November 2021 Update
Making the News
November kicked off with local municipal elections. The ANC’s national vote share dropped to 46%, down sharply from the 53.9% achieved in the previous (2016) local election. From having an outright majority in seven of the eight metros a decade ago, the ANC is now down to only Mangaung (Bloemfontein) and Buffalo City (East London). At 21.8%, the DA’s total vote share was down notably from just shy of 27% in the 2016 election. A large portion of the party’s votes in Johannesburg were picked up by ActionSA (lead by Herman Mashaba) – now the 4th largest party in Gauteng (9.4%). The EFF seemed to reach a ceiling at around 10% of the national vote. Their electoral support may have been capped by very low election participation from young voters.
The Mid-Term Budget Policy Statement (MTBS) was received well by the markets. The Rand strengthened against major currencies (1% -1.5%) and local bond yields declined with a flattening of the curve. Both the expected budget deficit (-6.6% vs -9.9%) and debt-to-GDP (69.9% vs 74.1%) ratios for the fiscal year improved relative to initial projections. The public sector wage bill, SOE’s and increased social protection were highlighted as the key risks to the medium-term fiscal outlook.
The US Federal Reserve announced that bond purchases would be trimmed in increments of $15bn per month, implying the end of the QE program by mid-2022. In line with various other emerging markets central banks, the SARB increased the repo rate to 3.75% (from 3.5%), noting upside risks to inflation due to higher oil prices and a weaker Rand.
Devastatingly, the month concluded with the discovery in SA of a new covid-19 variant which the WHO dubbed ‘Omicron’. Scientists have found that the variant has an unusually large number of mutations, and the fear is that it could be even more transmissible than Delta. New SA cases increased sharply throughout November, with the seven-day rolling average jumping from a low of 106 at the start of the month, to a high of 4700 on the 29th.
November was a volatile month for SA bonds, as yields initially rallied on the back of the MTBPS that showed a commitment to fiscal consolidation, which was well received by the market. Thereafter came the US CPI release in the middle of the month which surprised on the upside, and as a result unsettled global bond markets and emerging market assets. A stronger dollar, together with renewed Eskom load shedding, saw the rand weaken significantly from mid-month levels, and saw SA bond yields give up their earlier gains. The risk off tone culminated in a day of extreme risk aversion on the last Friday of the month with the Omicron variant rattling global markets and causing immediate travel restrictions into South Africa, a devastating blow to the fragile local tourism industry. These developments saw the rand weaken sharply and bond yields kicked 20-25bps higher across the curve, with the ALBI recording -1.2% on the day (yields did however recover slightly into month end).
Overall, the Rand weakened against all major currencies, most notably against the US$ by 5.6% and the All-Bond Index managed to claw back some gains into month end and returned +0.7% for the month; with 1-3 area returning +1.2%, 3-7 area +1.3%, 7-12 area +0.3% and 12+ area +0.6%.
A stronger dollar is typically bad for emerging market equities and November has been no exception given that the MSCI EM Index was down -4.1% in USD (but +1.1% in Rands). Despite SA equities also losing -0.9% dollar value in November, this was cushioned by Rand hedge shares within the Resource (+6.8%) & Industrial (+5.7%) sector listed on the JSE (which as a result ended the month up +4.5% in Rands). Conversely, the financial sector ended the month down -2.6% as most stocks within this sector are more exposed to the local economy. In terms of style investing, given this risk off backdrop, it is no surprise that the FTSE/JSE Value index underperformed the Growth index in November by -3.7%. Despite this underperformance, the value style remains ahead of growth by +21.9% over the last year.
Over the last year, local growth assets (equities and property) remain the best performing asset classes, up +28.5% and +44.3% respectively. The same trend played out for Developed market equities which delivered a return of +26.1% for the year while emerging markets equities fell short and only returned a meagre +6.3%. Longer term returns (i.e., 5 and 10 years) for local, DM and EM equities are now over 10%, which long-term investors can expect to be rewarded with per annum in the foreseeable future. In addition, local bond returns are reasonable at over 8% between the 1 year to 10-year point in time periods and ahead of EM & DM bond peers. Listed property on the other hand still needs to recover from the significant sell-off during March 2020 and as such show negative returns over the last 3 & 5 years and underperforms cash over the 10-year period by -0.9%.
Impact on Our Portfolios
It has been a difficult month for the CWM models given that value underperformed growth as a style. Performance has been divergent across the risk spectrum as the local conservative/defensive models (CWM Income, Defensive & RI-Defensive) performed ahead or in line with peers whereas those models with more equity exposure (i.e., more aggressive) underperformed peers during the month. As mentioned in previous monthly wrap-ups, we do not expect performance to be delivered in a straight line fashion and therefore expect periods where our models would underperform peers. One month’s relative performance would be considered short-term in nature and should not deter investors from their long-term goals. Given the tailwinds expected to favour value investing into the foreseeable future, our models remain positioned with a value bias in order to participate in these stronger expected returns going forward.
