Read this article to understand:
- Why a disciplined risk management approach to portfolio construction is crucial in the current climate
- The drivers of credit returns overlooked by beta-driven investors
- How the methodology used by our investment-grade credit team aims to enhance excess returns and mitigate downside risks
During the post-financial crisis years of tight corporate bond (credit) spreads and low dispersion in returns, positioning across the investment-grade (IG) market has become very reliant on the higher-risk credit beta – the direction of the overall market – rather than credit alpha, the excess return achieved over the relevant benchmark. As a result, excess returns tend to be strong during rising (bull) markets, but ‘widening spreads’ episodes, when markets become more negative and risk-averse, do not generate enough positive outcomes.
We are facing a period of greater macroeconomic uncertainty and market volatility, so a beta-led approach to credit investing will see larger and more frequent drawdowns.1
With spreads as compressed as they are in September 2024, now is a critical time to consider a more alpha-led approach, which maintains a disciplined, beta-neutral portfolio, while targeting high-conviction security and sector opportunities.
Figure 1 shows how, in recent years, spreads have tightened and how far BB corporate bond spreads have converged with the higher-quality BBB bond spreads.
Figure 1: Spread compression – BB-rated bonds inching towards lower investment-grade spreads (basis points)
Past performance is not a reliable indicator of future results.
Source: Aviva Investors, Bloomberg, ICE BofAML, BBB and BB Global Corporate indices, spread to worst. Data as of August 31, 2024.
Given the asymmetric nature of our asset class, protecting from the downside is critical. Gains from credit exposure typically build gradually over time, but drawdowns can be aggressive and spread blowouts relatively sudden. In this article, we explain how our portfolio construction methodology helps prioritise downside protection through an alternative approach to allocating risk, optimising returns and avoiding the beta bias.
Building an efficient and resilient portfolio for all seasons
Clients often ask how we go about constructing our portfolios and how we mitigate the inherent risks. A credit portfolio has a straightforward goal – to gain exposure to the global corporate bond market while maximising yield and avoiding defaults or downgrades as much as possible.
However, it is hard to overstate the importance of portfolio construction, as it can determine the success or failure of an investment strategy. In the current economic climate and credit landscape, an active strategy with a rigorous bottom-up credit analysis approach can play an essential role in mitigating investors’ exposure to broad market swings.
A portfolio with a high beta is more volatile and higher risk than the market
We have developed what we see as a unique portfolio construction method, which does not rely on the direction of spreads to generate positive excess returns. Remember that a portfolio with a high beta is more volatile and higher risk than the market, and is not the most efficient way of deploying risk.
Hence, rather than continually adding more and more high-beta (riskier) securities to outperform the benchmark, we target the same level of volatility as the benchmark and stay broadly within the boundaries of IG, which means limited off-benchmark exposure to other asset classes such as high yield corporate bonds, while at the same time optimising our excess returns.
Though it is never possible to eliminate all risk from a portfolio, the intention is to construct a structurally defensive portfolio against downside macroeconomic risks while aiming to deliver consistent excess returns above the IG benchmark through carry optimisation and security selection.2
The five steps to our portfolio construction process
Step 1: Define the universe
The starting point is to define and simplify the universe into custom buckets of securities with similar characteristics. The investment-grade credit universe as defined by the Bloomberg’s Global Aggregate Corporate Index is large. It contains over 16,000 securities issued by around 3,000 entities.
Standard sector breakdowns allocate securities by industry group without regard for the vast differences between individual bonds
Standard sector breakdowns allocate securities by industry group without regard for the vast differences in credit ratings, volatility, maturity and correlations. We have redefined credit sectors to focus the allocation and bucketing of risk into customised risk buckets that more effectively isolate risk and volatility.
To illustrate, just because Microsoft (AAA) and Dell (BBB) are both in the ‘technology’ sector, it does not mean they share similar risk metrics in the credit markets. Hence, we do not believe they should be viewed through a traditional sector-based approach within the same risk bucket.
Using over 20 years of historic data, we have redefined sectors of the credit universe to more effectively isolate securities into categories with similar risk and volatility characteristics. Figure 2 shows how our custom sector breakdown differs from a typical sector-based approach.
Figure 2: Custom sector framework
Note: For illustrative purposes only.
Source: Aviva Investors, October 2024.
