As we move further away from the turbulent period between 2007 and 2009, interest in credit has increased rapidly as investors globally search for extra return in a low yield environment. Credit can provide this potential extra return through its yield (spread). Since 2009, credit has performed strongly globally and spreads have tightened substantially. This tightening has increased the importance of considering what a reasonable expectation for the return from credit is and what strategies should be used to attain this return while managing the risks.

  • Credit is an attractive asset class that should provide long-term outperformance over government bonds
  • Active management of the exposure to credit can substantially improve performance but a sole focus on a targeted level of outperformance may not be the best strategy
  • This is particularly the case over shorter timeframes where target-seeking can be counterproductive
  • In fact, over the long term, following a mean- reverting strategy will deliver better returns than managing credit in pursuit of a target level of outperformance

Because credit provides a spread over bonds or swap, it is tempting to set a target level for outperformance over one of these measures. However, it is important to consider the strategy which can effectively achieve such a target and a reasonable timeframe for this level of outperformance. In approaching these questions, we increasingly encounter a dilemma that faces investors, especially those measuring performance against a shorter-term benchmark:

Even if we believe that credit spreads may widen and lead to underperformance at some time in the future, can we afford to give up on earning their current spread and underperform if spreads stay stable or continue to tighten for some time?

This dilemma means that, in our view, credit can never be managed using a 'set and forget' strategy. Active management of credit in terms of both exposure and security selection is crucial to achieve extra return without incurring excessive risk that may lead to underperformance. Many market participants believe that managers should target a specific level of outperformance to ensure the highest returns. However, our analysis shows that superior performance is actually achieved by using a mean- reverting strategy and avoiding following the herd.

Credit spreads are variable over time

Since 1995, US credit spreads have ranged by as much as 490 basis points (bps) between peak and trough. Chart 1 shows the average spread level over this period of 137 bps over swap. If we exclude the 24 months from June 2007-June 2009, however, the average drops to 118 bps over swap, an indication of just how severe the GFC was on the performance of credit.

Chart 1: US investment-grade credit spread to swap – Barclays Credit Index

US investment-grade credit spread to swap – Barclays Credit Index

Source: Barclays

The chart also demonstrates that spreads can remain above or below the average for sustained periods – for example, they were below average for four years from March 2003 to July 2007. This means that even if spreads are mean-reverting, the timeframe over which this reversion occurs may be so long that the cost of underweighting credit could lead to poor relative performance before any benefit can be accrued. In other words, although credit seems expensive when spreads are well below average, it can stay expensive for a long time.

As a result, the determination of 'a reasonable spread' becomes more complicated than it might seem at first. The estimate for a reasonable spread must take into account the expected performance of spreads over the holding period as well as the likelihood of defaults and whether the position is liquid or illiquid.

Credit investment strategies: Target-seeking and mean-reverting

How should one invest when spreads are tight and equally how much credit should one hold when spreads are wide? There are a variety of strategies available to investors, but we explore two contrasting representative strategies: target-seeking and mean- reverting. These two strategies are not the only options and, in general, credit managers apply aspects of both strategies but with a bias towards one or other approach.

Target-seeking strategy: Situation-driven investment style

This simple and common strategy basically involves responding to the current state of the market and trying to keep to a target return. Many investment managers are required to achieve a specified return above their benchmark. Under a target-seeking strategy, the level of spread is held more or less stable at a level equivalent to the targeted excess return level.

When credit spreads are tight, the target can be very demanding and the manager is pushed towards increasing risk, i.e. buying more aggressive credit (such as sub-investment grade or long- dated) at tight spreads. If spreads subsequently widen, then this strategy can generate severe capital losses and consequent underperformance.

On the other hand, when spreads are wide, credit is typically out- of-favour and the environment implies heightened risk in credit investment. At such times, a very conservative allocation to credit would be chosen since targeted returns above the benchmark are relatively easy to achieve through running yield even if a less aggressive portfolio is selected.

Mean-reverting strategy: Counter-cyclical investment style

The mean-reverting strategy is the complete opposite: this countercyclical strategy involves overweighting when credit is cheap (spreads are wide and credit is challenged) and underweighting when credit is expensive (spreads are tight and credit is popular). This strategy assumes that when spreads are wide they will tighten towards the mean and when they are tight they will widen towards the mean.

