Does the rand-dollar PPP actually tell us anything?

Views on the rand seem to be inherently controversial. This is not only because forming a view on the currency’s direction inevitably seems to be a fool’s errand, but it is a topic that South Africans find emotive.

So, when Methodical Investment Management’s fixed income portfolio manager Jean-Pierre du Plessis posted a rand-dollar purchasing power parity (PPP) chart on social media last week, he must have known he was courting a response.

Du Plessis’ chart showed that, for the first time since 2013, the rand was below its PPP level. This he calculated as R14.16 at the end of April.

However, calculating the ‘fair value’ of the rand is not an exact science. As PSG Wealth portfolio manager and strategist Schalk Louw pointed out in response, the starting point you chose for your calculation matters.

Different periods

Essentially, a PPP calculation is showing an inflation differential. In the case of the rand-dollar exchange rate, that would be inflation in South Africa minus inflation in the US. All else being equal, the rand should trend on that line.

But countries go through very different periods and inflation regimes. As a result, where you start the calculation could make a significant difference to the outcome.

“As an example, I used the last time the rand was trading at R1 to the dollar as a starting point, because then, in theory, your inflation differential should have been more or less the same,” Louw said in an interview with Citywire South Africa. “Using actual inflation differentials from that point on to date, and actual rand movement, you could a very different picture.”

On this view, the rand is still well above ‘fair value’, and in fact has never traded meaningfully below it.

Louw acknowledged that this isn’t an entirely serious calculation. In the 1980s, South Africa was under sanctions and exchange controls were in place. The value of the rand was not therefore being set in an open market.

However, it still raises the question of the usefulness of PPP as a measure. Louw’s preference is to use the below graph, showing the rand relative to PPP.

“I use this as a guideline,” Louw said. “It’s just to show us those really overly-negative or overly-positive environments where the rand really overshoots. We saw it in 1986 with the Rubicon speech, in 2001 when rand was manipulated, in 2016 when Zuma played with his ministers of finance, and we saw it when the rand blew out last year after the downgrade to junk.”

Rolling start date

Methodical’s Du Plessis takes a different approach. His preferred PPP calculation is a rolling 10-year geometrically weighted calculation, which is what is reflected in the first chart above.

“This gives you a weighted start time,” Du Plessis said. “I’m by no means saying that makes it 100% accurate, but I think it gives a good sense of whether the rand is historically cheap or expensive.”

He added that using a rolling 10-year geometrically-weighted average for inflation provides a better picture, in his view.

“If you use the differential over a very long term, the risk is that you end up with something that loses its relevance. I’m not suggesting that you are going to day-trade the rand versus its PPP, but you do want something that is relative to something that exists today, rather than something that was the case 50 years ago.

“That’s why I think it’s a better model to have a rolling start date. Whatever start date you have with PPP is going to be a problem because of the way it works. But if you have a rolling start date, and weight that towards the most recent past, then you do end up with something that is more anchored.”

Necessary, but insufficient

With these varying takes on PPP, then, how much does it really tell investors?

John Cairns, head of research at Rand Merchant Bank (RMB), said that the consensus from economists is that “PPP probably holds in the long-term”.

“The theory is a strong one, but the evidence is very mixed,” he added. “In South Africa’s case, I would argue that PPP is a necessary but insufficient explainer of the rand in the long-term.”

This is because, over time, the rand has weakened more than just inflation differentials would suggest.

“This you might think could be explained by South Africa’s low growth – the Balassa-Samuelson effect. But, in practice, the rand’s underperformance has not matched well with the ups and downs on the country’s long-term growth prospects.”

Cairns added that RMB runs various fair value estimates for the currency, with the chart below being one he finds particularly useful.

In this case, they have used the real effective exchange rates of both the dollar and the rand (the trade weighted exchange rate adjusted for inflation) and made a further adjustment for the rand’s tendency to depreciate faster than inflation. This model puts the PPP level on USD/ZAR in May 2021 at around 14.65. Like Du Plessis’s chart, it therefore shows the currency having now slipped into the expensive side of fair value.

Cairns added that RMB uses this PPP model as a useful tool to understand the valuation of the rand, although their forecasts are built on more complex behavioural models – which, he said, “go by the favourably termed behavioural effective exchange rate, or Beer”.

Patrick Cairns is South Africa editor at Citywire, which provides insight and information for professional investors globally.

This article was first published on Citywire South Africa here, and republished with permission.