As seen above, signifcant alpha was delivered in the last year for the local models with CWM Retirement Growth performing best, ahead of peeers by almost 5%. Since inception returns are reasonably ahead of peers aside from the regular income models that have marginally underperformed due to their allocation to local listed property.
On the global side, the models have underperformed peers during the month as exposure to EM (i.e., Coronation Global EM Flexible) detracted from returns. Underperformance is more pronounced for CWM Global Growth over the month and the last year because of the value style underperforming in November (i.e., Dodge & Cox Worldwide Stock detracting). However the returns for these models since inception, which is now precisely five years, is really pleasing with solid absolute returns in dollars and at the same time all funds out performing their peers by between 0.5 +1.2pa delivering alpha of +1.2%.
Inflation is very topical, with much of the developed world experiencing unsettling price pressures. For example, in October, the US recorded a CPI increase of 6.2% (the highest since November 1990); Germany saw its price level rise by 4.5% year-on-year (the highest since August 1993); and the BOE updated their outlook for inflation to peak at 5% in April 2022 (more than double their 2% target).
While many believe the current inflation spike to be transitory – owing to pandemic-related supply/demand imbalances – arguments for structurally higher inflation in the developed world also hold up. Using the US as an example, the key points in favour of sticky (i.e. persistent) inflation, include:
- The alignment of expansionary Monetary and Fiscal policy: the real US policy rate (i.e. the federal funds rate less inflation) is deeply negative (-6% p.a.) – last seen in the 1970’s. The Federal Budget Deficit as a % of GDP is also the largest post-WW2.
- Wage pressure: health concerns, direct stimulus checks, and a new “work-from-home” culture shifted the employee / employer balance of power. Workers are demanding more from employers, especially in terms of flexibility. Currently, there are more than R10m unfilled jobs in the US economy. Well-known companies like Amazon and others have raised their minimum wage and now offer sign-up bonuses.
- House prices: extremely low interest rates super-charged the demand for Property and resulted in soaring house prices. ‘Shelter costs’ comprise 40% of the US CPI basket and the impact on future rents is not yet reflected in the current inflation prints.
Similar scenarios are playing out in other developed countries. The implication is that – contrary to the last ±4 decades – a period of reflation may be on the cards with dire consequences for developed world bond yields. The graph below highlights the correlation between US 10-year bond yields and inflation:
Locally, since the SARB introduced inflation-targeting 20+ years ago, it has managed to temper expectations from ±6% to ±4.5% p.a. While the (new) figure can be derived from the bond market, it has also been used in practice by large price setters – the most recent example being the Government in its wage-bill negotiations.
Despite the recent 25bps rate hike by the SARB (3.5% to 3.75%), the well-anchored inflation expectations mean that the repo rate does not necessarily return to its 7% pre-pandemic level. The implication for fixed income investing is shown in the graph below: investors will probably have to take on more duration (i.e. longer-dated government bonds) going forward to earn returns previously offered by lower risk instruments (compare 3-Year NCD’s and 10-Year government bonds).
However, the combination of a steeper yield curve and lower inflation expectations has increased the margin of safety offered by our treasury bonds. The real 10-year yield has expanded to 5.6% (compared to c.3% in 2019). Using more than 20 years’ of data, it can be shown that a strong positive relationship exists between current real yields and subsequent 3-year bond returns: the current level (5.6%) implies an expected 3-year forward return of between 9%-12% p.a.
We therefore see two diverging inflation / interest rate paths: in the developed world, a period of reflation with concomitant higher rates; in SA, a lower inflation equilibrium and steeper yield curve. We remain cautious of longer duration developed world assets (including low-yielding, high-growth equities) and positive about local fixed-income. In the latter case, the longer duration required is being handsomely rewarded.
Finally, the scenario below looks at the impact of both negative and positive shifts in the SA yield curve over the next year. Assuming our 10-year government bond trades at a running yield of 9.7% and at par, a 200bps rise in the yield to 11.7% only results in small negative return (-2.3%) – thanks to the protection offered by the high current yield. If yields were to drop by 200bps, an equity-like return of 22.7% is realised (capital appreciation of +12%). In fact, even if yields were to stay the same, the current yield (+9.7%) plus roll yield (+2%) delivers a total return of 11.7%. All-in-all, a very enticing return distribution that is skewed to the upside.