Navigating the challenges of fixed income attribution
By Ian Thompson, Global Director of Portfolio Analytics, StatPro
September 25, 2018 | Confluence
Existing members can use the sign in option below.
Bobsguide members enjoy:
By Ian Thompson, Global Director of Portfolio Analytics, StatPro
September 25, 2018 | Confluence
Spend even a short amount of time in finance, and you’ll realize stock and bond investment managers are generally cut from different cloths. The former are often chatty, right-brain types, and the latter more analytical and left-brained.
Both asset classes can accomplish valuable things inside a portfolio, but that’s pretty much where the similarity ends; from a portfolio management perspective, they often serve fundamentally different functions, are heavily impacted by fundamentally different things, and are vulnerable to very different risks.
The investment process also presents vastly different challenges for both. At the risk of oversimplifying the equity world, stocks are relatively straightforward, at least from a data perspective – earnings, valuations, cash flows, growth rates, etc. The fixed income world, in contrast, is full of different flavors of instruments and data, each of which are affected by an extraordinarily complex mathematical web of macro and micro factors that can impact yield and return.
Perhaps just as importantly for the portfolio manager seeking to optimize allocations, figuring out precisely what variable resulted in what specific impact on a particular instrument’s yield (or vice versa) can be a fiendishly complex endeavor.
For fixed income managers, legacy applications and even spreadsheets have been the tried-and-tested method of tracking this mountain of data for years.
Creating complex models to measure how changing variables– such as yield curves – will affect the return outcomes of a subset of fixed income assets is imperfect, labor intensive, prone to mistakes and costly, yet many fixed income shops still look at scenario and attribution analyses this way. The StatPro Revolution platform makes use of new technology to make this process easier for fixed income managers and the wider set of stakeholders to digest.
We realized that we would have to provide a bridge between the performance team and the front office – strong analytics on what has happened in the portfolio on the one hand, coupled with the ability to decompose returns individually, portfolio-wide, and versus benchmarks on the other.
Done correctly, this information helps the portfolio manager understand which decisions have benefitted the portfolio (and by how much), and which ones haven’t.
We know a key challenge is ultimately about managing the data. Legacy applications and spreadsheets aren’t cumbersome because the math they use is incorrect; the sheer volume and complexity of the inputs they need make them unwieldy, slow and limited in scope.
At the same time, the complexity of the market data itself has skyrocketed – derivatives are commonplace, while many fixed income benchmarks have more than 20,000 holdings (compared to 500 in the S&P 500 index). This fact alone makes manual calculations of data points like relative attribution very difficult at best, and nearly impossible at worst. In fact, in our conversations with hundreds of clients, dealing effectively with the data is routinely ranked as a top challenge.
As with other markets, though, technology provides the solution. StatPro Revolution provides standard index data, as well as risk numbers (spreads, durations, yields, etc.) for both benchmark and off-benchmark securities.
It also includes data on not just bonds, but all the various instruments managers can use to gain exposure to this asset class – derivatives, ABS, swaps, swaptions, etc. This is a key distinction: Revolution is not just an “empty box” – we provide both the data and the software. In other words, both the raw information about these instruments as well as the sophisticated analytics to study it.
As a complete solution, the manager is freed up from needing one vendor to provide the data and a second to provide the software. The goal is a system that doesn’t dictate the approach, but one where the user does – ideally one where they can even include their own data if they’d prefer. Finally bridging the gap between systems that come with only their own specific pre-set data, or those delivered with none at all that may take years to properly implement and calibrate.
There is an old saying in finance that says get the data wrong, and the analytics are doomed. Having both under a single roof is important. It adds efficiency, and in Revolution’s case, scalability. The data demands of fixed income are an order of magnitude greater than in the equity sphere, which can mean there is a struggle to compile and calculate overnight and are often at capacity.
Platforms that can dynamically expand through the cloud, like Revolution, can shine in fixed income environments because of their elasticity when it comes to the vast volume of data, daily calculations and entries they can handle without adding additional resources.
