No part of this publication may be reproduced or transmitted in any form or by any means without permission from ICMA. This report is intended for general information only and is not intended to be nor should it be relied upon as being legal, financial, investment tax, regulatory, business or other professional advice. Users of this report should seek appropriate independent advice before entering into any kind of specific transaction. While the information contained in this report is taken from sources believed to be reliable, neither ICMA nor the author represents or warrants that it is accurate, suitable or complete and neither ICMA nor the author shall have any liability arising from or relating to the use of this report and its contents.
Executive Summary
In December 2025, the European Repo and Collateral Council (ERCC) of the International Capital Market Association (ICMA) conducted the 50th in its series of semi-annual surveys of the repo market in Europe.
The survey asked a sample of financial institutions in Europe, among other things, for the value and breakdown of their repo contracts that had been transacted by close of business on December 10, 2025, and that would still be outstanding after that date. Replies were received from 59 entities, mainly banks.
Data were also reported separately by the principal inter-dealer automatic repo trading systems (ATS), automated dealer-to-customer repo trading platforms, central counterparties (CCP) and tri-party repo agents in Europe, giving the size and composition of almost all electronic repo trading, central-clearing and tri-party repo collateral management in the region.
Total repo business
The total value of repos and reverse repos outstanding on the books of the survey sample on December 10, 2025, hit a record high of EUR 13,651.1 billion. This was EUR 1.2 trillion and +9.8% higher than the EUR 12,435.9 billion reached in June, and +24.6% higher than in December 2024. In other words, the sharp pick-up in growth seen in the first-half of the year was largely sustained in the second-half.
The drivers of growth in the second-half of 2025 would seem to have been the same as in the first-half of the year, namely, the market volatility and macro-economic uncertainty triggered by US trade tariff announcements, which fed demand for precautionary liquidity and prompted investors to seek shelter in the money market. However, despite volatility in other markets, repo market capacity proved adequate and repo rates generally remained stable, including at end-year.
The survey sample remained a net lender of cash to (and therefore a net borrower of collateral from) the rest of the repo market, as it has been since 2012, but this position has been unwinding since the end of QE in 2022, albeit subject to seasonal fluctuations.
Trading analysis
The trends in place since December 2023 largely continued until the end of 2025. Thus, the share of the outstanding interdealer repo that is traded on automatic trading systems (ATS) contracted further and reached an eight-year low of 25.1%. This reflected further growth in US dollar and US Treasury repo trading by the survey sample, most of which takes place off European venues.
GC financing again increased its share of ATS-traded positions, reaching 12.2% of end-period balances reported directly by the principal ATS. Its growth was fueled by new users and increased turnover but also the longer-than-average tenors of this type of business.
The first ATS-traded positions in securities issued by the EU were reported.
The relative positions of individual ATS continued to reflect their particular focus on different government securities but also increased competition, including recent new entrants.
Voice-broking attained an 11-year high of 13.7% of the survey, reflecting the usefulness of voice-brokers as an immediate and flexible means of extending a dealer’s market coverage when volumes surge; their importance in the dealer-to-dealer (D2D) market segment in currencies and collateral that are not traded electronically in Europe (including the US dollar and Treasuries); and perhaps also their role in GC financing.
There was no change in the survey share of OTC business.
Tri-party repo managed only a weak recovery to 10.5% of the survey sample, despite the attractiveness of repo to investors and the growth in the GC financing component of tri-party.
Automated trading systems serving the dealer-to-customer (D2C) segment of the repo market, and in particular, hedge funds, continued to grow rapidly, although the rate of growth decelerated.
Geographical analysis
The share of cross-border positions into and out of the eurozone extended its upward trend to reach a new all-time high and it continued to be, by far, the largest geographical component of the survey. This segment includes the trading of non-European currencies and collateral.
The upward trend in the share of cross-border positions into and out of the eurozone is reflected in the downward trend, since December 2023, in the share of anonymous business. The negative impact reflects the fact that anonymous business is CCP-cleared ATS trading, which is largely confined to European currencies and collateral.
In the position data reported directly to ICMA by the principal European ATS, the dramatic jump in the share of outstanding balances between non-eurozone counterparties was consolidated at a record level of 11.4%, assisted by the growth in GC financing.
The share of business with APAC counterparties recovered, although it remained well below its December 2022 peak. The recovery may have been helped by increased activity in yen repo mitigating the continued contraction in the share of JGBs, other APAC collateral and other APAC currencies.
Clearing analysis
The outstanding value of anonymous (CCP-cleared) repo trading by the survey sample, excluding GC financing, retreated in December 2025, close to its June level, reflecting the contraction in ATS CCP-cleared trading. However, GC financing bucked the negative trend and expanded by +46.8%, to increase its share of the survey sample to 2.0% from 1.5%, helped by growth in trading between non-eurozone counterparties. The growth of GC financing may also have been reflected in SFTR data in an apparent lengthening of the average term to maturity in CCP-cleared transactions in the EU.
The implied share of post-trade CCP-clearing by the survey sample, which has been declining since 2021, recovered slightly.
Cash currency analysis
The trends in the share of the euro and sterling in the survey were essentially flat, while the growth in the share of the US dollar seems to have resumed, largely at the expense of the yen.
Collateral analysis
US Treasuries remained the largest collateral holding of the survey sample and increased their share of the repo books of the survey sample to a new record of 17.8%. The counterparts to the rising share of US Treasuries were reduced shares for EU government securities and JGBs. However, Italian government issues remained the next largest holding, followed by UK gilts.
In tri-party repo, there was a sharp drop in the share of commercial mortgage-backed securities (CMBS), reflecting mounting concern over high commercial real estate valuations. The principal beneficiaries were government, covered and corporate bonds.
The share of AAA and A-rated collateral in tri-party repo grew strongly, largely at the expense of AA-rated collateral. BBB-rated collateral extended its downward trend.
There was a continued modest but widespread relaxation of haircuts on tri-party credit collateral, in particular, for covered and convertible bonds.
Contract analysis
Repos that have been guaranteed or indemnified (including the various forms of “sponsored” repo) continued to grow, reaching 5.8% of the outstanding positions of the survey sample. However, there was a shift in composition out of euro into dollars.
Repo rate analysis
The share of floating-rate repo showed robust growth over the second-half of 2025, to reach 19.5%, just short of the peak of 19.7% seen in December 2023. The latest change confirmed growing expectations of an end to the easing of interest rates by European central banks in the face of rising long-term inflation concerns.
Maturity analysis
The post-QE dominance of repos with one day remaining to maturity continued. However, there was a shift over the second-half of 2025 out of short-dated positions with between two days and one month remaining to maturity into the one to three-month bracket. One-month to three-month repo positions held by the survey sample remained highly seasonal, driven by collateral swaps over end-year.
There was a further shift in the average term-to-maturity of ATS-traded repo positions, away from a remainingterm-to-maturity of one day, into positions with one to six months remaining to maturity. However, some of this reflected wider coverage of ATS in the survey.
Maturity transformation provided by the survey sample to the rest of the market increased further with a jump in net cash borrowing at a residual maturity of one day and growth in net lending in all fixed terms between one week and six months.
Chapter 1: The Survey
On December 10, 2025, the European Repo and Collateral Council (ERCC) of the International Capital Market Association (ICMA) conducted the 50th in its series of semi-annual surveys of the repo market in Europe. The first of these surveys took place in June 2001 and the series now charts an unrivalled history of the development of the core segment of the European repo market over two and a half decades. During that time, the market has burgeoned and matured into an efficient and essential component of the regional financial system, while coping with several episodes of severe economic and financial turbulence.
The survey was carried out and the results analysed on behalf of ICMA by the author, Richard Comotto, under the guidance of the ERCC Committee.
1.1 What the survey asked
The survey asked financial institutions operating in Europe, who are members of ICMA, for the starting value of the cash side of repos and reverse repos that were still outstanding at close of business on Wednesday, December 10, 2025. The survey therefore measures the stock, or outstanding balance, of transactions that have not matured or been terminated by the survey date. Except for two questions (see below), it does not seek to measure the flow of transactions, or turnover, over the period between two successive survey dates.
The survey covers all types of true repo, which means agreements in which collateral is sold and repurchased, in other words, where collateralisation is by the transfer of legal title to the collateral rather than by the creation and attachment of a security interest, such as a pledge. Repo can take the form of repurchase transactions, reverse repurchase transactions, buy/sell-backs and sell/buy-backs. The survey does not cover synthetic structures.
The survey asked respondents to divide their data into repo (cash borrowing by the respondent) and reverse repo (cash lending by the respondent), as well as to break down the resulting positions by:
• location of their counterparty
• market segment
• cash currency
• type of contract
• type of repo rate
• remaining term-to-maturity
• method of collateral management
• origin of collateral
• some other categories.
In addition, participants were asked to report the outstanding value and composition of any securities lending and borrowing conducted from their repo desks.
Since 2017, the survey has asked for the number of new transactions and the value of turnover since the previous survey (these are the only questions in the survey which measure turnover). Since 2019, it has also asked for the numbers and types of legal agreements under which entities can transact repos.
Since June 2023, questions have been included in the survey about sponsored, guaranteed and indemnified repo.
An extract of the accompanying Guidance Notes for survey respondents is reproduced in Appendix A.
