We believe it makes sense to manage euro fixed income portfolios on an aggregated basis, i.e., with all the sectors in the same portfolio. This is because the Eurozone has experienced periods of systemic stress, which have affected all fixed income assets, and may well happen again. In such periods, when the correlation between asset returns either moves towards +1 or -1, it makes sense for one portfolio manager to have an overview of the whole portfolio, to make sure different sectors are not all overweight the same risk factor, and also to optimise off-setting hedge positions.

Recent market developments bolster the case

Roughly 18 months’ ago we published a paper arguing that fixed income assets in general, and in the Eurozone in particular, are often better managed on an aggregate basis because, in a systemic risk environment, a single portfolio manager is better placed to manage the portfolio’s overall risk and also take advantage of relative value opportunities between different fixed income sectors. The impact of the corona virus on financial markets this year, in particular between March and May, reinforces this view.

In particular, nearly all fixed income assets experienced unusually high levels of stress and illiquidity during March. The sudden shift in expectations to a severe economic slowdown had a predictably negative impact on assets like high yield and emerging market credit, but even some very high quality assets, like triple-A rated Australian and New Zealand government bonds, performed poorly at times because of liquidity and forced liquidation issues. A portfolio manager monitoring the details of all these risks together would have been able to control them better than an asset allocator who had delegated portfolio management responsibility to independent managers and didn’t have visibility of all the risks in their portfolios.

Equally important, while nearly all euro credit spreads have tightened considerably from the widest levels seen earlier this year, the degree and timing of the rebounds has varied across asset class. For example, euro sovereign spreads were the first to tighten in mid-March, in response to the considerable support the ECB offered through its enlarged QE programmes, but this rally then dissipated as concerns rose about the adequacy and funding of the euro area fiscal response. By early April, however, euro corporate credit spreads had started to tighten, with investment grade corporate credit in particular benefiting from the ECB’s expanded purchase programme. However, high yield credit also managed to perform very strongly in April as the scale of government support programmes for distressed corporates became more evident. By May, euro sovereign spreads started to tighten as the proposed EU Recovery Fund could become a vehicle for fiscal transfers between weak and strong countries. These shifts in relative performance took place within a few weeks (see Display 1). An investor managing a portfolio on a disaggregated basis would have struggled to take advantage of them, but doing so in an aggregated portfolio would have been more feasible.

Why the trend to disaggregation makes sense...
The trend in fixed income mandates in recent years has been towards management disaggregation. Rather than instruct a portfolio manager to manage a mixed portfolio of assets (credit, securitised, government bonds, etc.), end investors have allocated individual mandates to single sector specialists, e.g. appoint a government bond fund manager to manage the sovereign portion of the portfolio, an emerging market specialist to manage the EM component of the portfolio, etc. The logic for this is clear: the end investor can access best-in-class investment skill in each asset class. He/she can also optimise the diversification benefits by setting an asset allocation across sectors which best matches the investment goals, rather than using the market cap weighting of an aggregate index.

...and why it doesn’t (in fixed income)
However, we believe there is a compelling argument for aggregation, rather than disaggregation, when it comes to managing fixed income assets, especially in the Eurozone. This is because aggregation provides a better way to manage the risks stemming from the systemic crises which have periodically affected fixed income markets.

The nature of a systemic crisis is that it affects the performance of all assets, sometimes in ways which investors would not have previously expected. To make things more difficult, crises have varied in their length, nature and intensity, depending on whether the problem stemmed from a global financial crisis (e.g., 2008), a domestic banking crisis which contaminated the sovereign and rest of the economy (e.g., Ireland and Spain in 2010), a political crisis which affected the sovereign and the banking system, as well as other countries (e.g., Greece in 2010 & 2015), or, most recently, an exogenous economic shock emanating from a public health crisis.

Systemic risk events argue for aggregate portfolios
In such an environment, we believe it is advantageous to have a single fund manager who is aware of all the risks in a portfolio, and is better placed to manage the overall risk holistically. A particular danger of disaggregation is that fund managers, operating independently of each other, may implement similar and correlated risk positions in their portfolios, which amplify overall portfolio risk. Similarly, working independently of each other, they are unlikely to maximise the internal hedges which exist within a multi asset fixed income portfolio. Judged primarily on their own individual performance, they have limited incentive to help the overall aggregated portfolio achieve better risk adjusted returns.

For example, imagine a situation in which the macro fundamentals for country A were improving. In a disaggregated portfolio, the managers of the separate corporate credit, sovereign and securitised sleeves may all look to overweight country A. However, collectively this might lead to an imprudently large exposure. The manager of an aggregate fund would be better placed to judge if the overall exposure was appropriate, as well as the best way to gain that exposure, e.g., through financials or the sovereign debt.

In addition, he may be able to engineer a better expected return, on a risk adjusted basis, by owning a larger allocation of low risk government bonds to offset the high- yielding risky positions. He may not expect the government bonds to do very well, but thinks they are worth owning as insurance in case things turn out worse than expected and the risky positions do badly. In a disaggregated portfolio, individual managers are not aware, nor incentivised, to hedge each other’s risk positions in this way. 

Disaggregation may also slow the optimal re-allocation of assets between asset classes as opportunities and threats emerge. It can be time consuming to redeem funds, re-allocate them and redesign mandates if they turned out to be inappropriate.

