes most allocators actually deploy. After three years sourcing and structuring real-asset SPVs across the EU, here is the categorical map: twelve asset categories where tokenization economics genuinely work in 2026, five where the marketing has outrun the math, and the three tests every asset must pass before the on-chain wrapper is worth the overhead.
The map only makes sense after the filter. Before adding any asset class to the "tokenizable" list, the asset has to clear three tests. Most assets fail at least one. Most lists ignore that.
Test 1: predictable, contractually-defined cash flow. The asset must produce cash that an allocator who never meets the operator can model and underwrite. Rental income on a triple-net commercial lease. Operator rent on a battery container. Revenue per cycle on a documented working-capital structure. Royalty streams under publisher contracts. Cash flow that exists because of speculation (price appreciation, narrative momentum, "future adoption") fails this test regardless of how charismatic the founder is.
Test 2: deployable capital of at least €500K to €1M. Below this threshold the tokenization stack overhead (€2K-€8K platform setup, €5K-€25K legal for SPV incorporation, 0.5-1.5% per-raise issuance fees, KYC enforcement infrastructure) consumes too much of the per-unit margin. The economics only work above the threshold where the fixed-cost overhead amortises. Smaller deals should use a traditional private placement structure or a fund vehicle, both of which are cheaper to operate at small scale. The full economics are walked in the tokenization vs private placement guide.
Test 3: fits a standardised wrapper. The asset has to work inside one of the existing EU private placement structures (Romanian SRL, Austrian SPV, Maltese PIF, Luxembourg RAIF, Liechtenstein SPV) with a defined waterfall, defined allocator rights, defined reporting cadence, and an ERC-3643 token issuance pattern that the existing platforms (Tokeny, Polymath, Securitize) already support. Assets that require fully bespoke legal engineering can still be tokenized but they lose the operational benefit of the standardised stack and the cost advantage disappears.
Twelve asset classes pass all three tests cleanly in 2026. Five fail at least one and should not be tokenized regardless of the marketing. The rest of this guide walks each category.
Each category below passes all three tests, with the typical ticket size, the gross-yield range observed in the EU market, and a pointer to the relevant deal or guide on this desk.
Light-industrial, office, retail, and logistics property in EU jurisdictions with rule-of-law signals (DACH, Benelux, Nordic, regulated southern EU). The classic tokenization use case: a single illiquid building turned into tradeable fractions through an Austrian or Luxembourg SPV. Ticket-size threshold matters: under €2M the legal acquisition costs and ongoing SPV overhead consume too much yield. See the Austrian warehouse deal on the desk for a worked example of the structure.
Container or trailer battery units leased to licensed grid operators on fixed-rent contracts. Romania is the dominant EU jurisdiction because the day-night spread (15 to 25%) funds operator economics that support 15-20% lease rents. The asset stays on the investor's EU company balance sheet, with the operator capturing the spread and paying contractual rent. Two live deals on this desk: mobile BESS trailer at €80K and stationary BESS container at €500K.
Already-commissioned renewable installations with signed power purchase agreements (PPAs). The PPA provides the contractually-defined cash flow that the first test requires. Subsidies and feed-in tariffs add stability. The tokenized SPV pattern (Maltese PIF or Luxembourg RAIF wrapping the operating company) opens fractional access to assets that previously sat in dedicated infrastructure funds. Yields are lower than BESS but the cash-flow visibility is longer.
Networks of EV charging stations with utilisation contracts (corporate fleet operators, fuel retailer partnerships, municipal contracts). Revenue is per-kilowatt-hour delivered, modulated by utilisation. Higher operator risk than passive real estate but the network effect of pre-signed corporate users de-risks the cash flow meaningfully. Genuinely interesting category for 2026-2030 because the underlying market is growing 25-40% per year and the tokenized structure supports the fractional capital needed to scale.
Pools of loans to mid-market borrowers, structured as a tokenized fund vehicle. Janus Henderson's Anemoy products on Centrifuge are the high-end example (~$1B+ in tokenized treasury and CLO funds, see the Centrifuge RWA Roast for the structural critique). The tokenized SPV serves the same allocator base as a traditional private credit fund with marginally cleaner secondary mechanics. The category works at scale; below €5M the platform overhead does not amortise.
Short-cycle financing of trade receivables, export invoices, or import letter-of-credit positions. The classic working-capital pattern wrapped in a tokenized SPV. Cycles are short enough to course-correct (test 1 satisfied through documentation of completed cycles), capital is large enough for the wrapper to amortise (test 2 satisfied above €500K), and the structure fits the standard SRL or SPC pattern (test 3 satisfied). Centrifuge's original Tinlake stack served this category before they pivoted to fund vehicles.
Single-asset working capital structures (inventory cycles for commoditised goods, raw material pre-financing, equipment trading cycles). The Romanian solar trading SPV on this desk is structured as a single-investor SRL rather than a tokenized pool specifically because the €121K ticket sits below the tokenization threshold. Above €500K of pooled capital the tokenized pool pattern works well; below that, a direct SRL is cleaner.