This robust data capability means the portfolio manager is freed from the manual work involved in decomposing returns for both on- and off-benchmark securities. This is no small matter. Fixed income managers think about their portfolios in terms of spreads, durations and yields, and while benchmarks often provide this information, off-benchmark positions need to be calculated separately. Without a solution like Revolution, the manager has to construct this information themselves – most likely using workarounds and spreadsheets.
There is another advantage to using a platform for this type of work: Transparency. It is critical to every money manager that the numbers they are using to make investment decisions are accurate, and that they have easy-to-use toolsets necessary to investigate anomalies or go back to in time to research a conclusion. From the portfolio manager’s perspective, this can be quite useful in challenging conclusions, going back into the data to see where and how it was derived, calculated and attributed – all of which better informs future decisions and can help answer difficult questions when dealing with demanding clients.
Transparency is essential in modern institutional finance, and managers need tools that can help, not hinder this requirement.
This brings us to usability. While the old legacy applications and spreadsheets might get a portfolio manager to a similar result (given enough time), they are not very portable – and anyone who has tried to print a 30-column spreadsheet in any readable fashion can attest.
Knowing this, the StatPro team made usability a core priority for fixed income managers. Users are able to customize the analysis screens of their end results in an unlimited number of personal dashboards that mix and match user interface components as desired. Performance, attribution, risk numbers, etc can all be shared in a simple and easy to digest format, and managers can easily and quickly create output that is customized specifically for various internal and external stakeholders.
These are key features that developed as a result of feedback from clients. They wanted to build their own output using customized templates, measures and graphs, not be pigeonholed into the ones we thought they would want. Because of this customization, the number of ad hoc questions from stakeholders trying to understand complex data through the lens of pre-set templates can be reduced.
Ultimately, good fixed income analytics systems are about three things: accuracy, efficiency and flexibility.
The goal of good analytics is to provide actionable intelligence for the manager – whether it is to tweak the investment strategy or verify whether or not the investment process is working as expected.
The answers require robust tools able to crunch through reams of comprehensive data sets. For fixed income managers, it’s a powerful thing to be able to ask good data questions – and to know you can trust the answers.
Flexibility comes with what the manager does with this information. Getting into the weeds in a way that is both customizable and digestible opens a wide range of options normally out of reach for most firms. Even deciding how you’d like to express those analytics becomes possible – the ability, for instance, to re-assign or reallocate effects for particular instruments in an auditable/controlled data set through distinct rule sets allows the analysis to be tailored to the use of these instruments within the investment process.
The challenges of fixed income attribution are nothing new, but they are ever-growing.
However, technology that takes a fresh approach to the problem, ie understanding from the outset that the sheer scale of the data is a nightmare, the benchmarks are massive, the analysis is difficult and time consuming, and the communications lack transparency and customization, can be a significant help.
A modern interface with a modern system architecture leveraging cloud-based capabilities means the day-to-day life of the fixed income manager and analyst gets easier.
Many systems can answer specific questions, but few can help you get the story across. There is also value in being able to go from one extreme to the other – the data sheet on the one hand, and a customizable dashboard analyzing a government sleeve, a corporate sleeve, a private debt and a derivative overlay all at the same time.
Fixed income portfolio managers are typically at home with complex mathematical formulas, but the sheer volume of data and calculations required of them in modern finance makes usage of a platform almost mandatory.
With Revolution, we set out to help them not with just algorithms and grids of numbers, but an entire framework that is vastly easier to digest, eminently customizable, and relies on cloud technology to offer limitless growth. As these capabilities are layered into Revolution across several phases, managers will be able to break down their data into what’s really actionable, relevant and important to the portfolio, and output that information quickly, easily and in any format they’d like. Taken together, that’s a revolution in fixed income analytics, and it means managers have more time to spend doing what matters most – constructing and managing fixed income portfolios that deliver above-benchmark risk-adjusted returns over time.
The A-Z of financial technology solutions