J.P. Morgan Asset Management’s Paul Sweeting, Nick Gartside and Edward Gladwyn examine the impact of pension funds’ drive to corporate bonds
Over the last five years, UK pension fund allocations to corporate bonds have nearly doubled, to a little under 20 per cent. This is despite the fact that buying pressure has caused corporate bond yields and spreads to fall sharply over the same period, driving a reduction in return prospects. While at first this might seem surprising, there are a number of points to bear in mind.
Importantly, there has been a broad change in the risk appetite of pension schemes. Accelerating scheme closures have made the benefits increasingly certain and removed the impact of future pension accrual as a diversifier to investment risk. They have also reduced the appetite of sponsors to spend time dealing with pension scheme risk.
But any move towards de-risking has been against a backdrop of increasingly expensive government bonds. As a result, pension schemes – and other investors – have been keen to find other sources of income. Moving into investment grade credit gives a relatively low-risk way of enhancing yields, and so reducing the cost of the shift.
There exists, then, a question over causality in the relationship between corporate bond yields and the allocation to these assets. Yields might not just have been driving the changes in allocation – the changes in allocation could also have had an impact on yields and spreads. And this raises an important question: has the quest for yield driven prices to unnaturally high levels?
Evidence for a credit bubble – the counterfactual model
One way of looking at the question is to use counterfactual analysis. This involves describing the corporate bond yield curve using a range of macroeconomic factors. These macroeconomic factors are then used to work out what corporate bond yields would have been had the relationship between these factors and the yields remained unchanged. Constructing a counterfactual analysis makes an explicit assumption that the relationships exhibited in the training period should be maintained in the projection period.
In our earlier work on the impact of QE on UK pension schemes, we found that QE had depressed government bond yields significantly, principally during QE1. When applying the same model to corporate bond yields, the results were somewhat different. Not only did corporate bond yields rise above the levels expected at the start of QE1, they did so again at the end. QE2 precipitated a further widening of the difference between actual and expected, whilst the end of QE2 did not result in any significant narrowing of the gap – and at the end of 2012, this gap was 261 basis points.
It is possible that this is just indicative of QE doing what it is supposed to do: driving down yields and costs of borrowing for companies. However, the fact that yields were pushed below their natural levels even when the Bank of England was buying no bonds suggests that other factors are at work. The most likely culprit is the more general quest for yield, and on the basis of this analysis it has resulted in yields being held far below expected levels. But is this the only way in which corporate bonds should be judged?
Can companies cover their coupons?
Counterfactual analysis is only one way of looking at the value of corporate bonds and under our methodology it does not look directly at the sustainability of income. Income cover – and so the ability to maintain bond coupon payments - might be seen as a better measure of creditworthiness from the perspective of pension schemes.
Since the middle of 2011, a decline in income cover has indeed coincided with a tightening of credit spreads, the opposite to what might be expected. However, spreads are still wider than they were for the whole of the period from 2002 to 2007, and income cover was appreciably worse in the early part of that period. This does not suggest that we are in a credit bubble. It does not even suggest that corporate bonds are expensive; rather, it suggests that they are not as cheap as they once were.
A global outlook
The conclusion that corporate bonds are not expensive, just not as cheap as they used to be does not offer much comfort to investors – or much of a clue as to what investors should do if they require additional yield. Perhaps the best advice that can be given is to adopt a broader outlook. Consider, for example, the option-adjusted spread (OAS) on Baa-rated industrial corporate bonds available globally.
Even for a given sector and a given credit rating – so where there should be a degree of consistency between the bonds considered – there is a variation in spread available in each currency. Furthermore, emerging market corporate debt offers consistently higher spreads across the term structure.
Where do we go from here?
This analysis shows that over the short term – that is, the last couple of years – corporate bonds have got more expensive. However, they are not necessarily expensive when longer term relationships are taken into account. But can investors increase yields still further without taking on too much additional risk?
They can if they are prepared to look beyond the index. Specifically, it is worth considering unconstrained benchmarks, which allow investment in a wider range of currencies and credit ratings. The broader the range of opportunities, the greater the yield that might be generated, and the more diversified the portfolio.
Given that pension de-risking looks set to continue, leading to ever higher allocations to credit, it is important that such opportunities are considered – at the moment, these are likely to be the best way to beat the low yields currently available in the sterling corporate bond market.
