Fixed Income Attribution Whitepaper
Fixed Income Attribution Whitepaper
Fixed Income Attribution Whitepaper
ATTRIBUTION
Portfolio insights through analytical precision
may 2013
RBC Investor Services Limited 2013. RBC Investor Services Limited is a holding company that provides strategic direction and management oversight to its
affiliates, including RBC Investor Services Trust, which operates in the UK through a branch authorised by the Prudential Regulation Authority and regulated by the
Financial Conduct Authority and the Prudential Regulation Authority and in the DIFC through a branch authorised and regulated by the Dubai Financial Services
Authority. In Australia, RBC Investor Services Trust is authorised to carry on financial services business by the Australian Securities and Investments Commission
under the AFSL (Australian Financial Services Licence) number 295018. All are licensed users of the RBC trademark (a registered trademark of Royal Bank of
Canada), and conduct their global custody and investment administration business under the RBC Investor Services brand name.
The information is provided for professional clients. These materials are provided by RBC Investor Services for general information purposes only. RBC Investor
Services makes no representation or warranties and accepts no responsibility or liability of any kind for their accuracy, reliability or completeness or for any action
taken, or results obtained, from the use of the materials. Readers should be aware that the content of these materials should not be regarded as legal, accounting,
investment, financial, or other professional advice, nor is it intended for such use.
/ Trademarks of Royal Bank of Canada. Used under licence. *Trademark of RBC Investor Services Limited.
table of contents
fixed income attribution
01 | Foreword
02 | Fixed income in demand
03 | Generating additional returns
04 | New fixed income tools
05 | Limitations of traditional attribution models
06 | Designing a new attribution solution
07 | Refining the approach
08 | Assembling the blocks
10 | A practical example
15 | Summary
16 | Appendix
18 | About
19 | Contact
foreword
Dario Cintioli
Product Director
StatPro
12%
Equity 40%
Money Market 18%
Bond
20%
Bond 20%
Balanced-Mixed 10%
Equity
40%
Other/Unclassified 12%
Money Market
18%
12%
Equity 37%
Money Market 16%
Bond 24%
Bond
24%
Balanced-Mixed 11%
Equity
Other/Unclassified 12%
Money Market
Non-Federals
DEX U Non-Federals
Both fixed income investment strategies and the instruments themselves are evolving as the quest
to extract greater yields in this low interest rate environment continues. Beyond standard fixed rate
bonds, instruments have been developed that target changes in credit quality, such as FRNs and
CLOs. Others, such as MBS-based derivatives, have been developed to target specific segments in the
mortgage market. CMOs and basket derivatives go further and target more implicit aspects such as
default correlation among issuers.
Fixed income investing requires a different attribution approach that pinpoints the sources of risk
and return. It is clear that traditional sector-based attribution alone cannot adequately help investors
dissect performance figures.
Factor-based models are an improvement although their linear nature does not capture the nuances
associated with fixed income instruments. For example, in a more complex portfolio, it is crucial to
capture all the optionality embedded in callable bonds, the effects of convexity and all the
implications of effective time-to-maturity for mortgage pools. To capture all these nuances, the
engine driving a fixed income attribution system requires a robust library of pricing functions. These
pricing functions can then compute risk (sensitivity) figures consistently and thoroughly. By applying
these sensitivities to actual changes in the market environment, an advanced attribution report can
be produced. In this manner, all non-linear effects are captured and the attribution model aligns with
the investment process.
A clear example is reflected in an assessment of the risk factors affecting a simple domestic fixed-rate
bond, namely interest rates and credit spreads. The pricing function attached to this bond will input
interest rates and credit spreads and a computation process will use this function to output
sensitivity numbers. The sensitivity numbers are modified duration, modified spread duration, and
convexity. These numbers will then be applied to the changes in interest rates and credit spreads that
occurred in the time period in question. The combination of these changes and the bonds sensitivity
to these changes will provide a clear picture of where returns were made.
pricing
functions
risk
sensitivity
numbers
changes
in risk
factors
An important
element of the fixed
income investment
process is portfolio
stress testing.
a practical example
The following example illustrates the concepts discussed previously, from the perspective of both
a passive investor (with more interest in recent fund performance) and an active manager (with
interest in assessing the funds exposure to future shocks and making allocation decisions
accordingly).