As well as reports sent by respondents, data have also been contributed directly, since 2003, by the principal dealer-to-dealer (D2D) automatic repo trading systems (ATS) and by the main tri-party repo agents in Europe.1, 2 The latter have also reported tri-party securities lending since 2016. Data are now provided separately by the two principal dealer-to-customer (D2C) automated repo trading systems in Europe.3 These direct sources of data cover almost the entire populations of electronic trading and tri-party management providers in Europe, against which the size, composition and changes in electronic trading and tri-party repos executed by the survey sample can be placed in context. Members of the Wholesale Market Brokers’ Association (WMBA) contributed data on voice-broking directly between 2002 and 2017.
1.2 The response to the survey
The latest survey was completed by 59 entities belonging to 51 financial groups. There were no changes in the survey sample.
Of the 59 respondents in the latest survey, 43 were headquartered across 15 European countries, including members of the EU (35), Norway (1), Switzerland (1) and the UK (6). The EU respondents were headquartered across 12 of the 27 member states (there were again no respondents in the survey from Finland and Sweden, and none from the former Accession States). 31 EU respondents were headquartered across 10 of the 20 countries of the eurozone (which added Croatia in January). Other survey respondents were headquartered in Japan (5) and North America (12). 19 respondents were branches or subsidiaries of foreign parents, most of which (16) were located in the UK.
Many respondents provided data for their entire European repo business. Others made separate returns for one or more (but not necessarily all) of their European offices. Respondents were asked to report for both their UK and EU offices, where they had divided their European business post-Brexit. A list of the respondents that have participated in the ICMA’s repo surveys is contained in Appendix B.
1.3 The next survey
The next survey is scheduled to take place at the close of business on Wednesday, June 10, 2026. Forms will be sent out at least two weeks in advance.
Any financial institution wishing to join the next survey will be able to download copies of the questionnaire and accompanying Guidance Notes from ICMA’s website at www.icmagroup.org/surveys/repo/participate
Entities who participate in the survey will receive a confidential list of their rankings across the main survey categories.
The data received in the survey are used for no other purpose than to inform the survey report. Individual returns are seen only by the author. Only aggregated data are published and ICMA is not permitted to disclose data reported by individual respondents.
Questions about the survey should be sent by e-mail to reposurvey@icmagroup.org.
1 The reporting ATS were BrokerTec (CME), Dealerweb (who joined the latest survey), two eRepo (formerly TP Repo) platforms, two Eurex platforms, MTS Repo (Euronext) and SIX.
2 The reporting tri-party agents were Bank of New York Mellon, Clearstream Banking Luxembourg, Euroclear Bank, JP Morgan and SIS, who together account for the bulk of tri-party business in European repo. Agents not reporting included Citibank and Euroclear UKI (Crest), but the latter provided an estimate of the total tri-party repo positions managed by their DBV (Delivery-By-Value) service.
3 Tradeweb has provided data since 2020 and GLMX, who joined the survey in December 2023, from 2022.
Chapter 2: Analysis of Survey Results
The aggregate results of the latest two surveys (June and December 2025) and of the surveys in each December in the three previous years (2022-2024) are set out in Appendix C. The full results of all previous surveys can be found at www.icmagroup.org
Total repo business (Q1)
The total value, at close of business on December 10, 2025, of repos and reverse repos outstanding on the books of the 59 entities who participated in the latest survey surged to a record high of EUR 13,651.1 billion from EUR 12,435.9 billion in June 2025 (see Figure 2.1). This was a jump of some EUR 1.2 trillion in the aggregate size of the repo books of the survey sample, representing growth of +9.8% since the previous semi-annual survey and +24.6% year-on-year (from EUR 10,956.7 billion in December 2024). These growth rates compare with +13.5% and +11.9%, respectively, in the June 2025 survey. In other words, the sharp pick-up in growth seen in the firsthalf of the year was largely sustained in the second-half.
The drivers of growth in the second-half of 2025 would seem to have been the same as in the first-half of the year, namely, the market volatility and macro-economic uncertainty triggered by the sequence of trade tariff announcements by the US. These not only boosted demand for precautionary liquidity but also prompted investors to seek shelter in the money market.
Despite volatility in other markets, repo market capacity proved adequate and repo rates generally remained stable, including at year-end.4 One reason for the calm in the repo market may have been efforts by central banks to encourage routine use of liquidity facilities without fear of stigma.
2.1 – Outstanding value of total business by the survey sample
Figure
The survey sample as a whole has been a net lender of cash to (and therefore a net borrower of collateral from) the rest of the repo market continuously since 2012, when central banks started to offer longer-term liquidity to sustain the European financial markets in the aftermath of the eurozone sovereign debt crisis (see Table 2.1 and Figure 2.2). However, following the switch by central banks from quantitative easing (QE) to quantitative tightening (QT) in 2022, the net reverse repo position of the survey sample has been gradually unwinding, albeit subject to seasonal fluctuations (net lending temporarily increases at year-end). In the second-half of 2025, gross repo positions of the survey sample grew by +11.6%, outpacing gross reverse repo, which increased by +8.1%.
Figure 2.2 – Total repo versus reverse repo positions of the survey sample
ICMA survey methodology
The survey measures the value of outstanding transactions at close of business on the survey date. While the measurement of the flow of new repo between two dates is useful for some business and market analyses, the stock of transactions outstanding on one date was adopted because it gauges risk exposure and open interest in the market.
Note that outstanding value understates the share of shorter-term repos, given that such transactions run off faster between surveys than longer-term repos. Because repos traded on automatic trading systems (ATS) and cleared on a central counterparty (CCP) are typically very short-term, the consequence is that the share of outstanding balances that have been ATS-traded and CCP-cleared is smaller than their share of turnover. This can be seen by comparing published aggregate SFTR data on new and outstanding repos.
It also needs to be remembered that changes in outstanding balances can reflect cumulative changes in turnover but also variations in the tenors of new transactions or both.
Another important feature of the survey methodology is that it recognises repos from their transaction dates (when they are executed by the two parties and contracts are formed), rather than from their value or purchase dates (when cash and collateral are first due to be exchanged). This transaction-date basis means that the outstanding value measured by the survey includes forward repos, as well as unsettled new non-forward transactions. The latter include one-day repos that are transacted on or before the survey date, but that are not due to be settled until the business day after the survey date (tom/next repo) or on the following business day (spot/next repo). This gives greater weight to one-day repo than would measurement on a value-date basis.
The values measured by the survey are not adjusted for the reporting of the same transaction by two respondents. However, a study by the author (see the report of the December 2012 survey) suggested that inflation due to this problem of double-counting may not be very significant. Interestingly, a trade repository in Europe estimated that two-sided reporting was less than 30% under the EU Securities Financing Transactions Regulation (SFTR) and less than 15% under UK SFTR, which is consistent with the author’s estimate of double-counting.
The survey does not measure the value of repos transacted with central banks as part of their monetary policy operations, but it should include their investment operations in the repo market with survey respondents.
Measuring the growth in market size
In order to accurately gauge the growth of the European repo market (or at least that segment represented by the survey sample), it is not valid to simply compare survey totals. Some changes may represent the entry or exit of respondents into and out of the survey, mergers between respondents or the reorganisation of repo books across respondent banking groups. To offset the impact of changes in the structure and composition of the survey sample, comparisons are also made of the aggregate outstanding positions reported by a sub-sample of those entities who have participated continuously over the last three surveys.
In the December 2025 survey, all survey participants had taken part in the last three surveys, so there was no difference with the headline numbers.
Between June and December 2025, 36 of the 59 entities who responded to the latest survey expanded their repo books (down from 41 out of 59 in June). The repo books of another 20 respondents contracted over the same period (up from 15 respondents in the last survey).
The increases in the size of repo books in the second-half of the year was less dramatic than in the first-half but still strong. Thus, the median position-weighted change in all repo books was +8.3% compared to +10.1%; the unweighted-mean change for the respondents who expanded their repo books was +39.4%, compared with +58.8%; for those who contracted their books, the average reduction increased to -16.0% from -12.3% in the previous survey; and the weighted-mean change across all repo books fell to +16.1% from +17.2%.
The estimated turnover of the survey sample
Respondents accounting for almost 45% of the total value of the latest survey reported their repo turnover over the six months since the previous survey. Grossing up this reported turnover, by the combined share of the outstanding value of the survey of those respondents who did not report their turnover, suggests that the daily average turnover for the whole survey sample over the second-half of 2025 could have been EUR 2,400 billion per day, up +24.1% from EUR 1,934 billion in the previous survey and up +5.4% from EUR 2,276 billion in the second-half of 2024.
As turnover grew faster than the outstanding value of the repo books of the survey sample, there would appear to have been a shortening of the average term-to-maturity, which is consistent with the established seasonality of the repo market. The pattern is also consistent with published SFTR data. However, the estimated turnover of the survey sample remains a number that needs to be treated with caution.
Comparing survey and SFTR data
Data published under the Securities Financing Transactions Regulation (SFTR) in the EU and the UK show that the value, on December 12, 2025 --- the SFTR reporting date closest to the latest survey date --- of all outstanding repos reported to regulators was EUR 14,686 billion in the EU and EUR 11,223 billion in the UK, totalling EUR 25,909 billion (see Figure 2.3). This compares with a total of EUR 25,060 billion on June 13, 2025 --- the SFTR reporting date closest to the previous survey date --- and represents an increase of +3.4% over the second-half of 2025 and +5.4% over the first-half, somewhat lower than the growth in the survey.