Step 2: Calculate the expected returns
Just like any bond, each of our custom sectors will have a yield to redemption that constitutes the bulk of the expected return and a volatility measure based on the historical data we have captured.
We then add additional return components to define the expected return for each custom sector. These include:
- Sector overlays: These are our forecasts for spreads, derived from macro- and microeconomic views (on the economy and company fundamentals) as well as our analysts’ recommendations for the sectors. This also helps us tilt the portfolio to sectors we like from a fundamental perspective.
- Roll-down capital gains: The gains here are similar to the ‘roll-down’ concept in bonds (the capital gain generated by the natural fall in a bond’s yield as it approaches maturity). We have defined several maturity buckets for each sector (see Figure 2) and can calculate the roll-down impact from one maturity bucket to the next (lower-maturity one).
- Additional returns from cross-currency basis: There are times when it is possible to increase the yield on a bond by taking advantage of the cross-currency basis through foreign exchange (FX) hedging and currency management activities.3 This additional yield can be calculated for each sector.
Figure 3: Inputs in the calculation of expected returns
Note: For illustrative purposes only.
Source: Aviva Investors, October 2024.
Step 3: Set constraints
However, before seeking to maximise the expected return of the portfolio, we must first set some constraints.
We must also set constraints to remain within the limits of the portfolio mandate
Most importantly, we can adjust the volatility to be less than or equal to the benchmark to restrict the tendency to dial up beta.
We must also set constraints to remain within the limits of the portfolio mandate. These are flexible and range from curve, currency and sector limits to considerations such as liquidity.
For example, we can limit the allocation across maturity buckets. We can set a target for duration, so it can only deviate by plus or minus one year from that of the benchmark. To maintain liquidity, we can limit holdings to no more than five per cent of any issue. And we can take exposure to sectors we do not favour, such as tobacco, all the way down to zero.
Figure 4: Constraints
Note: For illustrative purposes only. ESG is considered alongside other risk factors, this is not binding on the investment process and the investment manager retains discretion (beyond the baseline exclusions policy and any specific portfolio parameters).
Source: Aviva Investors, October 2024.
Step 4: Optimise
Once we have defined our sectors, collated the data and set our constraints, we can optimise the portfolio.
The optimisation is a tool coded in Python.4 It aims to allocate to the most efficient risk buckets (our custom sectors) in the IG benchmark that work together to maximise excess return and minimise volatility in the portfolio.
Figure 5: Optimisation objectives
Note: For illustrative purposes only.
Source: Aviva Investors, October 2024.
The constraints limit the bias towards high beta and ensure we remain disciplined with our structurally defensive philosophy.
The tool is run on a monthly basis. The outcome is a framework of reports and charts that analyse each sector’s expected returns, split by different maturity buckets or a high/low beta lens, which we can use to decide where best to allocate risk.
Thus, the output from the optimisation can suggest the most efficient areas where to allocate risk to outperform, as shown in Figure 6.
Figure 6: Looking for optimal carry on a yield curve (basis points)
Source: Bloomberg, Aviva Investors. Data as of August 2024
Step 5: Downside protection
Several attributes of this process ensure a disciplined risk management approach that aims to outperform the benchmark and peers in a widening spread environment, while not giving up the upside during a bull market.
The process creates structurally defensive portfolios (remaining cautiously positioned throughout a market cycle), which are likely to deliver better risk-adjusted excess returns through carry optimisation. This is achieved by the different layers of the methodology.
Risk is typically rotated out of the portfolio to fund a new idea
Firstly, using custom sectors that bucket risk into high and low beta avoids the tendency to pile riskier and riskier securities into the portfolio.
Importantly, new ideas are not added incrementally to the portfolio. Risk is typically rotated out of the portfolio (i.e., individual investments are sold) to fund a new idea. This ensures we remain disciplined in our risk management, but also that we are balancing our best ideas by making the riskiest ideas (high-beta credits) compete with each other for allocation in the portfolio, and not ignoring the low-beta part of the market.
Secondly, the constraints are set to limit volatility from rising above the benchmark, which will add a layer of protection in a downside scenario. The duration, curve and sector limits also protect the portfolio from deviating too far from the objective.
A final downside protection method is built into the optimisation process. It uses scenario testing to stress the portfolio and ensures we do not underperform the benchmark during extreme bull and bear scenarios.
