Bridging Two Markets: Dealer Intermediation across Securities Lending and Repo

Repo and securities-lending markets trade the same sovereign bonds through separate participant universes, bridged by dealers who borrow bonds from beneficial owners and redeploy them as repo collateral. I model the lending fee as the dealer's repo deployment value: the bond's repo specialness, scaled by its redeployment intensity and split across a lending and a repo bargain. Lending fees therefore pass through repo specialness, most strongly for repo-central bonds when lending utilization is high. In matched securities-lending and repo data for euro-area sovereigns (2018-2024), the pass-through concentrates there and is muted elsewhere; bonds physically cross into repo, and fees, specialness, and repo volume comove. Auction reopenings confirm specialness is a supply-driven repo price. A lending fee is, in effect, a repo price.

Selected Presentations/Conferences: 3rd Frankfurt Summer School Deutsche Bundesbank – Leibniz Institute for Financial Research SAFE – Halle Institute for Economic Research (IWH)

Work in progress

Repo Collateral Reuse and Liquidity Windfalls ECB Working Papers series No 3147 ,  featured as a SUERF Policy note No. 403

with Sofia Marques Pereira and Victor Rodrigues-Gomes

Reuse of collateral within repo markets is key in enabling participants to meet their short-term financing needs, maintaining market efficiency, and establishing collateral valuations. In addition, part of the literature relies on the premise that some market players, in special large dealers, take advantage of their market position to obtain “liquidity windfalls” through haircut differences when reusing collateral. However, despite the importance of this mechanism for market functioning, the result is mainly theoretical, as empirical work exploring the effects of collateral reuse is scant. Through the analysis of a novel database on European Securities Financed Transactions, this study aims to help fill this gap. We show that around 11 percent of transaction volume is based on reused securities, with chains averaging three links. Besides, contrary to the liquidity-windfalls hypothesis, we find that dealers do not impose systematic haircut wedges when interposing between non-dealers.

Selected Presentations/Conferences: ECB Conference on Money Markets 2025 (Poster session, available here), Eastern Finance Association (EFA) Meeting 2026, Southwestern Finance Association (SWFA) Meeting 2026, Summer Workshop on Money, Banking, Payments and Finance 2026 (Study Centre Gerzensee, Bank of Canada, Board of Governors), Northern Finance Association (NFA) Meeting 2026, ESMA Research Conference 2026

Internal Securities Markets

with Felix Hermes, Benoit Nguyen and Davide Tomio

This paper provides evidence that global banking groups use their internal market to facilitate the access of non-banks to foreign centrally cleared repo markets. Using exhaustive repo transaction-level data, we provide evidence that intragroup repos, a market up to 20% as large as the entire repo market, operate as the middle leg of an intermediation chain in which non-bank investors transact with an affiliate of a banking group, which in turn sources or places the corresponding collateral and cash in the centrally cleared market through a second affiliate in another jurisdiction. The geography of intragroup activity retraces this cross-border circulation of collateral and currencies. Intragroup repo is predominantly cross-border, and the currency of activity is separated from the nationality of the group conducting it. Foreign banking groups dominate intragroup activity in euros, while euro area groups dominate it in dollars. This hidden indirect non-bank demand is associated with pricing in the cleared repo market segment, where bonds linked to the intragroup chain trade more special, implying that the footprint of non-banks on euro area repo pricing is larger than direct activity alone would suggest.

Private collateral chains and public collateral backstops

with Benoit Nguyen and Victor Rodrigues-Gomes and Iñaki Aldasoro

We study how central bank securities lending facilities interact with private collateral chains in European repo markets. Using transaction-level SFTDS data, we trace reused securities across successive repos and reconstruct nearly two million chains. Conditional on reuse, chains average 2.24 nodes beyond the original owner, implying a collateral multiplier of 1.37. Merging these chains with Eurosystem securities lending data, we find that facility usage does not lengthen chains or increase interconnectedness. Reuse is broadly unchanged, though effects differ across jurisdictions, and facility activity is positively associated with specialness. Public collateral therefore complements private reuse by easing scarcity without crowding out market intermediation.

Netting Effects in Bilateral Repo Markets

with Victor Rodrigues-Gomes and Michael Schmidt

Repo markets are central to euro-area monetary-policy implementation and to the ongoing debate on non-bank leverage, yet little is known about how netting shapes their functioning. Using transaction-level regulatory data, we construct portfolio- and transaction-level netting ratios for both close-out and balance sheet netting in the European bilateral repo market. We find that netting is strongly correlated with repo terms and with traded volumes; most strikingly, netted repos are far more likely to carry a zero haircut. These correlations, however, are driven by which counterparties net: once we compare transactions within thesame counterparty pair, they essentially disappear. Bilateral haircuts are thus set by the counterparty relationship rather than by collateral or netting, a finding that bears directly on proposals for minimum haircut floors. We further document that netting itself is procyclical: the share of bilateral exposure that can be netted falls at reporting dates, just when balance-sheet capacity is scarcest, as counterparties lose their netting partners. This drying-up of netting when it would be most valuable is directly relevant to the case for central clearing in euro repo markets.

Stock Market Participation in Germany, Wealth Effects, and Monetary Policy Transmission, R&R at the Journal of International Money and Finance

with Asli Mahmudova

We study whether equity valuation shocks around ECB announcements transmit into household spending in Germany. Combining household balance-sheet data from the Panel on Household Finances (PHF) with monthly local projections, we document three findings. First, equity ownership is rare, concentrated, and systematically patterned: the top wealth tercile holds over 90\% of directly held equity and nearly 88\% of broad equity exposure, with participation shaped by education, wealth, income, and trust. Second, announcement-driven equity shocks are followed by a delayed rise in retail activity that peaks at 10–11 months, with no detectable response in inflation at any horizon. Third, a calibration bridge that maps PHF equity exposures and heterogeneous marginal propensities to consume into a predicted aggregate consumption response is directionally consistent with the macro estimate and confirms that the top wealth tercile accounts for the large majority of the predicted spending effect. Together, the results characterise what we call a thin-tube wealth channel: equity valuation movements around ECB announcements do reach household spending, but transmission flows through a narrow, already wealthy segment of the population, generating modest demand effects and no measurable price pressure. The width of the tube is itself a policy-relevant variable — endogenous to participation rates, financial literacy, and product design — rather than a fixed structural feature.

Where Does the On-the-Run Premium Go? Futures Intermediation and the Cheapest-to-Deliver Bond in European Sovereign Bond Markets

The on-the-run U.S. Treasury is rich: it commands a price premium, special repo, and a high lending fee. Does this convenience yield survive where benchmark demand is intermediated through futures rather than the cash bond? Using a daily panel matching the cash, repo, and lending markets for six European sovereigns (2018–2024), I show it does not. Within bond, the cash premium is economically negligible and repo specialness reverses; the lending-fee gap chiefly reflects a freshly issued bond’s thin float rather than its on-the-run status. Scarcity is not eliminated but relocated to the cheapest-to-deliver bond, the focal point of futures-intermediated demand. The on-the-run premium appears to be a feature of cash-benchmark trading that futures intermediation suppresses.

AI-driven peer group selection using ML and numerical data-based analysis and its implication for Corporate Finance

with Mark Wahrenburg and Marin Shalari