By Sam Pierson, Director of Securities Finance, IHS Markit
Securities lending revenues for 2018 were the highest since 2008 as demand for equities, credits and government bonds have all trended higher. The challenge for beneficial owners has been that lendable inventory has increased at a faster pace than demand which has spread the returns over a larger base. The total lendable assets reported to IHS Markit crossed $20T for the first time in 2018, an increase of 110 percent since 2009, while loans out of that inventory have only grown by 23 percent.
As the securities lending market continues to grow, strategies for realizing the optimal mix of reward and risk exposure have also evolved. Accordingly, we are taking a fresh look at the measurement of securities lending returns and how they are reported to beneficial owners. At issue are a lack of clarity on what a peer group contains, along with a necessary look at return drivers which standard peer group filters lack the flexibility to capture.
In 2003, IHS Markit Securities Finance, then known as Data Explorers, pioneered the concept and paved the way for transparency in the securities finance industry. Since then the use of data within the industry has greatly increased and is now used not only for performance measurement, but is also embedded in the trading process from automated programs to the management of re-rates and intraday spot-rate checks. Securities finance data also powers income estimates and projections, highlights liquidity and depth of market (particularly for fixed income securities) and aids in the estimation of short interest.
While there has been an increasing use of the dataset across the industry, there is more ground to cover with regard to the analysis of securities lending returns. There have been developments and new functionality, but in the decade following the global financial crisis, this has not necessarily kept pace with the substantial change to the industry and the structure of programs. Even the term benchmark, with respect to securities finance, may be somewhat misleading and, given the focus and reforms across financial market benchmarks, a more fitting description which we will use henceforth is Securities Lending Performance Measurement. As the level of divergence and customization in lending programs has dramatically increased, focus has been on adapting to regulatory change, optimal revenue generation on specials and maintaining income. From a pure mathematical perspective, when using average returns as the benchmark, there should be an equal proportion of under and over performance. However, feedback from beneficial owners typically yields some variation of the “everyone beats the benchmark” refrain.
Sixteen years on from those early beginnings, the time is right for an industry wide focus on the issues that drive these outcomes to establish a global standard so that the industry as a whole, its observers and its participants, can have confidence that there is one global agreed upon methodology.
There are different views across the industry regarding inventory and lendable assets, which can lead to distorting performance measurement outcomes when calculating returns as Return to Lendable. This could mean that two identical Funds, both generating the exact same securities lending income, could have differing Return to Lendable outputs, simply driven by differing views of lendable inventory. Common areas of difference are restricted markets, restricted assets, regulatory lending limits, program-imposed buffers, foreign ownership limits on a company, client mandated holdbacks and buffers. More detailed consideration of these factors is required when considering returns relative to lendable inventory.
There is an inherent challenge in the way that the reference rate for Securities Lending Performance measurement is calculated. In no other part of financial markets is a reference rate calculated where a considerable majority of the input has a zero fee. At the end of 2018, according to IHS Markit, global utilization was 9.24% (lendable $19.1trn; on-loan $2.3trn), meaning that 90.76% of the lendable input had zero fee. The classification of lendable inventory into fee ranges will allow for greater specification of returns, ie. return on specials inventory.
Securities Lending Performance Measurement generally weights the assets across the industry to the same size and form as those belonging to the Fund being reviewed. This can create some distortion, due to different program structures and assets not actively lent, which leads to a higher proportion of Funds outperforming. Whilst the implementation by IHS Markit of a new preferred benchmark is aimed at starting to address this, further work is required. Due consideration must also to be given to alternative Securities Lending Performance Measurement metrics, specifically the inclusion of current SLPM rate and a new active only SLPM rate, which focus on returns on loans made rather than scaling returns by inventory.
Part of the explanation for the “everyone beats their peer group” narrative is the postcrisis emergence of different lending strategies, specifically participation in GC vs specials lending, where traditional peer group designations bely a potential range of strategies. The key to delivering meaningful benchmark reports going forward will be the combination of peer group clarity as well as consideration of style-based peer group filters.
A Fund’s investment objectives will greatly influence how optional trades such as scrips and cash/stock options are approached. A Fund that invests for income and a passive Fund will usually take the cash option whilst a longer-term investor may elect for the additional stock if their mandate allows it. At the portfolio level, the latter will usually have the most value (as the stock is often provided at a discount rate), but those forced to take cash effectively recover some of this discount via securities lending trade. Such securities lending transactions usually have exponentially high income, and that can disproportionately distort overall Securities Lending Performance Measurement. This is especially true when a portion of the universe does not need to engage in such transactions as they are already better off from selecting the stock option.
We view the process of performance measurement as a complement to the increase in regulatory reporting requirements. SFTR has mandated the reporting of some lending transaction details which had previously been challenging for market participants to produce in real-time. Looking at collateral details, the demand for more clarity from the industry has been significant and consistent, as has been the challenges in reporting for market participants. We anticipate that will finally change in 2019.
While we work with clients on the provision of additional collateral metrics we are also working on restructuring the collateral flexibility buckets which define collateral usage for the current performance measurement process. Of particular note is the isolation of equity collateral.
Accuracy versus Complexity
With so many variables that can materially impact Securities Lending Performance Measurement, one possibility is to have significant flexibility that allows greater customization in order that there is an exact match between a Fund and the peer group. However, such customization is harder to manage, can become too time consuming and lead to comparisons to a peer group of one. What then is the right balance?
We’ve taken one step in the direction of a selfselecting peer group based on fund characteristics and size, which was made available in Q4. The removal of funds with limited activity can be thought of as one of the first style-based peer group designations.
When a peer group has been selected for analysis it is important to review the total peer group inventory and returns profile in addition to the returns weighted to the client portfolio. Tracking the returns to the total pool for relevant asset classes supports understanding of the context for returns to the client portfolio and weighted benchmark. One advantage for multi-agent beneficial owners is the ability to generate reports which show top level performance as well as agent specific breakouts, which can be run against a consistent peer group. The disclosure of peer group inputs must be simple enough to understand and easy to replicate.
There are a few key threads here which we are focused on as we further develop the framework for securities finance performance measurement:
1. Clarity of inputs and size of peer group
2. Additional dimensions for specifying program preferences
3. Updated collateral specification
We are encouraged by the feedback we have received from our clients both within agency lending as well as the beneficial owner community for the next generation of securities finance performance management tools. We are committed to delivering a solution for beneficial owners which clarifies the trade-offs in program specification and provides a meaningful performance measurement framework. Progress toward that goal is underway and we’re excited about the roadmap ahead.