However, the steady growth of the overall European market is the net result of different trajectories in the EU and UK. So growth of 12.0% in the outstanding value of repo reported in the EU total in the first-half was trimmed back by a fall of -1.2% in the second-half, while a contraction of -3.0% in the UK in the first-half was more than offset by growth of +10.1% in the second-half.
The size of the survey increased to the equivalent of almost 53% of the EU and UK SFTR total, from some 50% in June 2025 and 46% in December 2024. While any comparison needs to be treated with caution, given the differences in methodologies and coverage, the survey clearly continues to cover a significant share of the European repo market.
Figure 2.3 – ICMA survey versus SFTR public data: outstanding amounts
Turnover in repo reported under SFTR between the week ending December 12, 2025, and the week ending June 13, 2025 --- approximately the same interval as that covered by the survey --- averaged EUR 3,251 billion per day in the EU and EUR 2,421 billion per day in the UK, totalling EUR 5,672 billion, compared with EUR 3,041 billion, EUR 2,376 billion and EUR 5,417 billion, respectively, in the previous semester (see Figure 2.4). This represents a rise in average daily turnover in the European market of +4.7% over the previous six-month average.
Growth was faster in the EU at +6.9% over the second semester and +5.3% over the first semester, averaging +12.6% over the year, compared to +1.9%, +0.6% and +2.5%, respectively, in the UK. Moreover, as turnover lagged outstanding value in the EU over the first-half of 2025 but outpaced outstanding value in the UK over the second-half, there would seem to have been a lengthening of the average tenor in the EU in the first-half and a shortening in the second-half, but the reverse pattern in the UK.
The turnover estimated in the survey (EUR 2,400 billion a day) rose to over 42% of the SFTR total from almost 36% in June but was marginally below the share of 43% in December 2024.
Figure 2.4 – ICMA survey versus SFTR public data: weekly turnover
Trends in place since December 2023 continued over the second-half of 2025. The share of outstanding repo traded on automatic trading systems (ATS) touched an eight-year low of 25.1%.5 The growth of voice-brokered trading reached an 11-year high of 13.7%. There was no change in the share of bilaterally-managed direct business and tri-party managed only a weak recovery.
The outperformance of the voice-brokered share may reflect the fact that voice-brokers provide an immediate and flexible means of extending a dealer’s market coverage when volumes surge, as during 2025. It was also reported
that voice-brokers have also been active in extending the geographical reach of GC financing. And voice-brokers will have benefited from increased interdealer trading in non-European currencies and collateral, as these are not widely traded electronically on European ATS, which compete with voice-brokers for European repo business.
Tri-party repo
The value of tri-party repo positions held by the survey sample reached a new all-time high of EUR 1,431.2 billion, up +12.9% from EUR 1,267.3 billion in June. Their share of the survey recovered to 10.5% from 10.2% in December.6
Gross cash borrowing by the survey sample through tri-party repo, measured as a share of the sum of gross triparty borrowing and lending, fell back to 75.1% from 76.3%. Net cash borrowed through tri-party repo by the survey sample was virtually unchanged at the equivalent of 5.3% of the total survey size. These changes were in line with the reduced net reverse repo position of the survey sample.
Tri-party repo fared worse outside the survey. The value of outstanding tri-party repo reported directly to ICMA by the two ICSDs (International Central Securities Depositories) and SIS (in Switzerland) --- as opposed to the positions reported by the survey sample --- contracted by -3.3% to EUR 906.4 billion, compared with growth of +15.3% to EUR 937.7 billion in the first-half of 2025.7,8 The deceleration in rate of growth in tri-party repo managed by global custodians was a little less severe, dropping by -0.2% compared with +14.1% over the first semester.
Table
2.2
– Comparative trading analysis
GC financing9
The second-half of 2025 witnessed a stellar performance by GC financing. Positions reported by the survey sample leapt by +48.0% over the second-half of 2025. GC financing’s share of the repo books of the survey sample increased to 1.9% from 1.4% in June, and its share of the tri-party business of the survey sample grew to 18.2% from 13.9%. The value of GC financing and its share of electronic trading reported directly to ICMA by the principal ATS in Europe rebounded to a record 12.2% from 8.9% (but note that not all GC financing is traded on ATS), while its share of tri-party positions reported directly to ICMA by the ICSDs and SIS jumped to all-time high of 41.8% from 27.8%.
The bulk of GC financing traded on the Eurex GC Pooling platform. This accounted for 86.6% of GC financing reported directly by the ICSDs and SIS. LCH’s GC financing service, notably its €GCPlus platform (to be renamed GCPlus), accounted for the remainder, having increased share from 11.6% in December 2024 and zero in December 2023 (see Figure 2.6).
6 The actual share of tri-party repo in the survey sample is thought to be higher than the reported share, given that some respondents, who are known to use triparty services, did not report their tri-party business.
7 The ICSDs are Clearstream Banking Luxembourg and Euroclear Bank.
8 The value of outstanding tri-party repo reported by the survey sample (EUR 1,431.2 billion) is larger than that reported by all tri-party agents (EUR 1,173.7). This difference is likely to reflect a number of factors, including the use of US tri-party repo by the survey sample and DBV tri-party repo in the UK, neither of which are reported directly to ICMA (the latter was GBP 115.3 billion in December 2025, equivalent to about EUR 176 billion, down from GBP 150.3 billion in June 2025).
9 GC financing repos are transactions cleared on CCPs and managed by tri-party agents. They may or may not be connected to an electronic trading venue.
Figure 2.6 – Monthly cleared nominal turnover on LCH SA €GC Plus
Source: LCH SA
Note: single-counted, nominal value of collateral
Table 2.3 – Numbers of respondents reporting particular types of business
Automatic trading systems (ATS)
It was noted earlier that the share of outstanding repo transactions that had been transacted by the survey sample across ATS dropped back sharply to 25.1% from 26.5% in June. On the other hand, the value of outstanding ATStraded repo reported directly to ICMA by the principal organised trading venues in Europe reached another high of EUR 2,115.6 billion from EUR 1,862.9 billion in June, growth of +13.6%, compared to +18.2% over the first-half of 2025. Growth in ATS-traded repo was therefore faster than the growth rate of the survey sample. However, the value reported by the venues in December included, for the first time, activity on Dealerweb. This means that the jump from June to December is significantly exaggerated. Stripping out the new Dealerweb data shows that growth in the rest of ATS-traded repo was more modest.
A significant component of ATS-trade repo was GC financing. As noted earlier, the share of GC financing rose to 12.2% from 8.9% of outstanding positions, although the growth in the share of outstanding positions will have been overstated, compared to the share of turnover, by the longer average tenor of GC financing repo. Indeed, turnover gives a distinctly different picture of the competitive position of the venues (unfortunately, the data is too incomplete to publish).
Previous survey reports have suggested that the contraction in the share of ATS repo since the second-half of 2024 reflected the increase in US dollar repo trading by the survey sample, as this would have bypassed European ATS, given their limited US dollar and Treasury trading (the share of dollar positions was just 0.1% in December). The retreat by European ATS in the second-half of 2025 is consistent with that hypothesis, given the recovery
in the shares of both US dollar and Treasury on the books of the survey sample (see the sections below on the currency and collateral composition of the latest survey).
Reports made directly to ICMA by European ATS showed a fall in the outstanding value of Italian government securities (-4.1%) and a drop in share to 40.3% from 46.8% of outstanding ATS-traded repos, although they remain the largest component of ATS-traded repo collateral positions, almost equal to the next three holdings.
The pattern of change in other collateral was also largely reversed in December. German government securities increased share to 15.4% from 12.4% and UK gilts to 13.4% from 8.0%, while French government securities fell back to 12.1% from 13.5%. One exception was Spanish government securities, which continued to increase share, reaching 7.6% from 7.4%.
Non-government securities accounted for just 0.1% of the positions reported by ATS (the most recent high was 0.8% in June 2019).
The first ATS-traded positions in securities issued by the EU were reported.
The monthly turnover figures published by BrokerTec and MTS, the two largest repo ATS in Europe, show the continued narrowing of the gap in market shares until December, when BrokerTec saw an unseasonal burst of growth (see Figure 2.7).
Figure 2.7 – Monthly turnover in European repo on BrokerTec and MTS
Sources: CME, Euronext
At BrokerTec, average monthly turnover fell -2.8% semester-on-semester but grew +3.0% year-on-year. Growth in MTS was +1.8% semester-on-semester and +8.6% year-on-year.
BrokerTec is likely to have benefited from the recovery in trading in German and UK government securities, in which it has the largest market shares, while MTS will have suffered from the contraction in Italian government bond repo, which is largely traded on MTS.
Eurex, which is the third major repo ATS in Europe, made substantial gains in the second-half of 2025. Monthly term-adjusted average daily turnover on Eurex Repo Special and GC (ERSGC) twice exceeded and twice almost equalled the all-time peak reached in October 2023 (see Figure 2.6a). Eurex Repo GC Pooling (GCP) spiked in July and November. ERSGC showed a similar pattern.
