Yesterday, StatsSA released the 2016 Second Quarter GDP numbers. Here’s the main flashcard from the media presentation:
But here are the other two that I think are just as important:
The problem is: we’re not good with geometric math. We like arithmetic math. So when we see a 3.3% increase in quarterly GDP, seasonally adjusted and annualised, what we don’t think is:
- “Oh, that’s only a real-life improvement of 0.0081 since the first three months of year.” or
- “After that 1.2% annualised contraction in the first quarter, this year has only seen a 0.49% increase since December.”
I’m not saying that this is a conspiracy or anything like that – this is just the way that GDP is measured. From a statistical standpoint, what’s important is that the measurement is consistent and comparable. And at some point in time, we decided that growth was more than just an additive process: like interest, it’s meant to compound. Growth begets greater growth, etc. So our measurement of GDP is geometric. But it also means that the number can be misleading if you’re inclined to ‘over-read’ it – and especially in circumstances where growth appears less…geometric.
Because what happens when growth slows for a quarter, and then picks up for a quarter? Well, that’s an awkward anomaly in the growth begets growth dynamic – because here’s what the measurement assumes:
- When growth is negative, less growth begets even less growth, compounded. And
- If there’s a recovery, you’re back to more growth begetting even greater growth for the next year (even though that’s not what a quarterly recovery actually implies).
So when you look at that 3.3% figure, the important question to ask is: why?
If the answer is “Oh, because we’re a growing country” – then that 3.3% makes some sense. And it should be roughly in line with the 6-month on 6-month annualised figures, and the year-on-year figures.
But if your answer is “Oh, because there was a recovery from a bad quarter” – then I think that 3.3% figure is nonsensical, and you have to focus on longer term measures. Like that 0.3% and 0.6% I pointed out earlier.
So why the 3.3% here?
Here’s the breakdown:
So basically, we’re here because Mining and Manufacturing had a much better second quarter – and those improvements had to be financed, and that product needed to be transported.
What happened with Mining?
Well, production levels were up. But also, so were global metal prices:
So the miners had a better quarter because their sales were taking place at higher prices. And would we call that a long-term trend that’s going to continue onwards and upwards until the end of the year at the same growth rate? Perhaps. But if not, then we’re talking about a re-adjustment.
According to StatsSA:
Notable increases were reported by the petroleum products, chemicals, rubber and plastic division and the motor vehicles, parts and accessories and other transport equipment division.
So two things:
- The Petroleum Industry benefited from a lower oil price; and
- The recovery in the mining industry seems to be supporting the rest of the increase in manufacturing.
Here’s the other side of the narrative:
Because our main exports are metals – and also, the Petroleum Industry, which has been actively supplying the Sub-Saharan region and benefitting from dollar revenue streams against a partially-rand-denominated cost base.
I realise that I may sound like a wet blanket, but that 3.3% number is not especially comforting. For these reasons:
- The math is weird, because the quarter-on-quarter results are so out of sync with the year-on-year numbers and the 6-month-on-6-month numbers.
- Much of the adjustment seems to have relied on a weaker rand, a weaker oil price, and stronger metal prices.
- Those are mostly external factors helping us along – not internal improvements in the local market.
- I’m not sure that we can legitimately annualise all that growth.
So treat with caution. At least, that’s my thought.