Hidden leverage refers to financial exposures that amplify risk without being explicitly recorded on a company’s balance sheet. This type of leverage is often embedded in complex financial instruments such as over-the-counter derivatives, which contributed to the collapse of Archegos Capital Management in 2021. Archegos used total return swaps to accumulate significant positions in a handful of stocks. It entered into similar trades with several different counterparties, allowing the firm to leverage its investments substantially. When stock prices moved against these positions, margin calls triggered a rapid unwinding of trades, causing billions in losses for global banks.
In the case of Archegos, individual counterparties were not sighted on the total size of the firm’s sizeable and concentrated equity swap positions, meaning they had imperfect information on which to manage counterparty credit risk. This leads to the fragmentation of risk, with no single institution able see the overall exposure. The fund had an incentive to exploit this opacity, as it allowed it to reduce the concentration add-ons it would otherwise face since prime brokers were unable to fully assess the true concentration of their position.
Trade repository data available to regulatory authorities offers a partial solution to this blind spot. Through the UK EMIR Trade Repository (EMIR TR) data collection, the Bank has access to details of all derivative transactions involving a UK-domiciled counterparty. This data set can be used to build a more comprehensive view of firms’ exposures across products, counterparties and markets. It can be used to identify concentrations of risk and monitor the build-up of hidden leverage within the non-bank financial system. The volume and complexity of the data (approximately 100 million rows of data a day) has required significant investment in the Bank’s data capabilities and upskilling of colleagues to make use of this data most effectively.
Monitoring tools
A key factor in the collapse of Archegos was the speed at which its positions grew. The firm rapidly accumulated large exposures in specific stocks, across multiple counterparties, which outpaced the ability of risk managers to respond. Prior to its default, Archegos ranked among the largest entities by size and growth of notional exposure in equity swaps in the EMIR TR data set. This was similarly seen across other similar data collections in other jurisdictions.
Chart 1: UK domiciled counterparties’ derivative exposures to Archegos
Footnotes
- Source: EMIR TR data.
While a rapid growth in notional exposure does not necessarily signal increased leverage or distress, it can serve as a useful early warning indicator when assessed alongside other supervisory data and risk metrics. In light of these lessons, the Bank now uses EMIR TR data to monitor large and fast-growing exposures in derivative markets. An internal monitoring tool developed by Bank staff tracks the size and growth of exposures across different derivative markets, allowing potential outliers to be identified and explored by the relevant supervisory teams.
The tool makes use of a range of visualisations to support this analysis. An example is shown in Chart 2, which plots each entity’s gross exposure against the three-month growth in their exposures at product level.footnote [1] footnote [2] This allows for comparisons to be made across the market and helps to highlight entities that are large and growing quickly relative to peers. Entities that fall within this high-growth, high-size quadrant are flagged within a catchment area for further analysis.
Chart 2: Illustrative view of large and growing counterparties in equity swaps
Footnotes
- Source: Illustrative EMIR TR data.
These exposures can also be explored at a more granular level. One approach involves examining how an entity’s exposure is split between different counterparties, shown in Chart 3. This helps in identifying whether they are concentrated or are spread across multiple counterparties. While transacting with several prime brokers can enhance resilience through counterparty diversification, it can also create blind spots. When exposures build across multiple prime brokers, no individual counterparty may have visibility of the overall portfolio, increasing the risk that leverage goes undetected, as was the case with Archegos.
Chart 3: Illustrative evolution of derivative exposure for Counterparty X in equity swaps
Footnotes
- Source: Illustrative EMIR TR data.
These exposures can also be broken down further to examine how an entity’s position in individual instruments is distributed across prime brokers, shown in Chart 4. Measures such as the Herfindahl-Hirschmann Index (HHI) can be calculated at the underlying product level to estimate the degree of counterparty concentration associated with individual long or short positions. A low HHI suggests that exposures in a particular product are being split across multiple brokers. This may reflect a conscious effort to diversify counterparty risk but can also contribute to hidden leverage by dispersing positions in a way that makes the overall risk less apparent. This type of fragmentation was highlighted in a Financial Stability Board (FSB) study on leverage in non-bank financial intermediation, which pointed to the equity swap market as being particularly susceptible to these dynamics.
Chart 4: Illustrative HHI and gross notional of entities’ positions in individual instruments
Footnotes
- Source: Illustrative EMIR TR data.
While this approach to monitoring provides insights into trends and outliers from exposures to derivative markets, it may only capture a partial view of portfolio leverage. Many non-bank financial institutions are active across jurisdictions; and positions that do not involve a UK-based counterparty fall outside of the scope of this data set, limiting Bank supervisors’ ability to see the full scope of an entity’s exposure.
Trading strategies can also be highly complex and dynamic, with exposures often hedged using products or markets not visible in this data set. This means the tool cannot fully capture risk. It does not look to establish a direct link between borrowing and investment activity, and may not capture netting, hedging or offsetting positions taken across different instruments. Even if it was possible to fully capture this, high leverage does not always lead to financial stability risks. Therefore, the tool is designed for high-level monitoring of relative changes to provide early signals that can prompt further investigation.
This post illustrates how transaction level data sets, such as the EMIR TR data, are used by the Bank of England to monitor the build-up of leverage in financial markets. While they can offer significant value, jurisdictional boundaries, complexity of trading strategies and gaps in data coverage mean they only form one part of a broader supervisory toolkit. As international efforts, such as the FSB’s leverage in non-bank financial intermediation report, look to strengthen data integration and transparency, these approaches will become an increasingly important part of the supervisory response to monitoring systemic risk posed by hidden leverage.
This post was prepared with the help of Harry Cowell, Gerardo Ferrara, Kilian Lamanna, Will Parry and Manesh Powar.
Share your thoughts with us at BankOverground@bankofengland.co.uk
Data used for Charts 2, 3 and 4 are for illustrative purposes only and do not incorporate actual EMIR TR data.
For interest rate and foreign exchange contracts, exposures are measured using DV01 (the change in value of a contract when subject to a 1 basis point shift in the yield curve), rather than gross notional to better reflect the risk profiles of these assets.
