November 20, 2026

Cross-Chain TVL and the Asset That Belongs to No Chain

Sagar Prasad
Portfolio Manager
Copper hexagon-grid header with line-art icons of two chain nodes joined by a bridge arc with an asset between them
In This Article
Share
Questions? Speak to our Team

On November 16, 2026, DefiLlama's protocol listings summed to $449.1 billion while the headline DeFi total value locked figure read $71.4 billion, roughly one-sixth of the parts. The difference is a deduplication policy, and its most revealing line concerns bridges: $67.9 billion of bridge balances are credited to no chain at all, because the origin and destination chains both claim the same locked coins. That is a defensible answer for a dashboard and an impossible one for a balance sheet, which has no column for an asset that is nowhere.

What the Headline Number Is, and Where It Puts a Bridged Coin

The balance-and-price arithmetic is the easy layer; the rules deciding which sums enter the published total are not. Working from the same listing data, it is possible to rebuild the headline at $71.1 billion, within $0.3 billion of the published $71.4 billion. The arithmetic reproduces; the exclusion decisions are what a reader has to go find.

Each exclusion is individually reasonable. Of the listings, 1,779 carry a liquid-staking or double-counted flag and hold $71.0 billion between them, liquid staking alone accounting for $41.7 billion.

Then there is the bridge decision, which is the one worth sitting with. Two chains hold a legitimate claim on the same coins, the aggregator cannot adjudicate, so it assigns the balance to neither. In software terms: when two parents claim one node and the schema offers no way to decide, the node is dropped from the tree rather than attached twice. The total stays honest. The location becomes undefined.

A Bridge Hop Is a Reclassification, Not a Transfer

Reporting cannot leave the location undefined. Under ASC 350-60, created by ASU 2023-08 and effective for fiscal years beginning after December 15, 2024, qualifying crypto assets are measured at fair value through net income. Wrapped tokens are the textbook scope problem, satisfying some criteria in ASC 350-60-15-1 while potentially failing others, which returns them to the older intangible model with one-way impairment.

Native ether and its bridged representation are therefore the same economic exposure on two measurement models. I have argued before that this distinction determines a position's risk profile, because a wrapped token is a claim on a bridge rather than on the issuer; it now also determines which accounting model applies. Note what the scope test never references: the chain identifier. Movement alone settles nothing either, since derecognition turns on rights transferred, rights retained and control, assessed through the ASC 606 model.

The operational failure follows quietly. A system that treats a hop as a disposal and a fresh acquisition books a realized gain or loss at the hop price and resets basis, though no economic event occurred: the position never left the manager's control and no counterparty changed. The symmetric error is netting: positions that cannot be shown to be the same asset against the same obligor cannot be offset, however fungible they look on a screen.

Four Numbers That Replace It

First, the in-scope share: what percentage of the book, by value, sits on the fair-value model rather than the intangible one. That ratio says how much of the portfolio marks through net income each period and how much only ratchets down.

Second, the encumbered share: units pledged, staked or lent, tracked apart from units freely available. Third, a reconstruction tolerance. Rebuild the firm-level figure from per-contract, per-chain, per-token leaves and compare it to the stored number; roughly 0.4 percent, where the $71.1 billion rebuild fell against $71.4 billion, is a fair bar.

Fourth, the count of unmapped hops, meaning positions whose cross-chain history has not been collapsed into a single economic movement. Administrators running multi-chain books treat burn-and-mint events and route hops as one movement precisely so transitions do not manufacture phantom gains.

What a Reportable Trend Looks Like

A healthy quarter shows the in-scope share stable or rising, unmapped hops at zero at every period end, encumbrance flagged at trade date rather than reconstructed at close, and a reconstruction delta inside tolerance without manual adjustment. Those four conditions mean the entity can answer, for every position, where the asset is and which model measures it.

The unhealthy pattern looks better on the surface. Headline exposure grows, the chain dashboard looks diversified, and the share of positions tied to a single enforceable claim quietly falls as capital moves into wrapped representations across more venues. The thing to watch for is a vendor's deduplication policy becoming the firm's measurement basis by default, because operations cites the published number instead of computing its own.

The consolation is that concentration makes the leaves tractable: the top fifteen protocols held $102.8 billion of the November 16 total, so a book touching the major venues across a dozen chains is a few hundred contract-level positions. Mapping that set to legal claims is a finite build, not a research problem.

What Is Getting Fixed

Standard setters are converging on the leaf-plus-attribute design. As of September 2026 the FASB had tentatively decided to expand ASC 350-60 to cover crypto assets commonly called wrapped or receipt tokens, where the holder has a right to receive another in-scope crypto asset, and had directed staff to draft a proposed update for a 60-day comment period.

The lending decisions are the better signal. At its August 2026 meeting the Board tentatively concluded that crypto asset lending should not cause derecognition: the lender keeps the asset and reclassifies it as encumbered, measured at fair value adjusted for borrower credit risk. That is one asset, one record, with a status attribute, rather than two assets in two places to be netted afterward.

None of this is yet a planning assumption. Writing on November 6, 2026, Forvis Mazars described the wrapped and receipt token question as still under evaluation with no tentative decision reported, while KPMG's September account described a scope decision taken and an update in drafting. Two firms reading one project characterize its status differently, which is itself the status. Lending remains undefined, liquidity pool tokens remain open, and until a final standard lands the classification work belongs to the preparer.

Cross-chain TVL is a usage proxy with a defensible answer for a disputed location: none. Institutional reporting has no such answer available, so the aggregate cannot be inherited. It has to be computed from leaves the reporting entity owns.

Sources:

1. Coinpaprika, "What Is Total Value Locked (TVL)?", DefiLlama data as of November 16, 2026, accessed November 20, 2026. 2. Forvis Mazars, "Digital Asset Accounting: How Far We've Come & What's Ahead," November 6, 2026. 3. KPMG, "FASB makes progress on crypto asset transfers project," September 2, 2026.

For informational purposes only. Not an offer to buy or sell any security. Available only to accredited investors who meet regulatory requirements.

‍

Recommended blog posts