Fixed Income Insight: navigating your bond portfolio through a sovereign downgrade

01 April 2016

South African investors typically gain exposure to South African bonds via bond funds that are typically benchmarked to the ALBI (the All Bond Index).

The ALBI is a bond index comprised predominantly of South African government bonds and bonds issued by Eskom. The poster child for South Africa’s uncertain economic outlook is the potential for a downgrade of South Africa’s sovereign credit rating to junk status by one of S&P or Moody’s. This is in turn is fuelling investor concern about the value of their bond holdings. Some investors are evaluating strategies for hedging their bond portfolios against a loss in value from a sovereign downgrade.

In our view, South African bonds have already experienced a sell-off and so much of the bad news from a sovereign downgrade is potentially already priced in. In any event, it is not clear to us that it is possible to delineate changes in an emerging market sovereign bond’s yield between changes due to that sovereign’s credit rating and its macroeconomic fundamentals versus those due to global risk aversion and liquidity factors.

In our view, whilst investors are expressing their concerns about a bond market sell-off by linking this to a sovereign downgrade, their concern is really one of a sharp and unexpected increase in SA bond yields from current levels. We do not rule out such an event, especially given the current, uncertain environment.

We outline a range of options available to investors wishing to hedge their bond portfolios against such an outcome. Our suggestions range from reducing the duration of their bond portfolios by switching all (or a portion) of their bond portfolio into a cash benchmarked portfolio to using options to protect against a fall in markets. We think that there is also merit in reducing exposure to bonds issued by state-owned enterprises as their credit spreads are likely to widen due to the close linkages between their financial position and implied support from the State.

Is a sovereign downgrade likely and when is this likely to take place?

Our current view is that a sovereign downgrade to junk status¹ is more likely than not, with the timing of a downgrade likely to be December 2016 or early next year. The medium-term budget policy statement scheduled for October is a key event that will influence the timing of such a downgrade and will dictate whether the downgrade will come as soon as December. We note that events of the past week have further clouded the outlook and a downgrade in June cannot be ruled out although we think that Pravin Gordhan’s efforts will stave off a downgrade in June, accepting that this is by no means certain. We will cover the outlook for the sovereign credit rating in a separate note to be released in the coming weeks. In this note our focus is on potential strategies for hedging an unexpected spike up in bond yields beyond what the market is currently pricing.

What is the likely impact of a sovereign downgrade on SA² bond yields?

We draw a distinction between the occurrence of a downgrade and the start of the bond market beginning to price a downgrade. Our view is that SA government bond yields have already begun to price a downgrade of the South African sovereign and that this is reflected in the more recent uptick in yields on government bonds. It is therefore not necessarily the case that sovereign bond yields will increase significantly from current levels (the R186 benchmark bond was yielding 9.25% at the time of writing) in response to an actual downgrade. That said, in line with the historical behaviour of bond yields on other emerging market bonds around the time of the downgrade, we would expect that bond yields would start to tick higher in the days prior to an actual downgrade and continue their rise following the announcement of the downgrade. To place this expected uptick into context our view continues to be that, after allowing for a possible December downgrade, that bond yields (as measured by the benchmark R186) will end the year between 9.25 and 9.5% p.a. which is broadly in line with market expectations.

In addition, the weight of empirical evidence³ suggests that more than half of the change in emerging market sovereign bond yields is driven by global risk and liquidity factors rather than by changes in a sovereign’s credit rating or macroeconomic fundamentals. It is therefore more likely to be the case that yields will react based on global risk factors at the time of the downgrade rather than the downgrade itself. For example, in the event that a downgrade occurs in an environment that is supportive of investment flows into emerging markets, then any adverse impact from the downgrade could be offset in the months following a downgrade by investor appetite for higher yielding emerging market bonds.

So, do South African investors need to be concerned about the value of their bond portfolios?
One of the reasons that investors in ALBI benchmarked bond funds have become increasingly concerned with a sovereign downgrade relates to the lengthening in the duration of the ALBI. As the graph below shows, the duration of the ALBI has steadily increased over the last few years as National Treasury has increasingly sought to term out their borrowing through the bond market as a way of mitigating their exposure to a higher interest rate environment. 



This strategy could be good for the country as it reduces the likelihood that the country will be a forced borrower at untenably high borrowing costs.

