Fixed Income What risks should we worry about? Simon Doyle, Head of Fixed Income & Multi-Asset

Transcription

1 February 2016 Schroders Fixed Income What risks should we worry about? Simon Doyle, Head of Fixed Income & Multi-Asset Overview When former Prime Minister Malcolm Fraser said that life wasn t meant to be easy he wasn t referring to bond investors. That said, this phrase sticks in my mind each time I turn on my Bloomberg screen and try to make some sense of the machinations of what was once thought of as the boring asset class. Before going further it needs to be said that we have found the going tough. We have not added value against the benchmark for the last year or so and both our investment process, which is anchored in value and some of the assumptions underpinning it, have been at odds with the pressures driving bond markets (particularly in the sovereign space). Likewise we ve been unable to satisfactorily reconcile the fundamental requirements of a defensive fixed income investor with a benchmark that has significantly changed its spots. As a consequence, we ve focussed more on absolute risk ahead of tracking error. In other words, how much duration risk we ve held has had to be considered alongside our relative duration positioning. On one level this has worked (volatility has been low and returns low but OK in absolute terms), but it has meant underperformance against the benchmark as economic uncertainty and central bank experimentation has changed the landscape and continued on balance to suppress yields. This article attempts to explain how we are managing fixed income in the current environment and to put some context around our recent performance. Our thinking may not be traditional, but in our mind it is logical and reflects how over time we expect to outperform the benchmark and deliver the outcomes that we believe our clients expect. Back to Basics The fixed income asset class is riddled with complexities that essentially don t exist in other assets like equities. Concepts of duration, convexity, curve and carry dominate investment discussions, while it s easy to get caught up in the fact that each security will vary in terms of factors such as credit quality, legal terms, coupons, maturity date and liquidity. At the end of the day though, these fixed income securities are inherently just loans. Furthermore, in making a loan, a rational investor should expect that at some future date their principal will be returned and they will receive adequate compensation for taking on the risk of the borrower (even if a sovereign) and for not having access to our capital for the loans duration. As a fixed income fund manager we are therefore lending our clients money to governments and corporates and should expect that in doing so we / they are appropriately compensated. Park this thought as I will come back to this shortly. The second key idea to explore is the appropriateness of fixed income benchmarks. I need to be careful here as it is easy to argue that our relative underperformance of late is the benchmark s fault. That s not my intention. However it is important to understand how there can be a significant misalignment between the fundamental objectives of fixed income investors (both in return and risk space) and the characteristics of the benchmark. Like equities, common fixed income benchmarks (like the Bloomberg AusBond Composite Index) broadly reflect the market portfolio, in terms of investible debt securities on issue and their characteristics. While it may make sense for equity investors, I d argue that owning the market portfolio in fixed income is intuitively flawed. This is because the market portfolio in a debt context is defined by the preferences and behaviours of borrowers (the bigger the borrowers, the bigger the exposure), both in terms of composition and characteristics. It is logical that as a borrower I want low borrowing costs, limited caveats and typically extended terms (long duration). However, as a lender it is logical for me to want adequate compensation, strong protections and my money back in a time frame commensurate with my investment goals. In other words, the link between the market (or benchmark) portfolio and my objectives as a defensive investor is tenuous.