Monthly term-adjusted average daily turnover on ERSGC averaged EUR 219.3 billion over the second-half of 2025, up +23.3% from EUR 177.9 billion over the previous half-year, while GCP averaged EUR 230.0 billion per day, up +19.1% from EUR 193.2 billion. A significant contribution to the gains by Eurex came from trading between non-eurozone counterparties. Their share of the data reported directly by the principal trading venues increased to 11.4% from 10.1%.
Figure 2.8a – Monthly term-adjusted turnover on Eurex repo trading systems
On an estimated unadjusted basis, the surge in GCP can be seen to have been largely driven by another dramatic extension in the term-to-maturity of transactions (see Figure 2.8b). The term-adjustment multiplier increased to 7.5 from 7.2 in the first-half. The increased term-to-maturity of positions at GCP likely continued to reflect the supranational institutions and official agencies in its user base, given that these entities prefer longer-than-average investment tenors.
Figure 2.8b – Monthly turnover on Eurex GC Pooling
Turnover on MTS --- which publishes monthly turnover data that are both unadjusted and term-adjusted --showed fairly steady growth in unadjusted activity until the end of the year but a boost to term-adjusted activity from longer maturities in July and September-October (see Figure 2.9). However, the average ratio of term-adjusted to unadjusted turnover at MTS in the second-half was unchanged at 2.6.
Figure 2.9 – Average daily term-adjusted and unadjusted turnover on Euronext MTS
Sources: Euronext
The share of ATS-traded repo positions that had been CCP-cleared, as reported directly to ICMA by the principal venues, increased to 97.4% from 95.6% in June. The share of uncleared positions exclusively reflected SIS.
Automated trading systems
ICMA received position data directly from both of the principal automated repo trading systems operating in the dealer-to-customer (D2C) market segment in Europe. These are GLMX and Tradeweb. The two platforms are believed to account for most of automated D2C trading in Europe.10
The growth in combined turnover of the two platforms over the second-half of 2025 was +15.2% semester-onsemester and +18.8% year-on-year in outstanding value. This compared with +20.3% and a revised +20.7%, respectively, over the first-half, and +25.9% and +18.8%, respectively, over the second-half of 2024. There has therefore been a further deceleration in the growth of turnover and a modest deceleration in outstanding balances over the second-half of 2025, suggesting a continued lengthening of average terms. However, the growth of the automated trading systems remains faster than the market as a whole, reflecting the fact that the burgeoning hedge fund sector is a major user of these automated trading systems but also suggesting continuing inroads into the wider D2C market segment.
On Tradeweb’s European platform, growth decelerated in the second-half of 2025. Outstanding balances contracted to USD 911.7 billion in the fourth-quarter from USD 920.9 billion in the second-quarter (-1.0%), but strong growth in the first-half of 2025 meant that outstanding positions in the last-quarter were +30.2% higher than in the same quarter of 2024.
In terms of turnover, there was modest growth of +3.5% to USD 290.9 billion from USD 281.1 billion over the second-half of 2025, but the strong performance in the first-half translated into growth of +44.6% between the second-half of 2025 and the same period in 2024 (see Figure 2.10).
Figure 2.10 – Monthly turnover and outstanding value in European repo on Tradeweb
Source: Tradeweb
On GLMX, turnover and outstanding balances in Europe, having slowed in the final quarter of 2024, were relentlessly steady thereafter (see Figure 2.11). Outstanding balances grew by +24.3% semester-on-semester in 2025 to EUR 5,017.9 billion and +47.9% between the second semesters of 2024 and 2025. Average turnover between the first and second semesters of 2025 grew by +17.9% and by +37.3% between the second semester of 2025 and the same period in 2024, to reach USD 1,428.0 billion over the last semester.
Figure 2.11 – Monthly turnover and outstanding value in European repo on GLMX
Source: GLMX
Geographical analysis (Q1.1)
Table 2.4 – Geographical analysis of the survey sample
Cross-border positions, particularly into and out of the eurozone, continued to expand over the second-half of 2025, at the expense of domestic and anonymous (CCP-cleared) positions. The growth in this cross-border segment reflects the fact that it includes the trading of non-European currencies and collateral, such as US dollars and Treasuries.
Cross-border positions into and out of the eurozone extended their upward trend to reach a new all-time high and remain by far the largest geographical component of the survey (see Figure 2.12a).
Figure 2.12a – Geographical analysis of the survey sample
Cross-border positions within the eurozone and domestic positions continue to trend sideways. Anonymous (CCP-cleared) business has been trending down since December 2023. The underperformance of anonymous business reflected, to some extent, the growth in the share of non-European currencies and collateral which are not widely traded on European ATS and therefore not CCP-cleared. The close connection between ATS and anonymous trading positions can be seen in Figure 2.12b (see also Figure 2.14a).11
Figure 2.12b – ATS versus anonymous trading by the survey sample
Table 2.5 – Geographical comparisons in December 2025 (June 2025)
In the position data reported directly to ICMA by the ATS, the dramatic jump in the share of outstanding balances between non-eurozone counterparties was consolidated at a record 11.4% from 10.1% in June and 3.7% in December 2024 (see Figure 2.13). This business is largely generated by GC financing.
Figure 2.13 – Outstanding value of ATS business by location of counterparties as reported by the ATS
The share of business with APAC counterparties recovered to 3.5% from 2.9% of the repo books of the survey sample. The recovery may have been helped by increased activity in yen repo mitigating the continued contraction in the share of JGBs, other APAC collateral and other APAC currencies.
Clearing analysis (Q1.2 and Q1.8)
The outstanding value of anonymous (CCP-cleared) repo trading by the survey sample, excluding GC financing, retreated in December 2025 by -11.0% to EUR1,647.4 billion, close to its June level of EUR 1,850.4 billion.
On the other hand, GC financing by the survey sample grew strongly by +46.8% to EUR 265.1 billion from EUR 180.6 billion. Its share of the survey sample increased to 2.0% from 1.5%.12
Overall, anonymous positions contracted by -5.8% to 14.5% from 17.0% of the survey sample, reflecting the contraction in ATS trading, to which CCP-clearing is intimately linked. This relationship is demonstrated in Figure 2.14a (see also Figure 2.12b).
Jan-21 Jul-21 Jan-22 Jul-22 Jan-23 Jul-23 Jan-24
Source: LCH
Note: LCH double-counted nominal value.
Turnover reported by CCP over the first-half of 2025 was lower and more volatile than in the previous semester (see Figure 2.12b). Data from LCH RepoClear showed a contraction of -0.8% in the monthly nominal value of CCP-cleared collateral over the second semester compared with growth of +4.3% over the first. This was the result of a contraction in both LCH Ltd (gilt repo) and LCH SA (euro-denominated repo) of -2.1% and -0.6%, respectively, in line with the overall contraction in ATS trading, compared to -5.9% and +6.4% over the first-half of 2025. Year-on-year growth was -7.9% and +5.8%.
Figure 2.14a – Monthly turnover on LCH RepoClear SA vs BrokerTec and MTS
BTec + MTS LCH SA (RHS)
2.14b – Monthly cleared nominal turnover on LCH RepoClear 2020-25
Source: LCH
Note: double-counted.
The recovery in the outstanding nominal value of collateral balances of repos that were cleared by LCH SA stalled. These balances --- calculated using the ICMA survey methodology --- contracted by -2.0% over the second semester compared with growth of +7.6% in the first (see Figure 2.15). However, balances at LCH Ltd recovered, increasing by +1.5% over the second-half of the year, compared with -5.3% over the first-half. The overall change at LCH was -1.6% compared with +5.9% in the first semester.
Given the weaker performance of turnover compared to balances, these data imply average term of CCP-cleared repo lengthened over the first-half of 2025, which is an expected seasonal change.
Figure 2.15 – Daily outstanding nominal value of cleared repos on LCH RepoClear 2018- 2025
Source: LCH
Note: double-counted.
Figure
At Eurex Clearing AG (ECAG) --- which is the CCP for Eurex Specials and GC Repo (ERSGC) and GC Pooling (GCP) --- and for which outstanding data using the ICMA methodology have been available since 2021, it can be seen that the growth in the outstanding nominal value of CCP-cleared collateral traded on both platforms was interrupted during the summer holiday period, but resumed over the rest of the year. In the second-half, it was ERSGC that led the way, growing by +27.0% compared to +16.8% in the first-half of the year (see Figure 2.16). GC Pooling expanded by +22.2% compared with +17.2%. Year-on-year, growth in GC Pooling accelerated to +58.3% from 27.0% and in ERSGC to +33.2% from 22.2%. The acceleration in growth may have reflected continued expansion out of Eurex’s Continental home market.
2.16 – Daily outstanding values of cleared repos on ECAG 2021-25
GC Pooling Special and GC Repo
Source: ECAG
Note: double-counted, stacked area chart
As shown in Figure 2.8b above, GC Pooling and ERSGC also experienced rapid growth in turnover. GC Pooling also saw a significant lengthening of the average tenor of transactions.
Figure 2.17 shows that the rapid expansion in Eurex Repo reduced the LCH share of outstanding cleared balances to 77.0% of the combined outstanding value at LCH and ECAG from 83.1% in June. These numbers exclude the share of Euronext Clearing (formerly, CC&G), which largely clears Italian government repo, as data are not published for that CCP. However, a significant share of its clearing activity is thought to be passed to LCH SA via a link between the two CCPs. There are also smaller repo CCP-clearing operations in Poland, Spain and Sweden.