Many euro fixed income assets have experienced several unanticipated periods of stress over the last 20 years, and may not have performed as anticipated. For example, the usual benchmark for euro sovereign portfolios prior to 2010 was a market cap weighted index of euro sovereign bonds. This was generally expected to be the “safe haven” component of a portfolio, delivering modest returns but performing well during periods of economic and financial stress. Few investors foresaw that, from 2010 to 2013, one issuer making up around 10% of the index (i.e., Greece) would default, others (Portugal and Cyprus) would lose their investment grade rating, some of the initially AAA rated sovereigns (Spain and Ireland) would require bail-outs and the largest member of the index (Italy) would become the most volatile euro fixed income asset.

Let’s explain and illustrate these points with reference to the historical data for euro fixed income assets since the beginning of monetary union in 1999. Whilst past returns may not be a good reflection of future returns, they do highlight some of the risks investors have experienced in the past and may face in the future.

A brief history of euro fixed income markets
Display 2 shows the OAS spreads for the major sectors of the Bloomberg Barclays Euro Aggregate index, i.e., Euro Treasuries, Govt Related, Securitised and Corporate credit. It provides us with a history of the euro fixed income markets and the relative performance of the main sectors. In particular, the period from 1999 until 2007 was characterised by very tight and stable credit spreads in most products, the only exception being when some corporate sectors (e.g. telecoms) came under pressure following the dotcom bust.

The first major crisis started in 2008 as the global financial crisis caused distress in euro credit markets, in particular for financials. Significant stress was also experienced in securitised markets, given their connection to the financial sector, forcing the ECB to launch the first covered bond purchase programme. Sovereign debt in countries with impaired financial sectors (Belgium, Holland, Austria) also cheapened as investors saw the financial sector’s problems as contingent liabilities of the state.

Credit spreads tightened sharply in 2009 as the global economy recovered and the crisis faded, but then widened out again from 2010 until 2013. This time the problem was more to do with the sovereign sector, in particular Greece, although in other countries (e.g., Spain and Ireland) the problem started in the banking sector and spread to the sovereign. As stresses built up within the economy and financial system, other countries (e.g., Italy, Portugal, and to a lesser degree France) were also pulled into the crisis as it exposed their vulnerabilities. Issuers deemed as safe havens (primarily the German sovereign) experienced additional demand.

A combination of central bank emergency measures and fiscal support helped end the crisis, leading to a period of sustained spread compression from 2013 onwards, apart from some corporate credit spread widening in 2016 and 2018 (due to economic growth concerns), and wider sovereign spreads in 2018 (due to heightened political risks in Italy).

We learn several things from this history:

1)      Crises affect all sectors: a systemic crisis, by definition, affects everything in the system. It’s unsurprising that the OAS spread for the Euro Aggregate Index is highly correlated with the ECB’s Systemic Stress Indicator (see Display 3). The severity may vary depending on how severe the crisis is, but finding good hiding places can be hard during these periods.

Even sectors which one would have thought have few fundamental links to each other turn out to be highly correlated. For example the credit spreads of government owned or sponsored agencies, which generally have very high and stable credit ratings versus sovereigns, are very highly correlated to corporate credit spreads (see Display 4).

2)      No two crises are the same: depending on why the crisis started, and also because investors and regulators respond to past crises, asset performance varies from crisis to crisis, e.g., 2008 was primarily about financials, while 2010-12 was about sovereigns, especially the “periphery”. This means the “beta”, i.e., the relative volatility of different assets, varies over time. The correlation between assets also varies over time, both depending on how each crisis builds and recedes and which assets it affects most.

3)      Divergences in performance provide diversification opportunities: the silver lining to a systemic crisis is that safe haven assets can provide protection to a range of risky assets. In particular, AAA rated sovereigns like Germany have performed very well in periods of stress. While the correlation between “risky” sectors in a crisis has risen, the correlation with Bunds has become more negative (see Display 5), providing a portfolio diversification opportunity, at least for those who have been able to invest in both high quality government bonds and spread product.

4)      Not all assets perform as expected: the 2008 crisis is remembered for how many structured credit products had far higher default risk than their credit rating would have suggested. In the 2010-12 euro sovereign crisis, investors discovered large portions of their sovereign bond portfolios, which were previously thought to have little or no default risk, were actually very risky, both in terms of credit spread volatility and actual default (with Greece restructuring its debt in 2012). While the Treasury OAS spread is far less volatile than other spreads, this hides significant variation within the sovereign universe (see Display 6). Italian Treasury OAS spreads have had periods of significantly wider and more volatile performance than any other sector of the index.

Why we fear further systemic crises
Unfortunately, we believe that the Eurozone will face further periods of systemic stress. Many of the structural and fiscal imbalances which have caused problems in the past have not been fully resolved, and while backstop and support mechanisms are better than they have been in the past, they are not strong enough to make investors feel entirely confident. Highly indebted sovereigns with low trend growth economies, most notably Italy, remain a key source of concern for investors, especially given the Greek experience made it clear that euro sovereign debt is not default-risk free. Another source of concern is the continued link between sovereigns and the banking sector, a potential source of contagion which runs in both directions. While the EU has worked hard to reduce risk in the financial sector, more substantial measures, like a Eurozone-wide deposit guarantee system, have yet to be achieved.

These lessons lead us to believe that it is better for euro fixed income portfolios to be actively managed on a holistic basis, i.e., with the potential performance and correlation of all assets taken into account.