Either ESKOM or Stats-SA are wrong

2015-11-16 19:58

In his latest quarterly briefing, Eskom CEO Brian Molefe has said that energy expert Chris Yelland's story showing that demand has reduced dramatically is "way off the mark".

I noticed that Yellands' estimates of demand contraction were close to mine posted a few weeks back, except he was using weekly data whereas I was getting monthly data from Stats SA. I would love to know where he gets his weekly data from. Nonetheless, I've got respect for both these gentleman, so I decided to double check my Stats-SA numbers, as well as update them with the latest ESKOM electricity output reading for September:

I'm sorry, but despite a jump in the latest seasonally adjusted output numbers, decades of studying economic time series tells me that regardless if you look at the peak-to-peak raw numbers or the peak-to-trough seasonally adjusted numbers, electricity output has fallen dramatically since it peaked in 2007, and continues to fall. The seasonally adjusted numbers (provided by Stats-SA) show current output at 2008 Great Recession levels. All of this despite an aggressive R213Bn build-program of new coal-fired and gas-fired power plants, wind-farms, solar-farms and what have you.

Now given that we all know (and have experienced) that ESKOM does not have enough electricity to support the demands of the economy (or else why did we have load-shedding?) and load-shedding has suddenly stopped, and electricity generation is falling, one can only come to one of the following conclusions:

1.I am interpreting all the data wrong.

2.ESKOM is right and the Stats-SA numbers are rubbish

3.ESKOM is wrong and the Stats-SA numbers do indeed show a demand crunch to 2008 recession levels

I think the way ESKOM presents its non-seasonally adjusted numbers can be (unintentionally) misleading. I took my monthly Stats-SA numbers and put them in the same format as that used by ESKOM in their briefing to get the following:

Yes, this shows only a 2.93 to 3.71% decline over the same month last year and the year before that, but this kind of comparison is a dangerous one with non-seasonally adjusted numbers in a highly seasonal time series. Let's play this game with the seasonally adjusted numbers:

Note how the 7% collapse over the 5 months from March to August suddenly becomes visible. Also the trend doesn't look so great anymore. I double-checked the seasonal analysis used by Stats-SA. This is not Mickey Mouse seasonal analysis- I managed to 99% replicate the Stats-SA seasonal adjustments using the latest X-13 ARIMA Seasonal Adjustment Methodology used by the United States Census Bureau. If the seasonal adjusted data is telling us we are at 2008 recession levels then I am inclined to believe it.

While we are at it, let's play this game going back to 2007 and show the high, low and average non-seasonal prints for each month of the year. We can see that for the 6-months since April we started outputting less Gigawatt-Hours than the 9-year average (dotted line).

But more concerning, the above chart shows that the months of June, July, August and September were NEW 9-YEAR LOWS! (below the red line)

I think looking at weekly non-seasonally adjusted numbers for this month and comparing them to just the same months of 2014 and 2013 may be inclined to encourage complacency. If you are going to compare, look all the way back to the peak of the last business cycle (2007) and also look at the seasonally adjusted numbers in a continuous time series.

Just because non-seasonal data is 2% less than the same time last year doesn't mean we are in the clear - what if the 4-months before were also 2-3% below their prior year prints? Looking at specific months of non-seasonal data and comparing them to prior years hides the cumulative effect of a series of under-performing months. That's why I prefer to look at continuous  time series in addition to the comparative ones used by Yelland and ESKOM. They certainly bring a few home-truths to the fore.

I respect the new management team at ESKOM and no doubt they have done an awesome job to clean up the horrible calamitous mess it was before. I am just as happy as the person next to me that there is no more load shedding. But I am inclined to throw my weight behind the "demand crunch" theory right now, until the data tells me otherwise. I'm betting this is going to start showing up in a raft of SA leading and coincident economic indicators leading up to December 2015.

If ESKOM can show why we are reaching the wrong conclusions here (in a civil manner I hope) I am the first to admit where I was wrong and adjust my theories accordingly. Besides, I rely on this sort of data to manage the risk of my own money in the stock markets and I would be delighted for someone to tell me I am reaching the wrong conclusions, to prevent me from making wrong investment decisions.

News24 Voices Terms & Conditions.

Join the conversation!

24.com encourages commentary submitted via MyNews24. Contributions of 200 words or more will be considered for publication.

We reserve editorial discretion to decide what will be published.
Read our comments policy for guidelines on contributions.
NEXT ON NEWS24X

AB praises selfless skipper

2010-11-21 18:15

Inside News24

 
/News
 

And the Paws24 and Hill's winners are ...

Find out who the winners of our Paws24 pet pics and Q&a competitions are...

 

Paws

Keep your family and pets safe from rabies
5 scientific benefits of owning a cat
Why we love cats
8 great natural remedies for your pet
Traffic Alerts
Traffic
There are new stories on the homepage. Click here to see them.
 
English
Afrikaans
isiZulu

Hello 

Create Profile

Creating your profile will enable you to submit photos and stories to get published on News24.


Please provide a username for your profile page:

This username must be unique, cannot be edited and will be used in the URL to your profile page across the entire 24.com network.

Settings

Location Settings

News24 allows you to edit the display of certain components based on a location. If you wish to personalise the page based on your preferences, please select a location for each component and click "Submit" in order for the changes to take affect.




Facebook Sign-In

Hi News addict,

Join the News24 Community to be involved in breaking the news.

Log in with Facebook to comment and personalise news, weather and listings.