2 Duration (yrs) Benchmark Yield (%) While in theory, it is possible to argue that the market should ultimately align the interests of borrowers and lenders meaning the market portfolio will reflect the composition and pricing of both sides, the flaw in this argument is that not all participants are equal. The challenge at present is that both the highly leveraged global economy and the intervention of central banks (the unequal participants) through both zero / negative interest rate programs and Quantitative Easing (QE) have distorted the market, driving a wedge between risk and return in the benchmark portfolio. In other words, the gap between the requirements of lenders and the characteristics of the main fixed income benchmark is historically wide. An examination of the main Australian composite benchmark (the Bloomberg AusBond Composite Index) shows that these issues are at play here, even though our policy framework is relatively normal. Not only is the benchmark becoming more concentrated (and therefore less diversified) as government issuance dominates (government/quasi government now constitutes around 88% of the index), it is also lengthening in duration as issuers (including the government) look to lock in what are historically low yields. For example the Commonwealth Government is now issuing 20 plus year bonds and is looking to further extend its issuance maturity profile. At the same time the prospective returns to investors for providing the funding is historically very low. Figure 1: Bloomberg AusBond Composite Index: Duration v Yield Bloomberg AusBond Composite Index - Duration v Yield 1% 9.0% 8.0% 7.0% 6.0% % % % % Benchmark BM Yield Source: Datastream as at 31 January 2016 The changes to the characteristics of the benchmark have significant implications. For example, in 2010 benchmark duration was around years and the benchmark yield around 5.5%. This meant that for an investor holding the benchmark portfolio, a 1% rise in yields would result in their 1 year return declining to 2% (5.5%-%). In contrast a 1% rise in yields today would result in returns for the equivalent period declining from 2.7% to -% (2.7%-4.7%). Another way to think about this is that for an index investor to see a negative return over a 12 month period, yields would only have to rise by around %. In 2010, a negative return would have required yields to rise by 1.6% (over 3 times as much). These distortions are equally significant in global benchmarks. Schroder Investment Management Australia Limited 2

3 Figure 2: Barclays Global Aggregate (Duration and Regional Composition) Duration (Yrs) Barclays Global Aggregate Yield and Duration Yield (%) Global Agg Duration Global Agg Yield (RHS) 0 Source: Datastream as at 31 January 2016 Barclays Global Aggregate Bond Index (By Country of Issue) Non-Japan Asia Australia Other Yield = 0.25% Duration = 8.6yrs Yield = 2.27% Duration = 5.5yrs Japan US UK Europe (Ex UK) Yield = 0.79% Duration = 6.4yrs Other Americas Source: Datastream as at 31 January 2016 It s appropriate now to return to my earlier point about lending our clients money and achieving adequate compensation. Schroder Investment Management Australia Limited 3

4 Percent In late January the Bank of Japan (BoJ) implemented negative interest rates on certain deposits adding to the growing list of central banks with a negative interest rate policy. The subsequent market reaction when coupled with increased uncertainty about the broader global growth outlook dragged Japanese bond yields lower. At the time of writing (11/2) yields on Japanese government bonds out to 10 years were negative. Likewise yields on German government bonds out to 8 years were also negative. This is worth thinking about for a moment. What it means is that if we lend money to the Japanese or German governments we will effectively be paying interest to do so and the shorter the term, the more we will be paying. While the key markets of Australia and the US are not there yet, the influence of the policy response in significant markets like Japan and Europe is dragging down bond yields across the globe. Figure 3: Comparative Global Yield Curves 0 Global Sovereign Yield Curves Cash 1Yr 2Yr 3yr 4Yr 5Yr 6Yr 7Yr 8Yr 9Yr 10Yr Aust US Ger Jap Source: Datastream as at February 2016 This is only rational from an investor s point of view under a scenario where real yields remain positive or where deflation is likely to intensify causing further declines in yields (into more negative yields). Even if this proposition is accepted, the question is how much exposure to low / negative yielding instruments is appropriate. This is where benchmarks come in. As duration has extended, investors linked to benchmark duration positions gain more exposure to this theme (directly or indirectly as in the case of Australian investors). Squaring the circle It is fair to say we did not expect yields to stay as low as they have for as long as they have. Our assessment of fair value that drives our forecasts is based effectively on nominal growth expectations and assumes a normal policy environment. With oil prices dragging down inflation expectations and central bankers rewriting the policy framework seemingly daily, these assumptions have not been met. Clearly a more benign nominal growth outcome would have given us more optimistic return forecasts for sovereign bonds but even if we d forecast lower yields, this would not have fundamentally altered the premise on which we ve constructed the portfolio. While positioning and performance against the benchmark is important (relative risk), I believe that it is the absolute risk position and the reward for taking it that matters most especially in the short to medium term. This is particularly pertinent in the current environment where asset prices are being distorted by significant central bank intervention. Schroder Investment Management Australia Limited 4