The tool calculates the benchmark’s performance during extreme scenarios
The tool calculates the benchmark’s performance during extreme scenarios (+3 to -3 standard deviation moves) and predicts the impact on returns and the corresponding changes for each custom sector.5 Each custom sector will have a bear-stressed and bull-stressed excess return while the tool ensures that in these scenarios we do not lose excess return versus the benchmark.
Figure 7 is an illustration of a stress test for a sector where we set a number of constraints into the programme such as: limiting excess returns’ volatility to be no more than the benchmark’s, ensuring that under each stressed scenario our excess return is greater than or equal to the benchmark, no short-selling and at least 15 per cent of the portfolio to be in low-beta issues.
As can be seen from the results of the stress test, we aim to not underperform the benchmark in even the most extreme downside scenarios and to outperform it in all others.
Figure 7: Stress scenario testing example
Note: For illustrative purposes only. Portfolio optimisation model. Hypothetical events and data shown do not reflect actual investment results and are not guarantees of future results. Potential returns are based on a number of assumptions, may not be realised and are subject to risk.
Source: Aviva Investors, August 2024.
Using the yield curve to help reduce volatility
Credit spreads do not compensate for duration (interest rate sensitivity) in a spread-widening market. Through the process of optimisation, we can maximise the spread in low duration while also limiting exposure to high duration.
We tend to allocate to high beta sectors and securities in the short-maturity buckets
The results of a typical output from the model are shown in Figures 8 and 9. This is an illustration, using actual spread and return metrics, to demonstrate compensating for spread widening through optimisation of excess returns. It shows that we tend to allocate to high beta sectors and securities in the short-maturity buckets and remain more defensive in the longer-dated maturities.
History has shown this to be the most efficient way to reduce volatility in the portfolio without giving up expected return. As can be seen from the bar chart on the right, the strategy has resulted in a lower drawdown for the portfolio versus the benchmark.
Figure 8: Current benchmark and portfolio (credit spread, basis points)
Note: For illustrative purposes. Credit spread data from Bloomberg for the benchmark (Bloomberg Global Aggregate Corporate Total Return Index Hedged) and Aladdin for the portfolio.
Source: Aviva Investors, Bloomberg, Aladdin. Data as of May 2024.
Figure 9: Drawdown for 50 basis points widening (returns, per cent)
Note: For illustrative purposes. Data from Bloomberg for the benchmark (Bloomberg Global Aggregate Corporate Total Return Index Hedged) and Aladdin for the portfolio. Includes information from an optimisation run for a portfolio in May 2024.
Source: Aviva Investors, Bloomberg, Aladdin. Data as of May 2024.
Why us – why now?
This is a critical time to consider using a manager that incorporates a disciplined risk management approach to portfolio construction and aims to limit the downside should spreads widen from their currently very tight levels.
Our process aims to outperform the benchmark, not through adding riskier, high-beta securities, but by optimising excess returns while targeting the same level of volatility as the investment-grade universe. Crucially, we do not deviate significantly from the IG market by investing in off-benchmark assets such as high yield, contingent convertible (CoCo) bonds or subordinated debt like many of our peers.
Our approach aims to optimise the yields that are available without relying on the direction of spreads to generate excess returns
In a climate of tight credit spreads, our approach aims to optimise the yields that are available without relying on the direction of spreads to generate excess returns, and to protect against the downside in a spread-widening environment.
In spread tightening periods, our process will draw on carry or coupon from the riskier names at the short end of the yield curve to deliver excess returns, while security and sector selection aims to deliver bottom-up alpha. But during spread widening episodes, exposure to high-quality names at the long end of the curve (long-duration allocation) should defend the portfolio against drawdowns by maintaining a core position in treasuries and defensive risk sectors.
While our optimisation tool serves as a useful starting point for our analysis to identify opportunities for risk allocation, it is not the only tool in our investment kit. A crucial element in our investment decisions is to build a portfolio around our best idiosyncratic ideas and tap into the credit research capabilities at Aviva Investors to populate our custom sectors with their recommendations.
We believe our differentiated portfolio construction process can produce more efficient returns and has a low correlation to those of peers. It can lead to better outcomes for our clients in periods of economic downturns and uncertainty. The corollary is that in good times, it is likely to deliver solid, albeit not stellar, returns.
