It will be interesting to see if the finance minister blames apartheid for the fiscal trouble that the medium-term budget policy statement on November 1 will have to clean up. Thankfully, the Treasury holds itself to a higher credibility bar than others, though. Instead, we are likely to get some more sober analysis and action.
The origins of this mess are interesting, because there are actually no surprises here. The wage bill settlement was marginally below what we thought it would be at the time of the budget and didn’t really come in any different from what the bond market was expecting. There have been no upside surprises in SOE bailouts and broader spending on a monthly basis is about where we thought it would be in the most recent data. Revenue similarly was expected by the bond market to be well below where the Treasury had pencilled it in at the budget.
So, what is the issue?
The revenue shock has been larger than expected. While in February we had pencilled in about R32bn of revenue undershoot (and consensus was about in the same area), the large seasonal tax take numbers recently have shown it is more likely a R65bn underrun.
On top of that global monetary conditions have tightened further and faster, as have domestic conditions. Credit growth has slowed, foreign investors have increasingly shunned local government bonds and local investors demand more compensation and higher real yields. Bond market liquidity — which hadn’t recovered since Covid-19 — has also deteriorated.
The market also didn’t entirely put two and two together in February and was still just about holding onto Treasury reassurance. But there are also the risks of a possible increase in issuance that is more alarming still.
Higher debt service costs are becoming apparent too, while SOE tail risk looks more likely to crystallise — particularly as the Transnet carnage becomes increasingly apparent.
All this is happening arguably with growth expectations now higher than in February and shifting higher still. The economy is more robust to load-shedding than expected in February and is getting more robust as more rooftop PVs are being installed in the short term. The estimates of Eskom and that of us are at about 4.7GW at end-July and a total of 3GW installed this year and next.
This is stark and important. Revenues have been this weak (stripping out prior tax year timing effects) despite much better growth and better resilience to load-shedding. As such load-shedding cannot be wholly blamed for the revenue shock.
The Treasury has now instructed departments to make huge cuts to spending this year and in future years to squeeze in the wage bill settlement, which has caused an outcry. Yet the Treasury was clear and transparent that it was not factoring in anything more for the wage bill in February.
The usual cries go up that this is awful and suggest the evil Treasury is taking great delight in swinging the axe. Yet cabinet approved this move and there is little other option. What seems to be forgotten — just six months from an election — is that the ANC and cabinet are actually remarkably fiscally conservative. Yes, they may not make the right structural decisions and put off making difficult choices, but they will not press the “blow the fiscus up” button and instead leave Treasury with no choice but to offer up “salami slicing” cuts.
There is an alternative prescription to fiscal laxatives, but it would seem to ignore all the evidence of the narrowness of the tax base, the depths within the personal income spectrum, the behavioural effects of wealth, financial flows or how international debt markets and currency markets work.
A mixture of laziness and misanthropy sometimes catches me off guard. I am tempted to say “let’s follow the prescription” and hike taxes massively and hike bond issuance just to see what would happen.
Yet logic rules. The National Treasury will present on November 1 that there is marginal room to issue more debt and, assuming they are politically blocked from VAT hikes, they will need to cut expenditure.