Comparing target-seeking and mean- reverting strategies

The mean-reversion strategy involves buying into falling markets (i.e. when the market is more inclined to sell) and selling into rising markets (i.e. when the market is more inclined to buy). Since this strategy is trading contrary to the market's trend, it is likely to more easily find liquidity and trade at levels close to valuation. On the other hand, the target-seeking strategy involves selling when others are selling and buying when others are buying. As a result, liquidity can be more challenged and when there are large market swings, trades may have to be further from valuation levels.

A significant difference between the two strategies is that with a target-seeking strategy, investment is highest when spreads are tightest whereas with mean-reverting strategies, investments are being minimised at this time. Since earning the spread is the means of recouping capital losses due to spread widening, the tight spreads and large allocation to credit imply considerably longer timeframes for target-seeking strategies to compensate for capital losses due to spread widening. On the other hand, a mean-reverting strategy invests most when spread most protects the investment. Managers may find a mean-reverting strategy uncomfortable at times when they are investing against the direction of the market, but as our analysis shows, this strategy actually delivers superior returns vs. target-seeking over the long term.

Performance comparison

To understand how these two strategies perform, we consider how and when they change their exposure to credit (see Appendix for methodology).

The period around the GFC provides a particularly good illustration of the extreme environments that credit can be exposed to. The magnitude of the performance impact during this period may be more than typically expected, but it provides a good example of the different phases of an extremely stressed environment.

Chart 2 shows the performance from October 2005 – May 2012 of these two strategies. To provide contrast, we show the performance of the Bloomberg Credit Index as well as a Treasury index in which the duration has been adjusted to match that of the credit index.

Often a manager's performance expectations are detailed in terms of a specific level of outperformance over a benchmark. Since there is no asset that can be bought to replicate such a return, we included in our analysis the performance of a theoretical benchmark which has a fixed excess return over (duration-adjusted) swap. The fixed excess is set so that it is the average spread over the period. This margin benchmark is driven by movements in swap rates and does not have the spread sensitivity to which credit is exposed.

Chart 2: Performance of different investment strategies
(October 2005 – May 2012)

Performance of different investment strategies

Source: BLOOMBERG, Nikko Asset Management

If one judged only by Chart 2, the performance of a mean- reverting strategy would seem compelling. However, examining separate periods within the timeframe demonstrates that no one strategy wins in all environments (see charts 3-6). The following charts show that the four strategies had very different outcomes depending on the market environment.

Chart 3: Performance of different investment strategies
(October 2005 – July 2007)

Performance of different investment strategies

Source: BLOOMBERG, Nikko Asset Management

Chart 3 shows that before the GFC, in a stable low-spread environment, leveraging up through a target-seeking strategy provided the best performance. In this environment, it is tempting to earn the extra spread but as Charts 2 and 4 reveal, this overweight can be very negative with a rapid sell-off of spreads more than reversing this outperformance within the space of a few months.

As Chart 4 shows, during the worst of the GFC, from mid-2007 until December 2008 shortly after the Lehman Brothers collapse, the best strategy would have been to have been out of credit completely, with Treasuries significantly outperforming.

Chart 4: Performance of different investment strategies
(July 2007– December 2008)

Performance of different investment strategies

Source: BLOOMBERG, Nikko Asset Management

Chart 5: Performance of different investment strategies
(December 2008– June 2010)

Performance of different investment strategies

Source: BLOOMBERG, Nikko Asset Management

Chart 5 shows that in the 'green shoots' period between December 2008 and mid-2010 when credit was volatile but on an improving trend, the best strategy would have been the mean-reverting strategy. This started aggressively overweight credit, made substantial capital gains and earned substantial spread while reducing the overweight into a rallying market.

Chart 6: Performance of different investment strategies
(June 2010– May 2012)

Performance of different investment strategies

Source: BLOOMBERG, Nikko Asset Management

According to chart 6, in the final period, credit spreads continued to rally and the mean-reversion strategy remained the most effective, but not as convincingly as in the 'green shoots' period.