Portfolios of apartment buildings managed as a single tokenized REIT or REIT-equivalent. The tokenization stack pays off above aggregate portfolio value of €10M because the per-unit overhead becomes negligible at scale. GromaCoin in Boston is the high-profile example (see the GromaCoin RWA Roast for the detailed critique of the structure at sub-scale). Done at proper scale with audited financials and a real secondary mechanism, multifamily tokenization works. Done at €5M with 200 holders it does not.
Tokenized exposure to data center buildings with signed multi-year tenant contracts (hyperscalers, enterprise cloud providers). The underlying asset is real estate plus high-spec mechanical and electrical infrastructure. The cash flow is rental income from creditworthy tenants. Standard EU SPV wrapper with ERC-3643 tokens has emerged as a workable structure for HNW family office access to data center yield. New category in 2026 but the economics are clean.
Specialised industrial equipment (CNC machines, food-grade production lines, energy infrastructure components) leased to operating businesses on multi-year fixed-rent contracts. Same structural pattern as the BESS lease but applied to non-energy equipment. The category works where the equipment is commoditised enough to retain market value, durable enough to survive the lease term, and tied to operators with real credit.
Securitised royalty streams from music catalogues, pharmaceutical patents, or IP licenses, tokenized through an SPV that holds the underlying rights. The cash flow is contractually defined through the underlying license agreements. The structure suits cross-border allocator bases because royalty streams are jurisdiction-agnostic by nature. A growing category in 2026, particularly for music IP, where existing platforms (Royal, ANote, others) have proven the pattern.
Tokenized pools of carbon credits from verified projects (forestry, methane capture, direct air capture). The category requires extreme caution on the verification side (Verra, Gold Standard, or equivalent registry quality) because the underlying credit must be defensible against future regulatory tightening. When the registry quality holds, the tokenized structure provides cleaner secondary mechanics than the traditional opaque OTC carbon market. When the registry quality does not hold, the entire deal is worthless. The asset works in principle; allocator due diligence on the registry is the critical step.
Two patterns across all twelve: each category has a ticket-size floor below which tokenization overhead consumes the margin, and each requires the operator counterparty to be reference-callable with track record on the specific asset type. The wrapper is the easy part; finding asset classes where both conditions clear is the harder part.
These five categories appear on most "what can be tokenized" lists. The economics do not support the claim in 2026. Each one has been tried at meaningful scale and the structural numbers are public.
A €200K single-family rental generates ~€12-15K of gross rent per year. Property management, vacancy reserve, maintenance, accounting, and SPV overhead consume €3-6K. Tokenization overhead (platform subscription, KYC, on-chain registry) adds another €2-4K per property at small portfolio scale. Net to investors lands at 3-4% gross, before any discount for thin secondary liquidity. RealT, Lofty, and GromaCoin all operate in this category and the structural problems are visible in the data: low monthly transfer volume, single-digit holder counts at the property level, persistent gap between marketed yield and net delivered yield. Traditional REITs at 4-5% dividend with deep market depth are a better tool below aggregate portfolio scale of €10M+.
Tokenized fine art has been the headline use case in marketing materials since 2018 and remains structurally unsolved. The underlying asset produces zero cash flow (you cannot underwrite "future appreciation"), the custody and authenticity problems are not solvable through a smart contract, and the wrapper for fractional ownership of a single illiquid art piece does not match any standard EU private placement exemption. Some platforms (Masterworks, Particle) have raised capital and acquired art, but the secondary liquidity is even thinner than tokenized real estate. Treat the category as speculation through a custody layer, not as a real-asset SPV.
Mid-market private equity deals with custom carry structures, multi-class share preferences, vesting milestones, and management option pools cannot be cleanly forced into the ERC-3643 standardised cap-table pattern. The tokenization platforms can handle them but the customisation eats the cost advantage that justifies tokenization in the first place. Traditional LP-GP fund vehicles with paper subscription remain the right tool for bespoke private equity. Tokenization adds value where the deal economics are standardised; bespoke PE is the opposite of standardised.
Distressed real estate, non-performing loans, turnaround equity. These categories require active operator management and discretionary decisions during the hold period. The tokenized SPV wrapper is structurally passive: allocators expect to receive distributions per a defined waterfall without ongoing decisions. Distressed strategies need a fund vehicle with a manager who has discretion. Tokenization can wrap the fund's LP interests, but it adds no operational value over standard LP paper at this stage of the asset's lifecycle.
The category that gets the most marketing attention and the least allocator scrutiny. A tokenized stablecoin pool is not a real-asset deal regardless of how the marketing frames it. A tokenized DeFi yield strategy is not real-asset exposure. A tokenized "digital gold" position is at best a crypto trade with a fancier wrapper. These can be legitimate investments on their own merits but they are not RWA in any meaningful sense and should not be evaluated using real-asset tools. The RWA Roast series covers several of these projects (Ondo, Centrifuge, Chainlink) where the marketed "RWA exposure" diverges meaningfully from what the underlying asset actually produces.