The accompanying screenshots depict an attribution platform that shows the performance of a
sample global bond portfolio initially set up as a 50-50 split between government bonds and
corporate issues, from January 2012 to January 2013.
From a historical perspective, it is evident that the portfolio is yielding less in 2013 than in 2012, yet
still above US LIBOR, which could reflect a general movement in rates. What is most revealing,
however, is that a majority of the portfolio returns are coming from corporates and their associated
spread movement.
From the return breakdown, the corporate sectors weighting increased above the original 50%
allocation to 52.5% while also contributing more to the overall return. The portfolios response to
changes in credit spreads had the largest influence on the portfolios returns, as the health of
industrial companies improved. A robust system should provide drill down capabilities in order to
examine contribution sources at a more granular level, as displayed here.
While this detail provides investors with an indication as to what is generating the returns, an active
manager might prefer to use an attribution system and incorporate it into the investment process.
Typically, an investment process has three steps:
1. Set investment objectives and policies
2. Select an investment strategy
3. Allocate funds to individual assets
An attribution system can be integrated into the second step. Once the objectives are established, a
strategy needs to be chosen and implemented. This involves evaluating and forecasting market conditions
then positioning the portfolio in light of this evaluation by allocating to existing or new assets. It is
also important to assess and forecast market conditions at a macro level. Based on a summary of RBC
Economics year-end research, inflation is projected to remain flat and some yield curves are expected to
steepen slightly. For the sample portfolio, the relevant forecasts would be in regard to the health of
the economy. With inflation in check, central banks can focus on employment and growth. Therefore,
GDP, in North America in particular, is expected to rise and credit spreads are expected to narrow.
Portfolio
returns are
dependent on
a variety of
contributing
factors.
Source: RBC Investor Services Investment Analytics Interactive
Please refer to the Appendix for additional economic indicators and insights.
Another important assessment is to investigate the portfolios exposure based on currency. The exposure
graphs of the sample portfolio are categorised by interest exposure and credit spread exposure.
The DV01 figures on page 12 reflect interest rates, spreads and inflation. The sensitivity to spreads in
the US and Canada are higher than in the other major markets.
With market projections and portfolio exposures in hand, allocation decisions can then be made. The
decision-making process can be broken down by factor classifications with risk factors that affect a
portfolio can be classified into yield curve factors, non-yield curve factors, and issuer specific factors.
Based on the example provided, changes in the shape of the yield curve did not contribute
significantly to the overall portfolio return. However, since the economic data presented indicated
the possibility of a slight steepening of the curve in the near future in North America and the
Eurozone (and larger changes in Australia and New Zealand), and also given the general rise in rates,
portfolio positioning can be undertaken in anticipation.
A general rise in rates poses the most concern for instruments with higher duration. From the
interest rate exposure graph and the DV01 graph, North America appears to have higher duration
and is expected to experience greater losses if rates rise. There are several ways a manager can alter
the portfolio. First, existing assets can be shifted to lower-duration assets (from US to Euro, for
example). Second, existing high-duration bonds can be swapped for new bonds in the same category
but with lower duration. And finally, the manager can enter into interest rate agreements, typically
interest rate futures. To lower portfolio duration, a manager would typically short (sell) interest
rate futures.
The expected steepening of the yield curve poses different questions and challenges. Non-parallel
shifts like a steepening are resolved using three common strategies: bullet, barbell, and ladder. A
bullet strategy concentrates the portfolio around a single point on the yield curve. A barbell strategy
concentrates the portfolio around two points on the curve, typically on either side of a potential
bullet and a ladder strategy evenly allocates the assets across the yield curve points.
None of the strategies are exclusive to any yield curve change. That is, bullets may not always be
employed for flattenings, or ladders will not always be employed for shifts. The choice among the
three strategies will depend on a thorough analysis of duration and convexity of each asset and the
anticipated changes in the yield curve at every point. A robust attribution system should provide
the duration and convexity for each bond under such an analysis.
From the yield curve risk factors, the discussion shifts to non-yield curve risk factors, which could
include optionality, volatility and pre-payments. Again, this requires a thorough analysis of each
bond. For example, when assessing callable bonds in light of rising rates, a manager would typically
examine the convexity figures of each bond and assess the impact the negative convexity inherent in
callable bonds. Callable bonds are also affected by interest rate volatility (vega). Both changes in
rates and in volatility can make up a sizable portion of the outperformance between callable and
non-callable bonds. Rising rates also pose issues for mortgage pools as changes in rates can have
simultaneous increases and decreases in prepayments, although not to the same degree. An
attribution system with full pricing functions should be able to provide a manager with convexity
numbers, vega contribution, and the impact of rising rates on the price of a mortgage pool.