5 Duration (yrs) Duration - Relative (yrs) As Figure 1 (above) highlighted, benchmark duration has increased significantly as yields have plunged. However, as Figure 4 (below) shows, the absolute level of duration in the portfolio has remained relatively constant. In effect this means that while our relative to benchmark risk has increased over this timeframe (given our material short vs benchmark), our absolute level of interest rate risk has not. This implies that had we not maintained a significant, and over time increasing, short duration position against the benchmark, our absolute level of duration risk would have increased materially something inconsistent with the continued declines in bond yields. Figure 4: Duration Schroder Fixed Income Fund v Bloomberg AusBond Composite Index Portfolio Duration: Absolute v Relative Portfolio Relative Source: Schroders, Bloomberg as at 31 January 2016 Another interesting way to look at this is to look at the yield received for each basis point of duration. This is a proxy for how much reward we are receiving for each unit of interest rate risk. In FY 2010/11 owning the benchmark portfolio returned to the investor approximately bps of yield (based on the 10Yr bond) for every bp of duration risk. The equivalent number now is around 0.6bps. In other words owning the benchmark portfolio is both signficantly more risky and signficantly less rewarding. Figure 5 shows how this ratio has changed over the period since inception of the Schroder Fixed Income Fund in Schroder Investment Management Australia Limited 5

6 Credit Government related Absolute duration contribution (yrs) BPS/Unit of Duration Absolute duration contribution (yrs) BPS/Unit of Duration Figure 5: Reward / Risk Ratio Bloomberg AusBond Composite Index - Duration v 10Yr /Duration Schroder Fixed Income Fund - Duration v 10Yr/Duration Benchmark 10Yr/Dur(BM) Portfolio 10Yr/Dur(Port) Source: Schroders, Bloomberg as at 31 January 2016 Diversification The final point on the benchmark is the lack of diversity. As Figure 6 shows, the relative diversity of the Bloomberg AusBond Composite Index has declined significantly. Figure 6: Bloomberg AusBond Composite Index (by Broad Sector) mv 10% 9% 8% 7% 6% 5% 4% 3% 2% 1% % Credit Supra Semi Govt Treasury Source: Schroders, Bloomberg as at 31 January 2016 Schroder Investment Management Australia Limited 6

7 Allocation Pre GFC, corporate exposure averaged in the low 30% range, in large part because of the lack of government issuance. In the post GFC / declining terms of trade environment corporate exposure has fallen sharply as government issuance as risen significantly. As a consequence being underweight credit more recently effectively removes any material diversification from the portfolio. Figure 7: Schroder Fixed Income Fund Credit Exposure 100% Schroder Fixed Income Fund Credit Exposure (incl/ RMBS) 80% 60% 40% 20% 0% Credit Source: Schroders as at 31 January 2016 As Figure 7 shows, we have been actively reducing our credit exposure (in fact in 2015 our credit exposure was around 20% (including RMBS) reflecting the tight spreads and deteriorating valuations). In the pre-gfc world this would have represented a position that was around 15% underweight the benchmark. However, based on today s benchmark weights it represented an 8% overweight position. The rationale behind this position was that in order to maintain some diversity we in effect maintained a modest overweight, which given the broader rise in spreads through 2015 actually detracted from relative performance. If only life were simple. Reconciling this with our Investment framework The comments above reflect our thinking on some of the structural dynamics of fixed income markets in the current environment. The challenge for us is to reconcile these factors with our broader investment process. Valuations are at the core of our investment process as we believe that over time this ultimately drives both future returns and risk. Value is of course not definitive and dependent on both our core beliefs and assumptions. At the core of our belief set is that idea that over the medium to long term sovereign bond rates should approximate nominal GDP - real economic growth (a function of productivity and population growth), and long run inflation expectations are the key drivers of these forecasts. To this end, there is a significant disconnect between current sovereign yields and what we would expect them to be under normal conditions (see Figure 8). Clearly this is a stretched assumption given the extent to which the major global central banks are re-writing the rule book. Yields are being driven lower (negative in many cases) with the aim of reflating their respective economies and avoiding deflation. Clearly our valuation framework has not been helpful in this environment. Schroder Investment Management Australia Limited 7