2.17 – Daily outstanding value of cleared repos on LCH RepoClear vs ECAG 2021-25 (EUR trillion, doublecounted: calculated using same methodology as the ICMA survey)
Sources: LCH and Eurex
Post-trade clearing
While the bulk of CCP-clearing is of repos transacted on ATS, some trading is executed directly between parties and then registered with a CCP post-trade (see Figure 2.16). Having peaked at 14.1% in December 2020, the share of this post-trade clearing by the survey sample has declined since 2021. However, it recovered slightly to 4.2% in December from 3.9% in June.
Figure 2.18 – Post-trade CCP-clearing by the survey sample
The value of CCP-cleared repo outstanding at end-week in SFTR public data for the UK and EU grew between the June and December ICMA survey dates by +7.7% and +0.6%, respectively, and by +2.1% for Europe as a whole. Between the December 2025 and December 2024 survey dates, the share of CCP-clearing grew by +5.1% in the UK, +15.9% in the EU and +13.4% in Europe. In other words, the growth of CCP-clearing decelerated in the EU but accelerated in the UK between the first and second-half of 2025.
The share of repos outstanding at end-week that were CCP-cleared averaged 44.5% in the EU in the second semester, compared with 43.8% over the first-half. In the UK, CCP-clearing continues to be less important, with a share of 17.2% compared with 16.8% (see Figure 2.20).
2.19 – Share of outstanding CCP-cleared repos reported under EU and UK SFTR (EUR trillion)
The average weekly value of CCP-cleared turnover reported under SFTR between the latest two ICMA survey dates grew by +6.5% in the EU and +6.6% in the UK. In Europe as a whole, CCP-cleared turnover grew by +6.5%. This compares with growth rates of +7.4%, -3.6% and +4.3%, respectively, over the first-half of 2025.
The share of new repos that were CCP-cleared averaged 51.8% in the EU in the second semester, little changed from 52.0% over the first-half. In the UK, there was an increase to 25.2% from 23.8% (see Figure 2.20).
2.20 – Share of new CCP-cleared repos reported under SFTR (EUR trillion)
Changes in outstanding values and turnover over 2025 imply that there has been a shortening in average term-tomaturity in CCP-cleared transactions in the EU but little change in the UK. In the case of the EU, this is contrary to the evidence from the CCP data.
Cash currency analysis (Q1.3 and Q1.4)
The trends in the share of the euro and pound sterling in the survey remain more or less flat. The growth in the share of the US dollar seems to have resumed, largely at the expense of the yen (see Figure 2.21).
Table 2.6 – Cash currency analysis
Figure 2.21 – Currency analysis of the survey sample
In tri-party repo, as reported directly to ICMA by the ICSDs and SIS, the euro continued to recover share, largely at the expense of sterling and the yen (see Table 2.7).
There was a contraction in the share of cross-currency tri-party repo, as reported directly by the ICSDs and SIS (but this reporting may be erratic).
Table 2.7 – Currency comparison in December 2025 (June 2025)
(13.1%)
(1.2%)
(3.9%)
(0.4%)
(1.4%)
(1.1%)
(2.8%)
(8.9%)
(2.7%)
(0.3%)
(0.8%)
Sources: Clearstream, Euroclear, SIS
Collateral analysis (Q1.9)
US Treasuries remained the largest collateral holding of the survey sample and they increased their share of the repo books of the survey sample to a new record of 17.8% from 16.5% in June. Italian government securities remained in second place but with a reduced share of 14.0%, down from 14.4% in June. UK gilts kept third place, with 11.8%, compared with 11.4% in the previous survey.
Figure 2.22a – Collateral analysis of the survey sample
Other eurozone
DKK, SEK
Other EU
EU institutions
UK
US
Japan
Other OECD
IFI
Others
Note: both government and non-government securities
The gap between the share of all US dollar-denominated assets and the larger share of the US dollar widened from 4.2 to 5.2 percentage points, equivalent to about EUR 825 billion. Only some of this gap is covered in the reported number for cross-currency positions (EUR 261 billion). More of the “missing” dollar-denominated collateral may be reported as eurobonds (up to EUR 283 billion) and in unclassified collateral (EUR 559 billion). EUR 459 billion of the latter is tri-party collateral, of which, the tri-party providers report 19.5% to be cross-currency.
The main counterparts to the rising share of US Treasuries were EU government securities and JGBs (see Figure 2.22b). The share of JGBs fell to 2.8% from 3.2%.
The share of government securities as a percentage of all European issues fell to 85.7% from 86.2%.
Figure 2.22b – Collateral analysis of the survey sample13
The contraction of the share of core eurozone securities has been led by Germany, followed by France (see Figure 2.22d). This has reflected erosion of their safe-haven status, heavy issuance, the loss of scarcity value as central banks shifted from QE to QT and political instability in France.
Figure 2.22d – Collateral analysis of the survey sample
The improved share of peripheral eurozone government securities largely reflects the recovery of trading in Italian government issues. From 2012 until 2022, this reflected ECB support. From 2022, it was boosted by the return of Italian banks to the repo market as ECB support was wound down. More recently, it has been driven by improved fiscal prospects and political stability.
Figure 2.22e – Collateral analysis of the survey sample
In the latest survey, German government securities recovered share (9.2% from 8,8%) but French government securities hardly changed (10.0% from 10.1% in June). Spanish government securities lost ground (to 4.0% from 5.4%) but increased their share of ATS-traded and tri-party repo.
On the other hand, the share of all Italian collateral fell back to 14.9% from 15.4%, and Italian government bonds to 14.0% from 14.4%. However, Italian government securities remain second only to US Treasuries in the repo books of the survey sample.
Table 2.8 – Collateral analysis of the survey sample
Securities issued by the EU and held as repo collateral by the survey sample grew slightly, to 0.6% of collateral holdings from 0.5% in December. Their share was somewhat larger in tri-party repo at 6.7%, up from 5.8% in June.
Tri-party collateral
Figure 2.23 – Collateral analysis (selected tri-party agents) by type of asset
In tri-party repo managed by the ICSDs and SIS, as reported directly to ICMA, there was a sharp drop in the share of commercial mortgage-backed securities (CMBS) to 2.4% from 14.4%, reflecting mounting concern over high commercial real estate valuations. The principal beneficiaries were government issues (46.2% from 41.1%), covered bonds (22.3% from 17.5%) and corporate bonds (12.3% from 9.3%). The pick-up in covered and
corporate bond shares reversed a decline from 28.2% and 16.1%, respectively, in December 2023. However, some part of the recovery in the shares of government, covered and corporate bonds is owed to the unidentified “other” category washing out of the data.
Table 2.10 – Tri-party repo collateral analysed by type of asset
Sources: Clearstream, Euroclear, SIS
Equity remains a much larger component of the collateral if the tri-party services of global custodians are taken into account (12.2% compared to just 0.7%, including convertible bonds).
Figure 2.24 – Historic collateral analysis (selected tri-party agents) by type of asset
Supranational Corporate
Securitised Equity
Other
Sources: Clearstream, Euroclear, SIS
French government securities took over from French non-government securities as the largest component of tri-party collateral reported directly to ICMA by the ICSDs and SIS (at 9.2% and 8.5%, respectively from 3.0% and 13.8% in June). UK gilts fell to third position, equal with EU issues (at 6.7% each, respectively from 8.6% and 5.8% in June).
Spanish government bonds increased share substantially to 5.7% from 2.3%. US Treasuries were largely unchanged (4.7% from 4.4%), as were German government securities (3.4% from 3.2%), while JGBs dropped to 0.8% from 2.7%.
Table 2.11 – Triparty repo collateral analysed by issuer --- largest changes (difference in percentage of total reported tri-party balances)
Accelerated recoveries in AAA and A-rated tri-party collateral combined to end the recent resurgence in the share of AA collateral. BBB-rated collateral extended its downtrend (see Table 2.12 and Figure 2.25).
There was a continued modest but widespread relaxation in average haircuts on tri-party credit collateral reported directly to ICMA by the ICSDs and SIS, in particular, for covered and convertible bonds.
Table 2.13 – Weighted-average collateral haircuts (all tri-party agents) analysed by type of asset
In December 2023, the ICMA survey started to ask respondents for the outstanding value of repo business which they had guaranteed or indemnified, including the various forms of “sponsorship”. The share of such repos was little changed at 5.8% of the survey sample in December 2025 from a revised 5.7% in June. The share of the euro in such repo positions fell to 41.4% from 53.7%, while that of the dollar jumped to 52.4% from 38.1%. It should be noted, however, that it continued to be the case that only a minority of survey respondents reported this type of activity.
The share of the ICMA Global Master Repurchase Agreement (GMRA) in the number of repo master agreements in place among survey respondents increased to 87.9% from 85.6%.
Repo rate analysis (Q1.6)
The share of floating-rate repo showed robust growth over the second-half of 2025, to reach 19.5% from 15.7%, just short of the peak of 19.7% seen in December 2023 (see Table 2.15 and Figure 2.27). The latest change confirmed growing expectations of an end to the easing of interest rates by European central banks in the face of rising long-term inflation concerns.
Table 2.15 – Repo rate comparison in December 2025 (June 2025)
The balance between gross borrowing and lending by the survey sample using floating-rate repo shifted more towards borrowing.
