Wednesday, 5 July 2017

The dog that sort of barked

MBIE's commissioned report on petrol prices has left everyone up in the air. The petrol companies have been sort-of fingered for profiteering - the report (page i) says "we cannot definitely say that fuel prices in New Zealand are reasonable, but we have reason to believe that they might not be" - but there's no smoking gun. "They could be ripping you off, maybe are, but who really knows" is an unsatisfactory outcome for the petrol companies, consumers and policymakers, and probably the report authors themselves.

I'll be coming back to this report over the next few days because I'm not especially happy with the outcome, but I'll just start with three observations to be going on with.

The first is that the report indirectly makes the case for 'proper' market studies, which the government has finally agreed to. That's no criticism of the people who carried out this report - Cognitus, Grant Thornton, and the NZIER, all capable and experienced folks. But it's frankly impossible to get to the bottom of anything without information-gathering powers that the report authors didn't have (but the Commerce Commission likely will when it gets going with its own reports). They got a lot of cooperation from the petrol companies, but that only takes you so far. Nor were they given the time the Commission would likely have been allowed (inquiry announced February 9, report delivered May 29).

The second is that the terms of reference hobbled the report from the git-go, with their heavy emphasis on short-term trends: "what is the return on average capital each year since 2011?", "What are the annual gross and net margins of each of the major businesses...What trends are apparent since 2011?" (my emphasis).

The report team politely pointed out (p92) that this didn't help:
The study period was also reasonably short – 2011 to 2015 – in an industry that is characterised by long-term pricing cycles. This carried a risk that those long-term trends would not be captured in the data we were using. We have had to bear this in mind when reaching our conclusions.
They did however have the wit to smuggle in one longer-term chart on the performance of the petrol industry (there are two other long series graphs in the report, on the world real oil price, and the link between the world price and the domestic price, but other than this one I'm about to show you, nothing on the local industry itself). Here it is, from page 1 of the report. It splits the petrol price into the petrol companies' costs (lighter blue) and their margins (dark blue) since 2004, expressed in real terms (2016 prices).

You'll notice that margins virtually vanished in the very difficult GFC period, and it gets you wondering about a cyclical explanation of petrol margins. Perhaps the petrol companies do well (like many other companies) when times are good (like now), but have to sharpen their pricing pencils when household and business budgets are stretched? Wouldn't it be nice to see a longer picture, over more cycles than just the GFC and the current expansion?

And there is one. On MBIE's own site. Here it is.

There's quite a plausible case that the strength of the business cycle is part of the explanation for variations in petrol margins. You can see the fall in margins after the '87 sharemarket crash (possibly conflated with anticipation of imminent deregulation), low margins again in the '90-'91 recession, better margins in the good years in the early to mid 1990s, and the sustained rise in the current expansion.

It's not a complete explanation. There was a sustained fall in margins in the good years of the early 2000s, so cycles can't be the full answer: there must be other longer-term trends going on, too. I'd be minded to dig out the HHI index for the industry, for example, as one of my first candidates to get added to the regression.

But either way the longer sweep of history clearly has something powerful to say about the state of margins at any single point in time: would there even have been an inquiry, if someone had pointed out that current margins are pretty much the same as they were twenty years ago? Quarantining the scope of the report to the last few years was a poor decision which prevented the report from developing the full value it might have had.

Finally, the report sensibly says that even if you harbour dark thoughts about what's going on, you'd want to be mighty careful about whatever regulatory sticks and carrots you reach for. Requiring some greater price transparency, for example, sounds good, but can backfire (page 87, emphasis in original):
While at first glance this type of regulation seems attractive and pro-consumer, it is a double edged-sword, although the second edge is not obvious. While these schemes give greater information to consumers, they give the same information to suppliers. That is, they increase the ability of suppliers to coordinate their prices.
One study of the German scheme found that prices for petrol increased by between 1.2 and 3.3 euro cents per litre as a result of the scheme, while the price of diesel increased by about 2 cents per litre.
And in general, the report says (p90)
Overseas experience suggests that even the most well-intended regulations can lead to perverse outcomes and unintended consequences.
Which is something else that the long-run MBIE graph shows: deregulation put a permanent dent in petrol margins. They've never returned to the regulation era levels. And it's a reminder, for those still minded to revisit the reforms of the 1984-90 Labour government, that these days regulation is at least intended to benefit the consumer. Before 1984, it was designed to enrich the producer.


  1. Half read it. Labels product differentiation as anticompetitive. Loyalty discounts are bad too which is a bit definitive for me.

  2. Yes, it's hard to generalise about pro or anti competitive impact of business practices, often comes down to individual context