8 10yr Bond Yield (%) 10yr Bond Yield (%) Figure 8: Yields v Nominal GDP ITALY UNITED STATES UNITED KINGDOM CANADA FRANCE JAPAN GERMANY AUSTRALIA Potential Nominal GDP (%) SWITZERLAN D ITALY UNITED KINGDOM CANADA FRANCE GERMANY JAPAN NEW ZEALAND UNITED STATES AUSTRALIA SWEDEN NORWAY Current Nominal GDP based on Core CPI (YOY %) Source: Schroders, Bloomberg The second factor we consider are cyclical factors and these have been equally difficult to interpret. Our framework has been broadly constructive on the US economy (albeit not bullish), indicated modest upside pressure on US core inflation, and therefore consistent with a modest Fed tightening cycle. I d argue that the surprise here has been the extent and duration of oil price declines that have seen US oil related investment collapse and headline inflation decline markedly. It s fair to say I have low confidence in where oil goes from here, but even stability in the oil price would start to see US core inflation edge higher. We also believe that some of the concerns about the downside risks to growth that are currently permeating global markets are a bit overdone. What we are now seeing is sharp adjustment downwards in inflation expectations globally and diminishing confidence in the ability of central banks to avoid a stalling global economy and/or to achieve their inflation objectives. The shift to negative rates in Japan (as well as Europe) is a desperate tactic by desperate banks but unhelpful if you re short duration against a lengthening benchmark. To summarise, we continue to think yields are too low, but are seeing increased uncertainty and risk around the macro backdrop. While we believe some of the prevailing pessimism is likely overdone, it does make the near term outlook more uncertain. Moreover, and notwithstanding the uncertainties around the outlook and the broader global policy dynamics, the risk / reward trade-off around Treasury bonds is deteriorating. In other words the downside risks significantly and increasingly outweigh those to the upside. Schroder Investment Management Australia Limited 8

9 Probability of Loss (%) Figure 9: Probability of loss (Major Sovereign Bond Markets) U.S Gov't Bonds Australian Govt Bonds German Government Bonds Japanese Govt bonds Source: Schroders as at 31 January 2016 Conclusion The aim of this article has been to highlight some of the complexity of the fixed income asset class in the current environment. While it would be relatively easy to recognise the uncertainty of prevailing conditions and reduce risk, it is not completely clear what a less risky position is. No doubt for some readers this means the benchmark and while this clearly would reduce our relative risk, it would significantly increase the overall risk of the portfolio. This is a significant point, particularly as our stated objective is first and foremost to deliver defensive outcomes for our clients. In reconciling these considerations against the complexities of the current policy environment, the implications of this for benchmark positioning along with our investment framework, we have taken the more challenging path by attempting to balance both absolute and relative risk considerations. While this has resulted in OK absolute returns and low volatility, we have clearly underperformed the benchmark. Yields may continue to decline, but that said, the prevailing low / negative yield environment does not make duration an attractive proposition. Hopefully, and unlike Malcom Fraser, we get out of this with our trousers still on. Disclaimer: Opinions, estimates and projections in this article constitute the current judgement of the author as of the date of this article. They do not necessarily reflect the opinions of Schroder Investment Management Australia Limited, ABN , AFS Licence ("Schroders") or any member of the Schroders Group and are subject to change without notice. In preparing this document, we have relied upon and assumed, without independent verification, the accuracy and completeness of all information available from public sources or which was otherwise reviewed by us. Schroders does not give any warranty as to the accuracy, reliability or completeness of information which is contained in this article. Except insofar as liability under any statute cannot be excluded, Schroders and its directors, employees, consultants or any company in the Schroders Group do not accept any liability (whether arising in contract, in tort or negligence or otherwise) for any error or omission in this article or for any resulting loss or damage (whether direct, indirect, consequential or otherwise) suffered by the recipient of this article or any other person. This document does not contain, and should not be relied on as containing any investment, accounting, legal or tax advice. Schroder Investment Management Australia Limited 9