Maturity analysis (Q1.7)
There was a movement over the second-half of 2025 out of short dates (one month or less remaining to maturity), except from positions with a remaining term-to-maturity of one day, out to the one to three-month bracket (see Table 2.16 and Figure 2.28). Overall, the share of short-dates fell to 64.4% from 68.4% of the survey sample. The contraction was in line with established seasonal patterns.
Table 2.16 – Maturity analysis of survey sample
Figure 2.28 – Maturity analysis of survey sample
One-month to three-month repo positions held by the survey sample remained highly seasonal, peaking in December, having troughed in June (see Figure 2.29). Significant activity in this residual-maturity band, and to a lesser extent in the three-month to six-month band, is driven by collateral swaps. These are exchanges of securities --- often conducted by means of back-to-back repos and reverse repos --- that are used to manage buffers of high-quality liquid assets (HQLA) required under the Liquidity Coverage Ratio (LCR), especially over end-year.
There continued to be seasonality (in the opposite direction) in open and short-dated repo.
Figure 2.29 – Maturity analysis of survey sample: non-forward terms
Repos with one day remaining to maturity continued to dominate outstanding transactions, replacing two-day to one-week terms, which tended to predominate during the period of QE (see Figure 2.30). Longer short-dates (one-week to one-month) had been trending down over the period of the survey but stepped up in size post-QE. The increased importance of short-dates has been at the expense of medium-term positions with remaining terms-to-maturity of six to 12 months.
Figure 2.30 – Maturity analysis of survey sample: breakdown of short dates plus open
Maturity transformation provided by the survey sample to the rest of the market increased further with a jump in net cash borrowing at a residual maturity of one day and growth in net lending in all fixed terms, but particularly between one week and six months (see Figure 2.31). However, a large decline in net lending between two days and one week reduced the negative funding gap of the survey sample to 5.8% from a record 8.9% of the survey total in June.
Figure 2.31 – Maturity analysis of survey sample: maturity transformation profile --- net reverse repo
Note: Each column represents one survey and each cluster of columns represents the change in the share of a particular tenor over surveys going back to December 2016. The red columns represent the latest two surveys.
In June, there was a shift in the average term-to-maturity of ATS-traded repo positions away from a remainingterm-to-maturity of one day, which dropped to a record low of 77.7% from 93.1% in December 2024, while positions with two-to-seven days remaining jumped to 13.3% from 1.5% and remaining short-dated positions increased to 5.7% from 2.7% (see Figure 2.32). In the latest survey, there was a further shift, but from short dates beyond two days (which fell to 17.0% from 19.0%) to positions with one to six months remaining to maturity (which increased to 4.5% from 2.9%). However, some of the latest redistribution of residual term-to-maturity may reflect the change in coverage of ATS.
2.32 – Maturity analysis (ATS)
Sources: CME, Eurex, Euronext, SIX, TP ICAP
Figure
In tri-party repo, there was a jump in open positions but a modest shift out of short dates (see Figure 2.33).
The ICMA survey measures the securities lending conducted on repo desks as a share of all the securities financing business executed on these desks. The share of securities lending in the latest survey recovered to 8.7% from 8.1% in June, but down from the recent peak of 13.7% in December 2024 (see Figures 2.34a and 2.34b). However, there may be a long-term decline in securities lending from repo desks.
2.34a – Product analysis of the survey sample
Fixed-term securities lending from repo desks recovered to 64.5% from 59.1%. Domestic lending recovered further to reach 27.2% from 20.9% but cross-border lending within the eurozone fell back to 35.5% from 43.3%.
The survey sample continued to be a net borrower of securities through securities loans, except in the case of eurozone equity.
2.34b – Historic product analysis of the survey sample
Table 2.18 – Concentration analysis of the survey sample
Figure 2.35 – Concentration analysis
Top 10
Top 11-20
Top 21-30
Remainder
Table 2.19 – Herfindahl Index for the survey sample14
The slightly higher Herfindahl Index in December was reflected in the shift to the right in the Gini coefficient curve (see Figure 2.36).
Figure 2.36 – Cumulative distribution of market share of the survey sample
Chapter 3: Conclusion
The European repo market, as represented by the ICMA survey sample, extended the surge in growth which built up in the first-half of 2025, to reach a record high of EUR 13,651.1 billion on 10 December 2025. The repo books of the survey sample grew by EUR 1.2 trillion or +9.8% since the June survey.
The drivers of growth in the second-half of 2025 would seem to have been the same as in the first-half of the year, namely, market volatility and macro-economic uncertainty triggered by US trade tariff announcements, which fed demand for precautionary liquidity and prompted investors to seek shelter in the money market. However, despite volatility in other markets, repo market capacity proved adequate and repo rates generally remained stable, including at end-year. This calm may have been due, among other things, to efforts by central banks to encourage the routine use of their liquidity facilities.
Further growth in US dollar and Treasury repo trading had the effect of pushing the share of interdealer positions traded on automatic trading systems (ATS) down to an eight-year low, reflecting the fact that European ATS tend to trade only regional currencies and collateral. US Treasuries continued to be the largest component of collateral in the European market.
Slower ATS trading had a knock-on effect on CCP-clearing. On the other hand, the surge in the growth of the repo market and its strong US component supported the continued growth in the share of voice-brokers, which reached an 11-year high.
The bright spot in ATS trading was in GC financing (which combines CCP-clearing and tri-party collateral management but is often also traded on an ATS). The growth in positions traded in this segment was fueled by new users and increased turnover, but also by the longer-than-average tenors of this type of business. A milestone in the ATS segment was the first trading of securities issued by the EU.
Automated trading systems serving the dealer-to-customer (D2C) segment of the repo market and in particular, hedge funds, continued to grow rapidly, although the rate of growth decelerated.
Tri-party repo was lacklustre, in terms of both market share and outstanding value, despite the attractiveness of repo to investors and the growth in the GC financing component of tri-party. Tri-party repo may have suffered from the ample liquidity elsewhere in the market.
In tri-party repo collateral, there was a sharp drop in the share of commercial mortgage-backed securities (CMBS), reflecting mounting concern over high commercial real estate valuations. There was also a continued modest but widespread relaxation of haircuts on tri-party credit collateral, in particular, for covered and convertible bonds.
The share of floating-rate repo showed robust growth over the second-half of 2025, almost reaching its all-time high, on the back of expectations of an end to the easing of interest rates by European central banks in the face of rising long-term inflation concerns.
The maturity distribution of the market was bipolar. Positions with one day remaining to maturity increased their dominance. At the same time, however, there was a shift out of longer short-dated positions (between two days and one month remaining) into the one to three-month bracket.
Maturity transformation provided by the survey sample to the rest of the market increased further with a jump in net cash borrowing at a residual maturity of one day and growth in net lending in all fixed terms between one week and six months.
About the Author
This report was compiled by Richard Comotto, who is Senior Consultant to the ICMA’s European Repo and Collateral Council (ERCC). He is also author of the ICMA’s Guide to Best Practice in the European Repo Market, Repo FAQs, Repo Market Guides, SFTR Task Force’s Repo Reporting Recommendations and CSDR Cash Penalty Best Practice Recommendations and FAQs, as well as being Course Director of the annual ICMA Professional Repo and Collateral Workshop. In addition, Richard provides technical assistance on behalf of ICMA, IMF, World Bank, Asian Development Bank and other organisations to developing repo markets around the world.
Appendix A: Survey Guidance Notes
The data required by this survey are: the total value of the repos and reverse repos booked by your repo desk that are still outstanding at close of business on Wednesday, December 10, 2025, and various breakdowns of these amounts, as well as the total value of all repos and reverse repos turned over the six months since the previous survey (which was on June 11, 2025).
Branches of your bank in other countries in Europe may be asked to complete separate returns. If your repo transactions are booked at another branch, please forward the survey form to that branch. If branches of your bank in other countries run their own repo books, please copy the survey form to these branches, so that they can also participate in the survey. Please feel free to copy the survey form to other banks, if you discover that they have not received it directly.
