The Anatomy of Financial Collateral: Dissecting the Myth of Collateral Chains in Repo and Derivatives Markets

A prominent post‑crisis narrative holds that the reuse of collateral in derivatives, repo, and securities‑lending markets generates ‘collateral chains’ that bind banks and shadow banks into opaque webs of interconnections and contagion channels. According to this narrative, when the collateral is reused under a financial collateral arrangement such as a repo transaction, the same asset is thought to support multiple secured transactions, so that a single failure to return collateral can trigger a chain of settlement fails and propagate shocks throughout the system. This narrative has become orthodox in post‑crisis debates and is reflected, for example, in the EU’s Securities Financing Transactions Regulation (SFTR), which links collateral reuse directly to complex collateral chains and systemic risk.

In my recent working paper, ‘The Anatomy of Financial Collateral’, I argue that the collateral‑chain story is largely a misleading myth. Once we look closely at how the law structures financial collateral, in particular the distinction between property and contract rights, and the operation of early‑termination and close‑out netting provisions, it becomes difficult to sustain the idea that extended chains of proprietary interests in the same asset can arise in the way the narrative suggests.

The starting point is the basic distinction between security financial collateral arrangements (SFCAs) and title‑transfer collateral arrangements (TTCAs). In an SFCA, the collateral provider grants a security interest over the collateral; in a TTCA, legal title to the collateral is transferred outright to the collateral taker. Collateral chains are usually perceived in the SFCA setting, particularly in repurchase agreements (repos) where reuse is allowed: the same securities are assumed to be pledged and re‑pledged again and again along a chain of transactions.

However, the law does something interesting when it allows reuse of financial collateral. Under the EU Financial Collateral Directive, once the collateral taker exercises a right of use over collateral posted under an SFCA, the collateral provider’s proprietary rights in those assets are automatically converted into a personal (contractual) claim against collateral taker for the return of equivalent collateral. This transformation or substitution matters because property rights ‘run with’ the asset, whereas contract rights bind only the immediate counterparty. After reuse, the original collateral provider no longer has a property right in the reused securities and cannot assert claims against downstream transferees. Legally, the chain of proprietary interests is cut at the first reuse. What remains is a chain of contracts, not a chain of property rights supported by the same asset.

The rationale for transforming proprietary rights into contractual rights appears to be threefold. 

First, absent such a transformation upon reuse, no prospective collateral taker would be willing to accept the collateral for fear that it might be subject to proprietary interests held by the original collateral provider or by subsequent collateral takers who have reused it along the collateral chain. High information costs would discourage collateral takers from accepting it. 

Second, in the absence of such a transformation upon reuse, the provider’s proprietary claim would follow the collateral each time it is reused. In the event of a default somewhere down the collateral chain, the original provider would have to locate the specific collateral, identify its current holder, and bring a claim against that party. Given that perfection of collateral occurs only through possession or control, such a search would be virtually impossible. Keeping proprietary rights of collateral providers intact would not only inflate search and litigation costs but also severely weaken incentives to post collateral that can be reused.

Third, the absence of such a transformation upon reuse would expose collateral to a ‘tragedy of the anti‑commons’ and lead to entropy and underuse. The reason is that the proprietary rights of each prior user in the chain would follow the asset. A multiple monopoly hold‑out problem would then arise because every collateral provider in the chain retains its own proprietary interest, a downstream collateral taker wishing to reuse the collateral would need the consent of each upstream provider. Each of these providers would be in a monopoly position to withhold consent and to extract the maximum possible rent from prospective rehypothecators. Economic theory predicts that, under such conditions, collateral will be underused. 

In addition to the above rationale, financial collateral enjoys robust early‑termination and close‑out netting protections. The obligation to return equivalent collateral is part of this netting exercise, so the collateral provider’s unsecured exposure following reuse is typically limited to the haircut, not the full market value of the reused securities. Netting thus mitigates precisely the contagion channel that the collateral‑chain narrative fears.

The paper further demonstrates that the structure of most financial collateral arrangements is fundamentally synallagmatic, with mutual and fluctuating obligations on both sides. Collateralization mirrors this structure: it is bilateral and symmetrical. Because exposures are dynamically re‑balanced through margining, and because each party holds collateral (cash or securities) against the other’s performance, it is misleading to characterize a chain of such bilateral relationships as being ‘supported’ by a single underlying asset. What appears ex-ante as one piece of collateral backing multiple trades is, in legal and economic reality, a series of offsetting contractual claims, each collateralized at the level of the immediate counterparty.

The central implication is that much of the finance and policy literature overstates the extent to which collateral reuse generates collateral chains and systemic interconnectedness. To the extent that reuse contributes to leverage, liquidity creation, or maturity transformation, these effects arise through well‑identified contractual exposures between immediate counterparties, not through an invisible web of property claims following a single security across the system. Systemic‑risk analysis and regulation should therefore focus on the concentration of bilateral exposures, market‑wide margin and haircut dynamics, and the robustness of netting and close‑out frameworks, rather than on a largely illusory metaphor of collateral chains.

The author’s paper can be read here.

Hossein Nabilou is an Assistant Professor of Law and Finance, University of Amsterdam, Amsterdam Law School.