These results show that different strategies are most effective at different times but suggest that overall the single most successful strategy over the longer term is to invest into credit using a mean-reversion strategy.

With the benefit of hindsight, it is possible to create an 'optimal' strategy, which is the combination of the best-performing strategy in each sub-interval (i.e. target-seeking prior to the GFC, investment in the Treasury index until December 2008 and the mean-reverting strategy from then on). The following table shows the performance of the various strategies over the entire period from October 2005 to May 2012.

Strategy Performance (% pa) Relative to Treasuries (% pa)
Target-seeking 5.7% -0.1%
Mean Reversion 7.6% 1.7%
Credit Index 6.5% 0.6%
Treasuries 5.8% -
Optimal strategy 8.6% 2.7%
Swap + Average Margin 7.7% 1.9%

Source: BLOOMBERG, Nikko Asset Management

It can be seen that the 'optimum' strategy outperforms substantially, although this outperformance is overstated since it does not reflect pricing costs of switching strategies and it is based on perfect hindsight, which is not a realistic assumption. However, it highlights that sticking firmly to any one strategy is sub-optimal and shows the potential upside through actively managing credit.

The analysis also shows that the credit index is not the best performer during any sub-interval and underperforms the mean-reverting strategy by 1.1% over the period, signifying the advantage of actively managing the weighting to credit rather than maintaining a constant exposure.

The analysis also highlights how all of these strategies struggle to achieve performance exceeding the constant average margin above swap. Over the longer period, the worst performing strategy is target-seeking. This highlights a key point in proper management of credit: Focus on the state of the market to determine aggression levels in credit – performance targets should be used as a guide to the average level of aggression and not as the actual target aggression level at all times.

In general, if a strategy similar to the mean-reverting strategy is employed when spreads are tight, the differentiation between aggressive and non-aggressive credit portfolios is small. However, when spreads widen, aggressive credit portfolios would be expected to take increasingly higher overweights to credit.

Conclusion: Astute active management can provide superior credit performance

Our analysis shows that target-seeking strategies can perform well for some period, but over the whole credit cycle they are ineffective. In other words, the most effective investing strategy over a credit cycle is to accept that the target will not be met when spreads are tight and seek to compensate through more aggressively taking advantage of spread widening.

We would also conclude that focusing on the difference from an 'average' spread is more effective over the longer term than focusing on a targeted level of outperformance. The mean- reverting strategy may be the most successful, providing superior returns to that of either a static credit portfolio or a focus on achieving a target return.

However, investors should note that the reversion can be very slow to happen and the strategy can underperform for a substantial period of time. Active monitoring of risk is vital when managing credit in times when spreads are far off the average. This is because it can be help to add extra value over the mechanistic mean-reverting strategy described above–if risk and spread factors are more carefully incorporated into the process, it is possible to further move towards the optimal portfolio strategy.

APPENDIX: Summary of Methodology

The base case is to compare the Treasury index to the credit index by first looking at the Treasury index and duration-adjusting it to match the credit index. The incremental monthly returns of the credit index against this adjusted Treasury index represent the credit-to-bond return.

Monthly returns for each strategy are then calculated as the Treasury return plus an excess return, which is

calculated as the exposure ratio multiplied by the excess return of the credit index over the Treasury return.

The exposure ratio is calculated differently for each of the two strategies:

• For the targeting strategy, the ratio is defined as the exposure needed to ensure the portfolio average spread matches the targeted spread; and

• In the mean- reversion strategy, the spread is defined as the current spread divided by the average spread.

In this discussion, the targeted spread is the same as the average spread, namely 111 bps.

Example: if the current spread for credit is 80 bps then:

• For the targeting strategy, the exposure ratio is 111/80 = 1.39 (approximately); i.e. the fund is 39% overweight credit; and

• For the mean-reversion strategy, the exposure ratio is 80/111 = 0.72 (approximately); i.e. the fund is 28% underweight credit.

The calculations use monthly returns to compute performance. This is equivalent to a trading strategy of monthly reweighting back to the calculated exposure level.

Although this methodology is quite simple, it provides a solid indication of how targeting and mean-reversion strategies perform comparatively. It does not of course capture all the possibilities that more regular and less mechanistic strategies could provide.