Guidance Notes
General guidance
a) Please fill in as much of the form as possible. For each question that you answer, you will receive back your ranking in that category.
b) If your institution does not transact a certain type of repo business, please enter ‘N/A’ in the relevant fields. On the other hand, if your institution does that type of business but is not providing the data requested by the survey, please do not enter anything into the relevant field. If your institution does that type of business but has no transactions outstanding, please enter zero into the relevant field.
c) You only need to give figures to the nearest million. However, if you give figures with decimal points, please use full stops as the symbols for the decimal points, not commas. For nil returns, please use zeros, not dashes or text. Do not use negative signs.
d) Please do not re-format the survey form, i.e. change its lay-out, and do not leave formulae in the cells of the underlying spreadsheet.
e) Include all varieties of repos, i.e. repurchase transactions (classic repos and pensions livrées) and sell/buybacks (e.g. simultaneas and PCT). There is a separate question (see question 2) on securities lending and borrowing transactions (including securities lending and borrowing against cash collateral).
f) Exclude repo transactions undertaken with central banks as part of their official money market operations. Other repo transactions with central banks, e.g. as part of their reserve management operations, should be included.
g) Give the value of the cash which is due to be repaid on all repo and reverse repo contracts (not the market value or nominal value of the collateral) that are still outstanding at close of business on Wednesday, December 10, 2025. This means the value of transactions at their repurchase prices.
h) “Outstanding” means repos and reverse repos with a repurchase date, or which will roll over, on or after Thursday, December 11, 2025. You should include all open repos and reverse repos that have been rolled over from Wednesday, December 10, 2025, to a later date and all forward repos and reverse repos that are still outstanding as forward contracts at close on Wednesday, December 10, 2025.
i) Give separate totals for (a) repos plus sell/buy-backs and (b) reverse repos plus buy/sell-backs.
j) The survey seeks to measure the value of repos and reverse repos on a transaction date basis, rather than a purchase date basis. This means that you should include all repo and reverse repo contracts that have been agreed before close of business on Wednesday, December 10, 2025, even if their purchase dates are later. An unavoidable consequence of using the transaction date is that tom/next and spot/next transactions that are rolled over will be counted more than once, e.g. a tom/next repo transacted on the day before the survey date and rolled over on the survey date will feature twice.
k) Give gross figures, i.e. do not net opposite transactions with the same counterparty. If this is not possible, please indicate that your figures are net.
l) Do not report synthetic repos.
m) You should include intra-group transactions between different legal entities or between foreign branches and the parent company.
Guidance on specific questions in the survey form
1.1 Transactions (1.1.1) direct with counterparties or (1.1.2) through voice-brokers should exclude all repos transacted over an ATS (see below). These should be recorded under (1.1.3). (1.1.1) should include repos executed on automated systems such as GLMX or TradeWeb (which offer a request-for-quote (RFQ) trading model). Repos executed on automated systems should also be included in (1.2.2), which measures all electronic trading.
(1.1.2) Transactions through voice-brokers should be broken down in terms of the location of the counterparties, rather than the location of the voice-brokers.
(1.1.3)“ATSs” are automatic or semi-automatic trading systems (e.g. BrokerTec, Dealerweb, Eurex’s platforms, MTS, eRepo and SIX Repo) but not voice-assisted electronic systems used by voice-brokers (where voicebrokers record and communicate transactions agreed by telephone or electronic messaging) or automated systems such as GLMX or TradeWeb (which offer a request-for-quote (RFQ) trading model). Nor does use of an ATS include trading assisted by electronic means of structured messages and confirmations such as Bloomberg’s RRRA and similar screens.
Transactions on automated trading systems (RFQ systems) should be included in both (1.1.1) and (1.2.2). Transactions through voice-assisted systems should be included in (1.1.2).
Anonymous transactions through an ATS with a central counterparty (e.g. Euronext Clearing (formerly CC&G), LCH, BME Clearing (formerly MEFFClear) and Eurex Clearing) should be recorded in either (1.1.3.4) or (1.1.3.5). (1.1.3.4) is for GC financing systems. These are ATS that are connected to a CCP and a triparty repo service. Examples include Eurex’s GC Pooling (GCP), LCH SA’s €GCPlus/GCPlus and LCH Ltd’s £GC. They do not include GC basket trading on ATS, in which the seller manually selects the securities to be delivered from a list prescribed by the ATS. This activity may be cleared across a CCP but does not involve a tri-party service and should be recorded in (1.1.3.5).
(1.2.1) This item includes all the transactions recorded in (1.1.3) plus any transactions executed directly with counterparties and via voice-brokers which are then registered with and cleared through a central counterparty. (1.2.2) Questions (1.1.3.1) to (1.1.3.5) measure repos and reverse repos transacted on automatic or semi-automatic trading systems such as BrokerTec, Dealerweb, Eurex’s platforms, MTS and eRepo but not voice-assisted electronic systems used by voice-brokers (where voice-brokers record and communicate transactions agreed by telephone or electronic messaging) or automated systems such as BrokerTec Quote, GLMX, MTS BondVision or TradeWeb (which offer a request-for-quote (RFQ) trading model). Business on automated systems should be reported as direct trades in (1.1.1) and included in (1.2.2), which asks for the total value of business transacted on any electronic trading system, whether automatic, semi-automatic or automated, and therefore including automated systems such as GLMX or TradeWeb, which offer a request-for-quote (RFQ) trading model. Electronic trading Is defined in terms of where the contract is executed and so does not include voice-assisted electronic systems used by voice-brokers or trading assisted by electronic means of structured messages and confirmations such as Bloomberg’s RRRA and similar screens.
1.5 “Repurchase transactions” (also known as “classic repos”) include transactions documented under the Global Master Repurchase Agreement (GMRA) 1995, the Global Master Repurchase Agreement (GMRA) 2000 or the Global Master Repurchase Agreement (GMRA) 2011 without reference to the Buy/Sell-Back Annexes, and transactions documented under other master agreements. “Sell/buy-backs” are therefore taken to include all transactions that are not documented. Repurchase transactions are characterised by the immediate payment by the buyer to the seller of a compensatory or manufactured payment upon
receipt by the buyer of a coupon or other income on the collateral held by the buyer. If a coupon or other income is paid on collateral during the term of a sell/buy-back, the buyer does not make an immediate compensatory or manufactured payment to the seller, but reinvests the income until the repurchase date of the sell/buy-back and deducts the resulting amount (including reinvestment income) from the repurchase price that would otherwise be due to be received from the seller. Sell/buy-backs may be quoted in terms of a forward price rather than a repo rate. Where sell/buy-backs are documented (e.g. under the Buy/SellBack Annexes to the GMRA 1995, 2000 or 2011), periodic adjustments to the relative amounts of collateral or cash - which, for a repurchase transaction, would be performed by margin maintenance transfers or payments - are made by adjustment or re-pricing. All open repos are likely to be repurchase transactions.
1.6 “Open” repos, which are reported in (1.7.3), are defined for the purposes of this survey as contracts that have no fixed repurchase date when negotiated but are terminable on demand by either counterparty. Open repos should also be included in fixed-rate repo (1.6.1) unless their repo rates are linked to interest rate indexes which will be refixed during the life of the repos, in which cases, they would be reported as floating-rate repos (1.6.2).
1.7 This section asks for the remaining term to maturity (not the original term to maturity) of repos to be broken down as follows:
(1.7.1.1) 1 day – this means:
• all contracts transacted prior to Wednesday, December 10, 2025, with a repurchase date on Thursday, December 11, 2025;
• overnight, tom/next, spot/next and corporate/next contracts transacted on Wednesday, December 10, 2025.
(1.7.1.2) 2–7 days – this means:
• all contracts transacted prior to Wednesday, December 10, 2025, with a repurchase date on Friday, December 12, 2025, or any day thereafter up to and including Wednesday, December 17, 2025;
• contracts transacted on Wednesday, December 10, 2025, with an original repurchase date on Friday, December 12, 2025, or any day thereafter up to and including Wednesday, December 17, 2025 (irrespective of the purchase date, which will vary).
(1.7.1.3) More than 7 days but no more than 1 month – this means:
• all contracts transacted prior to Wednesday, December 10, 2025, with a repurchase date on Thursday, December 18, 2025, or any day thereafter up to and including Monday, January 12, 2026;
• contracts transacted on Wednesday, December 10, 2025, with an original repurchase date on Thursday, December 18, 2025, or any day thereafter up to and including Monday, January 12, 2026 (irrespective of the purchase date, which will vary).
(1.7.1.4) More than 1 month but no more than 3 months – this means:
• all contracts transacted prior to Wednesday, December 10, 2025, with a repurchase date on Tuesday, January 13, 2026, or any day thereafter up to and including Tuesday, March 10, 2026;
• contracts transacted on Wednesday, December 10, 2025, with an original repurchase date on Tuesday, January 13, 2026, or any day thereafter up to and including Tuesday, March 10, 2026 (irrespective of the purchase date, which will vary).
(1.7.1.5) More than 3 months but no more than 6 months – this means:
• all contracts transacted prior to Wednesday, December 10, 2025, with a repurchase date on Wednesday, March 11, 2026, or any day thereafter up to and including Wednesday, June 10, 2026;
• contracts transacted on Wednesday, December 10, 2025, with an original repurchase date on Wednesday, March 11, 2026, or any day thereafter up to and including Wednesday, June 10, 2026 (irrespective of the purchase date, which will vary).
(1.7.1.6) More than 6 months but no more than 12 months – this means;
• all contracts transacted prior to Wednesday, December 10, 2025, with a repurchase date on Thursday, June 11, 2026, or any day thereafter up to and including Thursday, December 10, 2026;
• contracts transacted on Wednesday, December 10, 2025, with an original repurchase date on Thursday, June 11, 2026, or any day thereafter up to and including Thursday, December 10, 2026, 2026 (irrespective of the purchase date, which will vary).
(1.7.1.7) More than 12 months – this means;
• all contracts transacted prior to Wednesday, December 10, 2025, with a repurchase date on Thursday, December 11, 2026, or any day thereafter;
• contracts transacted on Wednesday, December 10, 2025, with an original repurchase date on or after Thursday, December 11, 2026 (irrespective of the purchase date, which will vary).
(1.7.2) Forward repos are now defined for this survey as contracts with a purchase date of Wednesday, December 17, 2025, or later, in other words, settling on T+5 or later. This definition has been amended to avoid an overlap with corporate/next transactions, which usually settle at T+3 or T+4.