However, this is also a zero sum game, and so the corollary is that investors in SA bonds could be lending money to the sovereign at interest rates that are too low. Put differently, it may be possible that investors could demand greater compensation for lending money to South Africa in the future. In this case it would be better for current lenders to SA to lend money for shorter terms.  In this way, they would have an opportunity to lend at higher interest rates in the event that South Africa’s borrowing costs increases in the future.

The ALBI currently has a duration of 6.9 years. This means that in the event that bond yields rise by 1%, a R10m investment in the ALBI would fall by approximately R690,000 to R9,31m.

What strategies are available to investors to protect against a sharp rise in bond yields?

For those investors who consider that the risk of a sovereign downgrade is not fully priced into the bond market, or that a sudden and unexpected rise in yields is likely, then an effective hedging strategy could be formulated and would depend on the precise nature of the bond exposure. For the purpose of this note we assume that the current bond exposure is through the ALBI, which in turn is a mixture of government and bonds issued by the State Owned Enterprise (SOE), Eskom.

In this case, the main hedging strategies would include:

i) Buying insurance against a sovereign downgrade
Credit default swaps (CDS) are a form of insurance against a bond issuer defaulting on their obligations. Many investors who are physical holders of a bond will often turn to this market to hedge themselves against a deterioration in the credit worthiness of the underlying borrower. Such contracts are also available to protect against a deterioration in the credit worthiness of a sovereign borrower, including South Africa. The fact that this market trades daily means that there is a readily available market price for the cost of obtaining such protection. In the event that there are concerns about the credit quality of the sovereign then the costs of obtaining such insurance as indicated by the CDS premium or, in market parlance, the “spread”, would increase.

So, how could we use this market?  Well, if the CDS market was not pricing in a downgrade we would enter into a CDS contract to buy protection against a future deterioration in the credit worthiness of the SA sovereign. When this deterioration did occur and the CDS market began to price it, the cost of obtaining such protection would increase. Our previously purchased contracts would become more valuable and we could on-sell it at a profit. The profit would then help offset some (or all) of the losses on our physical holdings of South African government bonds.

However, as can be seen from the graph below, which shows the CDS spread for South Africa, an increase in the cost of obtaining such protection already occurred over the second half of last year, with a sudden spike over December, as concerns over South Africa’s credit worthiness began to ratchet up. 



In a phrase: the proverbial horse has bolted; in our view, the CDS market is already pricing a sovereign downgrade and therefore we do not see an opportunity at the current time to use this market to purchase protection against a downgrade.

ii) Reduce interest rate duration

Reducing the duration of the ALBI portfolio to, say, zero, would almost entirely mitigate the risk of a further rise in yields and result in no loss of value in the event that bond yields rose further from today’s levels. The converse to this is that if bond yields fell, the value of the investment would not rise. In this position, the investor is protected from a rise in bond yields but would also forego any gain from a fall in yields.

With today’s 25 basis point rate hike cash becomes a bit more attractive from a yield perspective. That said, investors are best placed discussing the precise strategy with their bond manager before deciding on a course of action, since the bond curve is still upward sloping so that shifting into shorter duration instruments (the extreme being cash) may be penal if the anticipated sell off does not materialise.

iii) Use options to protect against a market sell-off
Much like options can be used to protect against a fall in equity markets, they can also be used to protect against a bond market sell-off. 

As an example of a protection strategy to protect against a sudden rise in yields over the next one year, we can use the R186 to illustrate the type of protection available to investors using options to hedge their bond portfolios.

Example: an investor in the R186 (current yield 9.25%) could be protected from the yield on the R186 rising  above 10.25% in exchange for foregoing any increase in the value of their bond portfolio should the yield on the R186 fall below 8.93%. Clearly the investor would suffer some fall in value from yields rising (between 9.25 and 10.25%) but such a strategy would provide protection should yields continue to rise above that.

Options can be expensive, however, we think they have a place in the risk management toolkit provided the fund manager can properly price, execute and manage these strategies on an ongoing basis. Subject to any regulatory constraints, clients should consider including them in the list of instruments they permit their fund managers to use.

iv) Reducing SOE exposure versus the benchmark
In our view credit spreads on the bonds of State Owned Enterprises (SOE’s) will likely reset higher as the listed credit market begins to price the likelihood of a downgrade. Unlike with the government bond market, both absolute yields and credit spreads in the listed credit market tend to react much slower and so these spreads, in our view, could widen out in the coming months. So, as this market begins to reprice we believe that it will become more costly for SOE’s to borrow. This will mean that their credit spreads will widen.

