Irem Erten, Ioana Neam?u, John Thanassoulis
Review of Finance, Volume 30, Issue 3, May 2026, Pages 995–1028, https://doi.org/10.1093/rof/rfaf076
We study the impact of ring-fencing on bank riskiness using short-term money markets. Ring-fencing is when the government restricts some banking activities to a subsidiary of the group whilst restricting intra-group transfers. Exploiting confidential data on sterling denominated repo transactions, we document that banking groups subject to ring-fencing are perceived to be safer – repo investors lend to ring-fenced groups at lower rates – and that the safety perception is amplified during times of market stress. Our paper suggests that structural reforms can create a ‘safe-haven’ bank in the financial system.
The UK is one of the few countries that implemented ring-fencing in the aftermath of the Global Financial Crisis of 2008, though the EU considered it. We use the interest rates in the UK government bond repo market to capture market participants’ perceptions of counterparty risk. In the UK most repo transactions are negotiated bilaterally. This setting isolates counterparty risk, since most transactions are overnight and backed by safe collateral.
A key feature of the UK repo market is that counterparties trade with multiple dealers within short time windows. In effect this means we can identify how counterparties reward dealers who are ring fenced repeatedly, and predictably, on the same day in near-identical conditions. So we can identify the risk premium counterparties demand to lend to banks, and how this premium shifts with ring-fencing.
We find that the ring-fenced dealer groups can borrow in the overnight repo market at rates which are 3.5% lower than the median repo rate. This statistically significant ring-fencing bonus indicates that repo investors perceive ring-fenced groups as safer following the reform. We show that this effect is driven by the ring-fenced subsidiary (RFB) which can borrow at about 10% lower rates than the median repo rate. Meanwhile the non ring-fenced subsidiary (nRFB) is not materially affected. This effect can be seen in the figure which shows, with no controls, the spread between repo rates secured by the subsidiaries of the ring-fenced banking groups as compared to the groups not subject to ring-fencing.
Further analysis confirms that ring-fencing laws are required for the bonus to manifest. Banks cannot recreate the bonus unilaterally by creating a domestically focussed retail banking subsidiary within the group. One ring fencing rule is that the Liquidity Coverage Ratio needs to be satisfied at the RFB subsidiary level, and not purely at the highest legal entity within the jurisdiction. This means that the RFB benefits from liquidity insurance from the wider group – but credibly the RFB subsidiary does not contribute to the insurance for other parts of the group. Hence the RFB subsidiary will have good liquidity, even in stress times. And this is rewarded with lower borrowing rates in the repo market.
The paper confirms that the effects are even stronger in stress times, such as the Covid pandemic. Hence we conclude that ring fencing can generate a safe haven bank which counterparties prefer to trade with, especially in times of stress.