Orbisa: Actionable Securities Lending Intelligence

Informative metrics and customised projects

Hamlin Lovell
Originally published on 30 May 2023

EquiLend is a leading capital markets fintech company founded in 2001, which offers a wide suite of technology and data solutions for the financing industry – specifically securities financing, including trading, regulatory compliance, post-trade and data and analytics services. 

Its award-winning data business began in 2012 with the flagship DataLend platform, which is geared to wholesale players such as beneficial owners, agent lenders and broker dealers. The first foray into the buy-side came in 2021 with the launch of Orbisa, a leading securities lending data platform for hedge funds and other buy-side players. “Our wholesale data has now been transformed into a data and analytics solution for the buy side,” says Nancy Allen, who heads EquiLend’s Data & Analytics Solutions division.

Orbisa has generated datasets curated to hedge funds and the broader buy-side community based on data derived from various datasets, some of which are exclusively sourced from the firm’s market activity.

The data can feed directly into the investment process and help to generate buy and sell signals and construct portfolios.

Nancy Allen, Head of Data & Analytics Solutions, EquiLend

In addition to Allen, Orbisa’s seasoned and hedge fund savvy team includes: 

  • Mike Doyle, Director of Sales North America, who previously managed hedge fund relationships at some of the largest prime brokers including BAML, Goldman Sachs and Morgan Stanley;
  • Grant Davies, Director of Sales EMEA, a securities lending, swaps and collateral management expert with experience from BNY Mellon, J.P. Morgan, HSBC, Goldman Sachs and iShares;
  • Dimitri Arlando, Head of Data & Analytics Solutions EMEA, former BNY Mellon, State Street and Northern Trust securities lending relationship management executive;
  • Michael Gentile and Matthew Bernard, Orbisa consultants, who previously co-founded a hedge fund data management company that was acquired back in 2016; and 
  • a team of data and analytics specialists across EquiLend’s global locations

The appeal is broad: “Hedge fund users of the data include those managing quantitative global macro, equity long/short, equity market neutral, credit long/short, convertible bond arbitrage and merger arbitrage strategies. Clients are in all regions, and they range in size from start-ups to multi-billion-dollar funds,” says Allen.

The most obvious use case is keeping abreast of the cost and availability to borrow for existing and potential short positions, and gauging the risk of “short squeezes”, which have recently spelled the nemesis of some funds and generated triple-digit annual returns for others. “Indeed, the data can also feed directly into the investment process and help to generate buy and sell signals and construct portfolios,” points out Allen, who also serves as chair of the New York chapter of the Women in Securities Finance (WISF) organisation. 

$30trn

The Orbisa data set comprises USD $30 trillion of physical lendable securities, including USD $8 trillion of fixed income securities.

Data coverage, sourcing, cleaning and validation

The Orbisa data set comprises USD $30 trillion of physical lendable securities, including USD $8 trillion of fixed income securities, of which around USD $2.5 trillion (USD $1.5 trillion of fixed income) is on loan. The data coverage is global, including activity from all active lending markets worldwide. Equity securities cover all market capitalisations, including some small caps and micro caps that are lent – or could be. The fixed income securities are mainly government bonds and corporate bonds, including investment grade and high yield, which can be lent through GRMA repos or GMSLA stock loans. The data is sliced and diced by asset, region, sector, company and issuer. 

“We believe that our data coverage is as close to comprehensive as is possible, based on flows and comparisons. We have specifically received positive client feedback on Asian market coverage, where lending data is otherwise hard to come by, confirming our own data,” says Allen.

The data comes from a unique combination of sources from across the EquiLend ecosystem: client-submitted data, trading data from the NGT (Next Generation Trading) trading platform for electronic trading of equities and fixed income, and post-trade data collected every 15 minutes from the client base. Clients not only supply a portion of the raw data, but also help to shape how the output is designed: “We have a symbiotic relationship with clients. Our product team consults clients to understand their needs, and our specialist team works towards specific client needs for optimising alpha and returns in the lending market,” says Allen. Clients appreciate the quality of the data: “Clean and accurate data is critical, and we have had positive feedback. We invest in data validation and cleansing, which includes measures to avoid double counting,” she adds. 

