Monitoring collateral calls

The purpose of Bank Overground is to share our internal analysis. Each bite-sized post summarises a piece of analysis that supported a policy or operational decision.
Published on 06 March 2025
Yields on gilts and government bonds across other advanced economies moved higher in the first half of January. The moves were predominantly driven by global factors, with a focus on fiscal policy in advanced economies.footnote [1] Changes in government bond yields can trigger calls for additional collateral in both repo and derivative markets. In historic periods of market stress, such collateral calls have amplified price movements by forcing investors to sell assets. During the January period of yield increases, however, collateral calls were much lower than in historic stresses such as the March 2020 dash for cash (DFC) and the Autumn 2022 liability-driven investment (LDI) crisis and remained manageable. This post explains our approach to monitoring collateral calls.

Market developments in the first half of January

Chart 1 plots the rise in gilt yields in the first half of January. It shows that increases were more pronounced at medium and longer maturities and were sharper during 7 and 8 January (Phase 1) before slowing in the next four trading days (Phase 2). For instance, the 10-year gilt yield increased by around 20 basis points in Phase 1 and a further 10 basis points in Phase 2. The moves were predominantly driven by global factors, with a focus on fiscal policy in advanced economies. Thus, government bond yields also rose elsewhere in Europe and in the United States (Chart 2). Although the moves in gilt yields attracted some commentary, they were considerably smaller than in recent episodes of gilt market stress as well as the hypothetical stress used in the Bank’s 2024 System-wide exploratory scenario (SWES) exercise (Chart 3). During the DFC and the LDI stresses, spikes in yields led to a spiral of collateral calls and forced asset sales, which further amplified yield moves and saw central banks intervene to restore market functioning. As we explain below, such dynamics were not observed in January as collateral calls were much smaller.

Chart 1: Gilt yields during the first four weeks of January 2025

Line chart showing 1 year, 5 year, 10 year and 30 year yields of UK government bonds (ie gilts). Longer-term gilt yields rose between 6 and 14 January and reversed afterwards. Yields rose more sharply between 6 and 8 January (Phase 1) than in the period between 8 and 14 January (Phase 2).

Footnotes

  • Source: Datastream from LSEG.

Chart 2: Cumulative changes in international 10-year government bond yields during the first four weeks of January 2025

Line chart showing cumulative changes during the first four weeks of January in 10-year government bond yields of the United Kingdom, France, Germany and the United States. The chart shows strong comovement, with all of the 10-year yields rising between 6–14 January, after which much of those moves are reversed.

Footnotes

  • Sources: Datastream from LSEG and Bank calculations.

Chart 3: Trough-to-peak changes in gilt yields (a)

Bar chart comparing the trough-to-peak changes in 5 year, 10 year and 30 year gilt yields for selected periods of gilt market stress and the SWES scenario. Increases in yields were by far lowest during the January period and highest during the LDI episode.

Footnotes

  • Notes: DFC = March 2020 DFC, LDI = Autumn 2022 LDI crisis and SWES = SWES.
  • Sources: Bank of England, Datastream from LSEG and Bank calculations.
  • (a) Changes in yields between close of trading on 6–14 January (January 2025), 9–18 March 2020 (DFC) and 21–27 September 2022 (LDI). The hypothetical SWES stress is assumed to occur over a two-week period.

Collateral calls on gilt repo

As gilt yields rose (and prices declined), cash borrowers in the repo market would have been required to post more collateral against their loans. This may have come about as a result of a collateral shortfall on outstanding repos that borrowers were asked to redress or through the rollover of maturing repos into new contracts that required more bonds as collateral given their lower valuations. In either case, we assume that the collateral calls faced by non-bank financial institutions (NBFIs) on their repos between 6–14 January would have been equal to their outstanding positions at the end of 6 January multiplied by the percentage change in the value of the collateral between then and close of trading on 14 January.

Data on repo positions are collected under the UK Securities Financing Transactions Regulation (SFTR data).footnote [2] Specifically, for different types of NBFI, we analysed data on their outstanding repo (ie cash borrowing) and reverse repo (ie cash lending) positions grouped into buckets reflecting the maturity of the underlying collateral (0–2 years, 2–5 years, etc.). We then estimated price changes for the collateral in each bucket by multiplying the change in yield of the gilt closest to the mid-point of the bucket by its modified duration. The product of these positions and price changes gives our estimates of collateral calls on repos and collateral receipts on reverse repos.footnote [3]

Chart 4 shows the results. Pension and LDI funds had the largest estimated collateral calls on a net basis of around £3 billion. This reflects the leveraged positions in gilts that they take through repo borrowing to help match the duration of their liabilities. Hedge funds were also estimated to have faced net collateral calls on their gilt repo positions. These positions may have been part of a wider investment strategy, however, making their overall collateral needs harder to judge. Part of them may have related to interest rate swap positions, to which we now turn.