Issuer-specific risk factors generally centre on the credit quality of an issuer, which can be affected
by sector, capital structure and company-specific operational concerns. The RBC Economics analysis
included does not break down anticipated credit spread movements by sector, but it generally
implies that there will be a tightening of credit spreads across the board. From the exposure graphs,
the portfolio stands to make gains in an environment of credit spread tightening, more so in North
America than in any other region.
It is important to note that the impact of all risk factors (yield curve, non-yield curve, and issuer) is
simultaneous. While a manager may want to move bonds from higher duration US bonds in light of
rising rates, this decision must be tempered with the realisation that US bonds stand to gain the most
in a credit spread narrowing environment. An analysis of each bond and its corresponding risk
numbers from the attribution system would be helpful in this exercise.
summary
appendix
about
rb c i nves t o r s ervice s
RBC Investor Services is a specialist provider of investor services to asset managers, financial
institutions and other institutional investors worldwide. Our unique approach to domestic and
cross-border solutions, combined with award-winning client service and presence in 15 markets,
helps our clients achieve their ambitions.
RBC Investor Services ranks among the worlds top 10 global custodians with USD 3.0 trillion
(CAD 3.0 trillion) in client assets under administration and is a wholly owned subsidiary of Royal
Bank of Canada, one of the largest and most financially sound banks in the world.
statpr o
StatPro is a global provider of portfolio analytics for the investment community. Our cloud-based
services provide vital analysis of portfolio performance, attribution and risk. Hundreds of investment
professionals use our cloud services directly or through a fund administrator/partner to perform
analysis, reporting and distribution every day.
With nearly 20 years of experience and expertise, we believe analytics should be sophisticated
yet simple and useful as well as secure. StatPro data coverage includes global equities, global
bonds, global mutual funds, most families of benchmarks, FX rates, sector classifications and
much else besides.
StatPro has grown its recurring revenue from less than 1 million in 1999 to around 30 million at
end December 2012 and currently enjoys a renewal rate of approximately 93%. StatPro floated on
the main market of the London Stock Exchange in May 2000 and transferred its listing to AIM in
June 2003. The Group has operations in Europe, North America, South Africa, Asia and Australia
and approximately 350 clients in 30 countries around the world. Approximately 80% of recurring
revenues are generated outside the UK.
contact
statpro
americas
Andrew Peddar
Group Chief Operating Officer, Boston
[email protected]
John Lockbaum
Head, Investor Services, Canada
[email protected]
Brent Reuter
Head, Investor Services, US
[email protected]
europe, middle east and africa
Sbastien Danloy
Head, Investor Services,
Europe & Offshore
[email protected]
Padraig Kenny
Head, Investor Services, Ireland
[email protected]
Simon Shapland
Head, Investor Services, UK
[email protected]
Philippe Legrand
Head, Investor Services, France
[email protected]
Marco Siero
Head, Investor Services, Switzerland
[email protected]
Jos Mara Alonso-Gamo
Head, Investor Services, Spain
[email protected]
Mauro Dognini
Head, Investor Services, Italy
[email protected]
Marc Vermeiren
Head, Investor Services, Belgium
[email protected]
Cormac Sheedy
Senior Executive Officer, Middle East
and Africa
[email protected]
asia pacific
David Travers
Head, Investor Services, Asia Pacific
[email protected]
Andrew Gordon
Head, Investor Services,
Hong Kong and North Asia
[email protected]
Diana Senanayake
Head, Investor Services,
Malaysia and Singapore
[email protected]
Dario Cintioli
Product Director, Milan
[email protected]
Kate Maryniak
Head of Business Analysis, London
[email protected]
James Harkness
Manager Business Development,
Toronto
[email protected]
Chris Leverette
Risk Analyst, Client Services, Toronto
[email protected]
Rachael Cooper
Global Marketing Manager, London
[email protected]
@Rachael_StatPro
Swati Bhoumick
Marketing Manager, Boston
[email protected]
@Swati_StatPro
rbcis.com
stratpro.com