(1.7.3) Open repos in this field should equal open repos in item (1.6.3).
1.8 Please confirm whether the transactions recorded in the questions in (1.6 and 1.7) include your tri-party repo business. Some institutions do not consolidate their tri-party repo transactions with their direct or voice-brokered business because of delays in receiving reports from tri-party agents or the complexity of their tri-party business.
(1.8.1) and (1.8.2) should not include any repos transacted across GC financing systems and recorded in (1.8.3).
1.9 “Eurobonds” (also known as “international bonds”) are defined as securities held outside national central securities depositories (CSD), usually in an ICSD such as Clearstream or Euroclear, or a custodian bank; typically with the ISIN prefix XS; often issued in a currency foreign to the place of issuance; and sold crossborder to investors outside the domestic market of the place of issuance. Eurobonds should be recorded in (1.9.30-33), except for those issues by “official international financial institutions”, which should be recorded in (1.9.20).Eurobond does not mean a bond denominated in euros.
(1.9.20) “Official international financial institutions, including multilateral development banks” such as:
African Development Bank (AfDB)
Asian Development Bank (AsDB)
Bank for International Settlements (BIS)
Caribbean Development Bank (CDB)
Central American Bank for Economic Integration (CABEI)
Corporacion Andina de Fomento (CAF)
Council of Europe Development Bank
East African Development Bank (EADB)
European Bank for Reconstruction and Development (EBRD)
Inter-American Development Bank Group (IADB)
International Fund for Agricultural Development (IFAD)
Islamic Development Bank (IDB)
Nordic Development Fund (NDF)
Nordic Investment Bank (NIB)
OPEC Fund for International Development (OPEC Fund)
West African Development Bank (BOAD)
World Bank Group (IBRD and IFC)
Securities issued by the EU (but not individual EU members) should now be included in the new question 1.9.37. EU issuers include:
European Commission
European Financial Stability Mechanism (EFSM)
European Financial Stability Facility (EFSF)
European Investment Bank (EIB)
European Stabilisation Mechanism (ESM)
European Union (EU)
(1.9.21) “US Treasury” includes bills, notes and bonds, including floating-rate notes, issued by the US central government but not securities guaranteed by that government, such as Agency securities.
(1.9.23) “Japanese government” includes bills, notes and bonds issued by the Japanese central government but not securities guaranteed by that government.
(1.9.25) “Other OECD countries” are Australia, Canada, Chile, Iceland, Israel, Korea, Mexico, New Zealand, Norway, Switzerland and Turkey.
(1.9.26) “Other non-OECD European, Middle Eastern & African countries” should exclude any EU countries. (1.9.34) “Equity” includes ordinary shares, preference shares and equity-linked debt such as convertible bonds.
2.1 This question asks for the total gross value of transactions with a transaction date on or after Thursday, June 12, 2025 (the day after the previous survey date), to and including Wednesday, December 10, 2025 (the latest survey date). In other words, it asks for the turnover or flow of business over the six month interval and includes all business transacted since the last survey date, even if it has matured before the survey date. This section is therefore different from the rest of the survey, which asks for the value of business outstanding on the survey date, in other words, the stock of transactions.
2.2 This question asks for the number of individual transactions with a transaction date on or after Thursday, June 12, 2025 (the day after the previous survey date), to and including Wednesday, December 10, 2025 (the latest survey date), even if it has matured before the survey date. In other words, this is the number of tickets written.
3 This question asks for the cash value of any repos in which the survey participant is not a principal but provides a guarantee, indemnity or similar credit support. This support could be through a facility such as DTCC Sponsored Repo, LCH Sponsored Clearing or Eurex ISA Direct, or could be a bilateral arrangement.
4 “Total value of securities loaned and borrowed by your repo desk” includes the lending and borrowing of securities with either cash or securities collateral. Exclude any securities lending and borrowing done by desks other than your repo desk. If your repo desk does not do any securities lending and borrowing, this line will be a nil return.
5.1 “Active” means about once a week or more often.
For further help and information
If, having read the Guidance Notes, you have any further queries, please e-mail the independent survey administrator at reposurvey@icmagroup.org.
Appendix B: Survey Participants
List of respondents
ABN Amro Bank
Allied Irish Banks
AXA Bank Europe
Banc Sabadell
Banca d'Intermediazione Mobiliare (IMI)
Banca Monte dei Paschi di Siena
Banco BPI
Banco Santander
UniCredit Bank Austria (Bank Austria)
Bank fuer Arbeit und Wirtschaft und Oesterreichische Postsparkasse (Bawag)
Bank of Ireland
Banque de Luxembourg
Banque et Caisse d'Epargne de l'Etat
Barclays Capital
Bayerische Landesbank
BBVA
BNP Paribas
Bundesrepublik Deutschland Finanzagentur
Caixabank (including Bankia)
Caixa d'Estalvis de Catalunya x Bankia SA (formerly Caja de Ahorros y Monte de Piedad de Madrid (Caja Madrid))
CA-CIB (formerly Calyon)
Citigroup
Commerzbank
Canadian Imperial Bank of Commerce and Credit (CIBC)
Commonwealth Bank of Australia
Confederación Española de Cajas de Ahorros (CECA)
Credit Suisse Securities (Europe) Ltd
Danske Bank
Daiwa Securities SMBC Europe
Dekabank Deutsche Girozentrale
Deutsche Bank
Deutsche Postbank
List of respondents
Belfius Bank (formerly Dexia)
Banque Internationale Luxembourg (formerly Dexia BIL)
DNB Bank ASA
DZ Bank
EFG Eurobank Ergasias
Erste Bank der Oesterreichischen Sparkassen
Euroclear Bank
European Investment Bank
Fortis Bank
Goldman Sachs
HSBC
HSBC France
HSH Nordbank
Unicredit Bank Germany (Bayerische Hypo-undVereinsbank)
ICBC Standard Bank
ING Bank
Intesa SanPaolo
Jefferies International
JP Morgan
KBC
KfW
Kingdom of Belgium Federal Public Service Debt Agency
Landesbank BadenWürttemberg, Stuttgart
Landesbank HessenThüringen -Girozentrale (Helaba)
Lloyds Bank Commercial Banking
Lloyds Bank Plc
Macquarie Bank
Bank of America Merrill Lynch
Mitsubishi Securities International
Mizuho International
Morgan Stanley
National Australia Bank
National Bank of Greece
Nomura International
Norddeutsche Landesbank
Girozentrale
Nordea Markets
Norinchukin Bank
Nova Ljubljanska Banka d.d.
Nykredit Bank A/S
Piraeus Bank
Post Italiane
Appendix C: Summary Of Survey Results
Q1 What are the total gross values of cash due to be repaid by you and repaid to you on repo transactions maturing after survey date? (figures in EUR billions)
Of the amounts given in response to question (1) above:
1.1 How much was transacted:
direct with counterparties
• cross-border in (other) eurozone countries
•
in non-eurozone countries
through voice-brokers
• in the same country as you
• cross-border in (other) eurozone countries
• cross-border
on ATSs with counterparties
• in the same country as you
• cross-border in (other) eurozone countries
• cross border-border in non-eurozone countries
• anonymously across a GC financing system
• anonymously across a central clearing counterparty but not GC financing
• total through a central clearing counterparty
• transacted across any electronic system
1.2 How much of the cash is denominated in:
• EUR
• GBP
• USD
• SEK, DKK
• JPY
• CHF
• other Asian and Pacific
1.3 How much is cross-currency?
1.4 How much is:
•
1.5 How much is: •
• open
1.6 How much fixed and floating rate repo is (1.6.1) for value before (survey date) and has a remaining term to maturity of:
• 1 day
• 2 - 7days
• more than 7 days but no more than 1 month
• more than 1 month but no more than 3 months
• more than 3 months but no more than 6 months
more than 12 months
open
1.7 How much is tri-party repo:
• for fixed terms to maturity
total tri-party repo
1.8 How much is against collateral issued in:
Austria
• by the central government
• by other issuers
Belgium
Denmark •
Finland
France
•
Germany
Greece •
Ireland
Italy •
Luxembourg
Netherlands •
Portugal
•
Spain •
Sweden • by the central government
by other issuers
UK
•
other issuers
US but settled across EOC/CS
other countries
Bulgaria
• by the central government
• by other issuers
Cyprus
• by the central government
• by other issuers
Czech Republic
• by the central government
• by other issuers
Estonia
• by the central government
• by other issuers
Hungary
• by the central government
• by other issuers
Latvia
• by the central government
• by other issuers
Lithuania
• by the central government
• by other issuers
Malta
• by the central government
• by other issuers
Poland
• by the central government
• by other issuers
Romania
• by the central government
• by other issuers
Slovak Republic
• by the central government
• by other issuers
Slovenia
• by the central government
• by other issuers
Other EU members by central government
Other EU members by other issuers
• by official international financial institutions
Japan
• Japanese government
• Other Japanese issuers
Other Asian & Pacific OECD countries in the form of fixed income securities, except eurobonds
Other OECD countries in the form of fixed income securities, except eurobonds
issued by Asian & Pacific entities
of unknown origin or type
collateral in tri-party which cannot be attributed to a country or issuer
gross values of repo & reverse repo with APAC
Q2 What is the total value of securities loaned and borrowed by your repo desk: to/from counterparties in the same country as you •