Eskom bonds account for approximately 7% of the ALBI index. Being “short” these bonds relative to their benchmark weight will lead to outperformance vs. the index.

This is the one part of the local listed bond market where we believe that a sovereign downgrade may not yet be fully priced in and we foresee opportunities to achieve a measure of protection.

v) Increase exposure to inflation-linked bonds
Inflation-linked bonds tend to be mostly held by domestic investors, much more so than nominal bonds.  In a sovereign bond sell-off, especially one led by foreign investors, nominal bonds tend to react much more viciously than their inflation-linked counterparts. Any sell-off in inflation-linked bonds is also generally met with buying support above a real yield of 2% p.a.

The graphs below illustrate this point. The first graph below shows that the yield on the nominal, R186 (10-year) bond rose by more than 1.5% in December last year compared to a mere 0.25% rise in the yield of the I2025 (an inflation-linked bond maturing at a similar time) and as shown in the second graph below.






It may make sense for bond investors to consider switching a portion of their holdings in nominal bonds into inflation-linked bonds on a duration neutral basis.

Our current view is that a sovereign downgrade to junk status is more likely than not, with the timing of a downgrade likely to be December 2016 or early next year.

Are there any other measures I could take to protect my bond portfolio against the prevailing market uncertainty?

Given our view that bond investors are not necessarily specifically concerned with the risk of a sovereign downgrade but rather the risk of a sharp rise in bond yields from an unexpected event such as the events of last December then, in our opinion, investors should also consider strategies that provide protection at an overall portfolio level. For example, it may be as effective to consider hedges that could act as a proxy hedge for a decline in the value of a bond portfolio. These may include:

• Hedging exposure to the Rand by looking to further increase exposure to foreign currency assets within their specific regulatory constraints.
  
  
• Hedge or selectively sell out of those equities that are most susceptible to a sovereign downgrade – ideas here include hedging /selling down exposure to Big 4 SA banks and Telkom.

  Big 4 SA banks may experience a sell-off in their share price as the outlook for economy deteriorates even further.

• Increase exposure to Rand-hedged stocks
  Many of our locally listed SA companies have offshore earnings that act as a natural hedge. Should a sovereign downgrade translate into an even weaker currency, these stocks could act a natural hedge.  

Conclusion
Local bond investors face an increasingly uncertain outlook. Whilst it may be too late to protect against a sovereign downgrade, the risk of sharp rise in bond yields may still materialise given the heightened economic and political uncertainties. The gradual lengthening in duration of many bond portfolios due to the relatively larger issuance of longer-dated government bonds means that, with the possible exception of those investors matching long-dated, guaranteed liabilities, a rise in bond yields may be even more detrimental than was previously the case. Various strategies exist to provide protection against an unexpected bond market sell-off including reducing the duration of the bond portfolio and rethinking allocations to credit within the actual bond holdings. 

________________________________

1 It is important to draw a distinction between South Africa’s foreign currency rating and its local currency rating. A downgrade to sub-investment grade by S&P, and possibly Fitch, relates to the foreign currency rating. The local currency ratings are still multiple notches above sub-investment grade and we do not see these deteriorating to sub investment grade in the near term.

2 We do not cover the impact on offshore bonds although we note that a) empirical evidence would again suggest that yields on offshore hard currency SA sovereign bonds are more likely to influenced by global macroeconomic risk factors and liquidity and less influenced by the specific macroeconomic fundamentals of the sovereign and b) the hard currency bonds are included in many emerging market bond indices against which many global fund managers are benchmarked. A single notch downgrade increases by any one of the agencies to junk status increases the likelihood that SA hard currency bonds fall out of that index and so some selling is likely to occur for technical reasons. This will lead to a spike in yields. Whether those yields settle back to previous levels is likely to be more correlated with global fundamentals.   

3 Local Currency Sovereign Risk by Wenxin Du (Federal Reserve Board) and Jesse Schreger (Harvard University) 10 December 2013