Thirty-minute strategy session for operators. Bring the asset, the rough raise target, the timeline, and the allocator base you're reaching. I'll show you which structure fits inside the 12 categories above, whether tokenization is the right tool at your ticket size, and what the all-in cost looks like at the structure that matches. If your asset is in the 5-that-don't list, I'll tell you that too.
I cannot source every asset class on the list. Real expertise compounds within specific categories. The desk currently focuses on three of the twelve, with five live deals across them.
| Category | Live deal | Ticket | Yield |
|---|---|---|---|
| Battery energy storage | Mobile BESS trailer | €80K | ~15% gross |
| Battery energy storage | Stationary BESS container | €500K | up to 20% gross |
| Working capital cycles | Solar trading SPV | €121K | ~22.6% gross / cycle |
| Commercial real estate | Austrian warehouse | €2.8M | ~7.7% gross (projected) |
| Energy infrastructure | Brașov clean-energy park | €12.5M | ~9.1% gross |
Additional structures in Hungary, Slovakia, and Estonia are sourced when the economics qualify, primarily within battery storage and working capital. New asset categories enter the desk only when there is reference-callable operator depth in the category and the unit economics work cleanly. Sourcing breadth without sourcing depth produces broken deals; this is why the desk stays narrow.
For the right way to evaluate any tokenized deal regardless of which desk sources it, the 9-point DD checklist applies directly. For the structural choice between tokenized SPV and traditional private placement, the comparison guide walks the trade-offs.
Three categories sit on the edge of tokenizable in 2026 and could move firmly into the 12 by 2027 if the underlying market matures and the legal stacks catch up.
Insurance-linked securities at fractional scale. Catastrophe bonds, reinsurance-linked notes, and parametric insurance pools have clean cash flows and standardised structures. The tokenization wrapper has not yet been adopted at meaningful scale, partly because the existing institutional ILS market is functional without it. Could enter the 12 in 2027 if a credible platform builds the legal stack.
Tokenized supply chain finance at scale. Trade finance (category 6 above) covers the invoice side; full supply chain finance (multi-tier vendor financing, inventory-financing across multiple SKUs, just-in-time working capital) has more structural complexity but materially larger total addressable market. Likely to mature in 2027-2028 as the legal patterns from trade finance generalise.
Tokenized infrastructure debt. Existing infrastructure debt funds work well for institutional LPs. The tokenized SPV pattern could open this to family office allocators who currently lack access. The legal complexity (multi-jurisdiction, multi-asset, long-duration) is higher than current tokenized categories but the underlying cash-flow predictability is unusually strong.
Three categories that will probably stay off the list permanently: highly-cyclical retail (margins too volatile), early-stage venture (no contractual cash flow), and sports betting receivables (regulatory positioning fails test 3 in most EU jurisdictions).
Twelve categories make economic sense in 2026 Europe: commercial real estate (€2M+), battery storage, operational solar and wind, EV charging, private credit pools (€5M+), trade finance, working capital cycles, multifamily residential (€10M+), data centers, industrial equipment leasing, royalty streams, and verified carbon credits. See section 02 for the full walk through each.
Five categories where the marketing outruns the math: single-family residential under €5M aggregate, fine art and collectibles, highly bespoke private equity, distressed assets requiring active management, and "crypto-native" assets dressed as RWA. See section 03.
€500K of deployable capital is the practical floor. Below this, the platform setup, KYC infrastructure, and legal SPV overhead consume too much of the per-deal margin. Above €500K the overhead amortises; above €5M the cost advantage over traditional private placement compounds meaningfully. See section 01.
It depends on asset, structure, ticket size. Commercial real estate at €2M+ tokenized through standard EU SPV structures works well. Single-family or small-portfolio tokenized real estate delivers thin yields after fees with near-zero secondary liquidity. Run the 9-point DD checklist on any specific deal before committing.
Working capital cycles (15-25% gross per cycle on short-cycle commodities like solar panels, trade receivables, equipment trading), Battery storage on operator lease (15-20% gross in Romania specifically), and certain trade finance pools (10-18% gross). All require operator counterparty quality that survives reference calls.
Romania has the widest day-night power spread in the EU (15-25%), driven by accelerating solar installation against limited existing storage. The grid operator captures the spread, takes the risk, and pays fixed lease rents that support those yields. In Germany the spread is 5-8% and the same lease structure would yield 5-7%. Geography matters. Read the BESS investment pillar for the full market picture.
Battery storage (two live deals), working capital cycles (one live deal), commercial real estate (one live deal), and energy infrastructure parks (one live deal). Five live deals across three of the twelve categories. See section 04 for the table.
All three sit on the edge of tokenizable in 2026 and may enter the 12 by 2027 as the underlying markets mature and the legal stacks catch up. See section 05.
Apply the 9-point DD checklist regardless of which advisor sources the deal: asset, operator, jurisdiction, SPV structure, token mechanics, allocator rights, distribution waterfall, reporting, exit. The checklist works the same for any tokenized real-asset deal in any of the 12 categories. Read the DD checklist.