The data can also feed directly into the investment process and help to generate buy and sell signals and construct portfolios.

Nancy Allen, Head of Data & Analytics Solutions, EquiLend

Metrics monitored 

This data generates metrics gauging changing availability, scarcity and the cost of borrowing a security. Key metrics include lendable inventory value and quantity, measured through unencumbered long positions, as well as loan value, availability and utilisation rates – the amount of inventory being lent, drilling down to the individual security. These can provide an early warning for squeezes: “As securities approach 100% utilisation, fees can increase,” says Allen. Hedge funds wanting to avoid “overcrowded” shorts may heed daily short interest measures on loan quantity versus publicly available float. Some short squeezes arise from shortages of stock for closing out shorts, and here days to cover (on loan quantity against 30-day average trading volume) may be informative. 

Fee levels are shown in the Orbisa rate, which indicates borrow rates for the buy-side, brokers and wholesale brokers. Fee trends in terms of long-term and short-term fee momentum and volatility are monitored. “Depth” of borrowers and lenders means the number of each, which may also give clues about market dynamics. 

Lending market flashpoints and heatmaps vary between regions. In 2022, the largest sources of equity lending revenue in the US included so called “meme stocks” such as Lucid Group, Beyond Meat, GameStop and AMC. In contrast, in the EMEA region, the top ten sources of lending revenue were dominated by blue chip large cap stocks, such as Total, Axa, ASML and BNP Paribas. Average lending costs differ between markets: for instance, last year equity borrow rates in Sweden cost roughly twice as much on average as in France, Germany, United Kingdom and Switzerland, while lending costs in Asia Pacific are generally higher than either the Americas or EMEA. In Asia Pacific, South Korea and Taiwan taken together provided eight of the top ten equity lending revenue stocks in 2022, when short sellers were focused on technology cycles and valuations.

Repos and voting

The claw-back of stock borrow can ruin a short trade. Lenders may repo securities for various reasons, such as exercising proxy voting on equities. “Beneficial owners understand the opportunity costs of not voting. They partner with agent lenders to make sure they can recall to vote,” says Allen. 

Security borrow quantities and prices can be fixed for a term or may be open – and can then fluctuate. This is reported, as are re-rates, which show changes in security financing rates, as markets heat up or cool off. 

Orbisa highlights this market colour but does not locate specific sources of the borrow, since the data is anonymised.

This data is also aggregated to identify market trends. EquiLend’s 2023 The Purple publication tracked nearly USD $10 billion of revenue for securities lenders in 2022, which was the highest wholesale annual revenue figure for beneficial owners since 2018, which in turn had the highest since 2008. “It was a record year for corporate bond revenues, as rising interest rates and credit spreads increased demand for shorting them. Special collateral was an important driver,” says Allen. “Hot” securities (costing between 101 and 300 basis points to borrow) and “special” securities (costing over 300 basis points) combined contributed over half of revenues.

Delivery and ongoing innovation

The data can be easily accessed through various portals: raw data feeds, Excel add ins, API, and user interfaces, on screens, desktops and proprietary systems. It can also feed into trading algorithms. All securities have standardised market identifiers, which can be easily integrated into client systems and analytics.

Most clients pay fixed subscription fees for the data, but some also opt for bespoke consultancy projects. Datasets are being enhanced all the time with new tags, initiatives and metrics. For instance, time stamping was added recently. A new metric, the Liquidity Score for Orbisa, ranks 23,000 corporate bonds by ease of borrowing, based on a range of criteria, including utilisation, shares outstanding, lender concentration and fee.

An ongoing project is honing and refining user interfaces and APIs to further improve flexibility and customisation.

More broadly, additional sources of data could soon give Orbisa more pixels to play with as the US Securities and Exchange Commission (SEC) looks to publish a cut of securities lending data: “The industry welcomes transparency for the benefit of all market participants – and we expect that the SEC 10c-1 proposal on securities lending transparency will enable us to provide even more actionable insights to our clients,” says Allen.