US treasuries had a volatile quarter, with the benchmark 10-year yield starting at 2%, then falling to as low as 1.47% at the end of August, as a result of escalating trade concerns. Yields subsequently retraced to 1.79% as the hunt for yield resumed after dovish policy action from the Federal Reserve and other key central banks. The US treasury yield curve remains inverted – meaning that short-term yields are higher than longer-term yields. These inversions are noteworthy because they have historically preceded recessions.

In South Africa, real sovereign credit spreads ticked up during the quarter, with fiscal concerns coming under closer scrutiny. The SA government is facing a double whammy of increased spending to bail out struggling state-owned entities (SOEs) – with Eskom being the largest burden – together with revenue under-collection due to slow economic growth and rising unemployment. This will necessitate increased bond issuance if the shortfall cannot be sufficiently met by expenditure cuts elsewhere.

Despite the prevailing economic malaise, the domestic credit market has proven to be rather robust, with 2019 issuance on track to exceed last year’s R115bn. Auction activity has been dominated by banks and corporates, while SOEs have largely avoided public auctions due to negative investor sentiment around the sector. Demand for credit assets also remains robust. Therefore, credit spreads have continued to tighten, although the rate of compression is slowing down relative to prior years.

During the quarter, we switched out of short-dated Eskom bonds into longer-dated bonds (both are government guaranteed) in order to capture the term premium, as the long bonds had a higher credit spread relative to the benchmark. We took advantage of a mid-August sell-off in bonds by adding a little bit of duration to the Fund. We also added some bank AT1 paper, which is floating rate and provides attractive credit spreads relative to SOEs, corporates and bank senior paper.

The Fund remains conservatively positioned, with duration significantly shorter than the All Bond Index and sufficient liquidity to take advantage of any sell-offs in the bond market that may provide attractive entry points.

Commentary contributed by Londa Nxumalo

North is seldom obvious in the midst of a storm, but the future direction of investment markets feels particularly uncertain. One indication of this uncertainty is that it has become surprisingly easy to provide compelling arguments for diverging bond views.

Global bond bulls argue for lower yields for longer due to deflationary demographic, debt and technology trends. On the other hand, bears point to current abnormalities, such as negative bond yields, a boom in junk-rated corporate credit and overconfident central bankers. South African bond bulls focus on high real yields, stable inflation and excessive pessimism. Just as convincing, South African bond bears have an armoury of ammunition following South Africa’s decade of economic divergence: high government debt, bankrupt and incompetent state-owned enterprises, negative per-capita economic growth and deteriorating intellectual, political and physical capital, to name a few.

Given the right advocate, all of the above arguments make sense. It is, however, unlikely that divergent outcomes will remain equally probable. Instead, reality is likely to deteriorate or improve, with different investment consequences. Unless one has a crystal ball, investors are being asked to ''pick their poison'' and accept the associated pain, should the future turn out different.

Our poison is to position the Fund for the middle path, despite this being the least probable actual outcome. By doing so, we accept that the Fund will likely underperform on a relative basis if the bond bulls turn out to be right, or experience poor absolute returns if the bond bears are correct.

We have chosen this path because we cannot predict future macroeconomic outcomes. Instead, our goal is to improve the probability of reasonable absolute returns across a variety of scenarios. Practically, this means the duration of the Fund is lower than the All Bond Index, the average yield of the Fund is close to the Index yield, and a high percentage of the Fund remains invested in liquid money market instruments and government bonds to allow the Fund to respond to opportunities.

Duration was marginally lowered over the quarter. The most notable development was steepening of the South African bond curve as the market priced in short-term interest rate cuts, while long-term rates bore the brunt of weak fiscal conditions. For some time, we have believed longer duration bonds offer the best relative value. In hindsight, this turned out to be a mistake or too early. The recent curve steepening further reinforces our preference for a combination of long-duration government bonds, short-duration money market, and mid-duration fixed-rate corporate credit.

New appointments
We recently welcomed two new fixed-income portfolio managers: Thalia Petousis and Londa Nxumalo. Thalia has over seven years’ experience in fixed interest and joined Allan Gray in 2015 as a fixed-interest trader. She will co-manage the Allan Gray Money Market Fund. Londa also has seven years’ investment experience and joined Allan Gray in 2017. She will co-manage the Allan Gray Bond Fund. Both managers have proven to be invaluable members of the investment team, and their expanded roles will allow our clients to benefit further from their expertise.