Chart 4: Estimated collateral calls on gilt repo by sector between 6–14 January

Stacked bar chart showing incoming and outgoing collateral for gilt repo transactions for the hedge fund, insurer, investment fund and pension/LDI fund sectors. Net collateral outflows were highest for Pension/LDI funds (around £3 billion), followed by hedge funds and other sectors.

Footnotes

  • Sources: Datastream from LSEG, SFTR data and Bank calculations.

Collateral calls on interest rate swaps

The increase in interest rates in early January also led to a repricing of interest rate swaps. Counterparties who had made commitments to pay fixed rates of interest gained, while those committed to pay the ‘floating’ market rate of interest lost. To manage the counterparty risk associated with such developments, losing counterparties transfer collateral to gaining counterparties in amount equal to the change in the market value of the contract. These transfers are known as variation margin (VM).

We estimate VM calls on the largest categories of interest rate swaps (ie overnight indexed swaps, other fixed-for-floating swaps and inflation swaps) denominated in major currencies (GBP, USD, EUR, JPY, AUD and CAD).footnote [4] We do so in a similar manner as for collateral calls on repos, by combining data on positions and price changes. Our derivative positions data is collected under UK European Market Infrastructure Regulation (UK EMIR TR data), while Chart 5 shows selected changes in swap rates that occurred both in Phase 1 and Phase 2. For instance, and not surprisingly, the 18 basis point and 6 basis point rises in the 10-year GBP swap rate during these two phases were similar to the increases in the 10-year gilt yield. By comparison, trough-to-peak changes in the 10-year GBP swap rate during the DFC and LDI stress periods were 47 basis points and 117 basis points respectively.

Chart 5: Shifts in overnight index swap zero curves between 6–14 January 2025

Line chart showing shifts in GBP, USD and EUR overnight index swap zero curves for tenors between 1 year and 50 years for the two periods 6–8 January (Phase 1) and 8–14 January (Phase 2). During Phase 1 upward shifts were highest in long-term UK OIS rates. In Phase 2 upward shifts were higher for USD and EUR rates compared to UK rates.

Footnotes

  • Note: Any missing maturities were interpolated using cubic splines.
  • Sources: Datastream from LSEG and Bank calculations.

Chart 6 shows our estimates of the VM calls triggered by the January swap rate changes. As for gilt repo, pension and LDI funds faced the largest collateral calls on a net basis, having to pay £1.5 billion. This also reflects their use of swaps for asset-liability matching. Insurance companies, which also use swaps for this purpose, faced net VM calls of £0.6 billion. Hedge funds and investment funds also faced net VM calls, although these were somewhat smaller.

Chart 6: Estimated VM calls on interest rate swaps by sector between 6–14 January 2025

Stacked bar chart showing incoming and outgoing variation margin calls for interest rate swaps for the hedge fund, insurer, investment fund and pension/LDI fund sectors. Net variation margin outflows were highest for Pension/LDI funds (around £1.5 billion), followed by insurers and other sectors.

Footnotes

  • Sources: Datastream from LSEG, EMIR TR data and Bank calculations.

Comparison with historical and hypothetical market stresses

Adding up across the different sectors, we estimate that NBFIs active in UK markets faced net collateral calls of around £4 billion on their gilt repo positions and net VM calls of around £3 billion on their interest rate swaps. As shown in Table A, these amounts are small compared with collateral needs in the March 2020 DFC and the Autumn 2022 LDI crisis, and also in comparison with the hypothetical collateral needs in the SWES. Consistent with that, the Bank did not detect any signs of forced sales as a result of collateral calls. It will, however, maintain – and further develop – its toolkit for assessing collateral calls.

Table A: Collateral calls of NBFIs active in UK markets

Period/scenario

Repo

Swaps (a)

Repo and Swaps

January 2025

£4 billion

£3 billion

£7 billion

DFC (a)

N/A

£12 billion

N/A

LDI (b)

£53 billion

£13 billion

N/A

SWES

N/A

N/A

£86 billion (c)

Footnotes

  • Sources: See links in table.
  • (a) Interest rate swaps, except for the DFC figure, which also includes FX swaps.
  • (b) Pension and LDI funds only.
  • (c) NBFIs need to meet approximately £94 billion of margin calls in aggregate, with £86 billion of those liquidity needs arising from VM calls.

This post was prepared with the help of Daniel Krause and Nicholas Vause.

This analysis was shared with Governors in January 2025.

Share your thoughts with us at BankOverground@bankofengland.co.uk

  1. See pages 30–31 of the February 2025 Monetary Policy Report.

  2. Our estimates of collateral calls rely on the completeness and accuracy of positions reported in both the SFTR and EMIR TR (see below) data sets. These data sets are subject to ongoing checks to continuously improve the quality of the data.

  3. To the extent that repos are over-collateralised, providing some headroom before collateral calls are made, these estimates may be too high.

  4. While these interest rate swaps account for the majority of non-banks’ outstanding interest rate derivatives in terms of notional values, it remains possible that they could face (additional) VM calls on other interest rate derivatives (eg interest rate futures). In addition, our estimated liquidity needs do not take into account any depreciation of collateral posted as margin.