Commentary contributed by Mark Dunley-Owen

The current talking point in fixed income markets is the inversion of the US yield curve. This refers to the yield differential between long- and short-dated US government debt securities, for example, the 10-year bond and three-month Treasury bill. While the US economy is improving, the longer bond typically offers a higher yield than the short Treasury bill bond to entice investors to move out of riskier assets, such as equities, and extend fixed income duration. As the economy peaks, the US Federal Reserve raises short-term interest rates to control inflation, while investors move into longer-duration bonds, which typically outperform other assets in the down cycle. The yield curve differential goes to zero or negative, in which case the yield curve is said to be inverted.

This is important because an inverted yield curve has preceded every US recession since the mid-1950s. A simplistic explanation for why this happens is that the US fixed income market is the deepest investment market in the world, making it as good an economic predictor as one can get. In the current cycle, the yield differential between the US 10-year bond and three-month Treasury bill has declined from positive 380 basis points in mid-2009 to negative five basis points in March 2019. The conclusion being drawn is that if history repeats, a US recession is imminent.

Yield curve inversion is a popular topic of conversation because of its past predictive accuracy and simplicity. However, it ignores other factors which impact bond yields, the most important of which is the increased role played by Central Banks in manipulating interest rates. It is also unusual that something that everyone is talking about turns out to meet expectations; if this were the case, investing would be easier than it is. We are thus reluctant to assume a global recession based on a single data point. A more interesting question is what would happen to South African bonds if a global recession did occur? South Africa typically tracks the world; therefore, the country’s already weak economic growth should be expected to deteriorate further. This would be negative for South African risk assets, and the prices of equities would likely fall. It is, however, not clear what would happen to the prices of South African bonds as investors move out of riskier countries, such as South Africa, and move into lower risk assets such as bonds. Contrary to expectations, a global recession may be positive for SA bonds if falling US bond yields drag SA bond yields lower, and prices higher.

One should also remember that the SA economy has diverged from the global economy over recent years as negative domestic events have overwhelmed positive global trends. SA bond prices thus already reflect a concoction of local risks, in contrast to the prices of developed market bonds, which reflect the near-perfect bond conditions of recent times. It is probable that South African assets would outperform global assets if reality turns out to be not as bad as feared in South Africa, and not as good as hoped in the developed world. The Fund’s positioning remains consistent with recent quarters, offering above-inflation yield with some downside protection from below benchmark duration. Changes were made to the portfolio when opportunities presented themselves, including a small reduction in Eskom and increase in Land Bank and FirstRand exposure.

Commentary contributed by Mark Dunley-Owen

Eleven percent of the Allan Gray Bond Fund is invested in Eskom bonds. Eskom is technically bankrupt, and appears unlikely to be able to pay the interest, let alone the capital, on its R420bn gross debt. All of the Fund’s Eskom holdings are however government guaranteed, meaning the South African government guarantees to take over the debt obligations if Eskom is unable to do so.

Eskom bonds yield approximately 1.25% higher than comparable government bonds, an attractive yield pick-up if one believes that the credit risk of government guaranteed Eskom bonds is equivalent to the credit risk on government bonds. In other words, the government will honour the guarantee. We believe this is likely and are comfortable holding Eskom bonds despite the company’s perilous financial condition.

The technical implications of the government not honouring its guarantee are not clear, but it is probable that the rating agencies would consider this a sovereign default and South Africa would be downgraded multiple notches below its current rating. The South African government has defaulted before, but not since the abolition of apartheid or since being upgraded to investment grade. The practical implications of doing so would be severe. South African risk would re-price materially higher, meaning the rand would weaken, share prices of South African companies would fall, and bond yields would rise. The already high cost of servicing the government’s debt would rise further, fewer investors would be willing to fund the country’s twin fiscal and current account deficits, and economic growth would fall. South Africa cannot afford this scenario.

Our view remains that South African bonds offer attractive yield with limited risk of a large price move either up or down. The Fund maintained duration higher than its historical average, and average yield in line with the benchmark JSE All Bond Index. There was little activity over the quarter, illustrated by both the Fund size and modified duration increasing marginally. Liquidity was invested in long duration government bonds