Breaking Bertrand

New Zealand's two large supermarket chains seem to compete with each other for market share.

Anyone who's done undergrad IO knows that if you have two competitors who can each take the whole market by having a slight price edge over the other guy, you can wind up with a Bertrand outcome. The duopolists behave as though they're perfect competitors. Taking the whole market is valuable. 

Now, if they could sign some cosy deal with each other, splitting up the market between themselves, that would be best from their perspective. Undercutting each others' prices to take the whole market is great and will always be in the interest of each of them in the moment, but Cournot rents are far higher. 

But they can't sign such a deal because it would be illegal under competition law. 

And that made this proposal ... interesting.

In a cross-submission following the competition watchdog's public hearings last November, Coriolis MD and founder Tim Morris said the only way to break the power of the two main supermarket owners in NZ – Foodstuffs and Woolworths/Countdown – is to force divestments when market share gets too large. 

“If you want competition in traditional supermarkets in any meaningful timeframe, you will need to force separation (at Foodstuffs) and or divestment (at WW/Countdown)," he wrote in a submission published by the commission as part of its inquiry into supermarket pricing. Foodstuffs operates as a cooperative whereas Countdown and its associated brands are part of the ASX-listed Woolworths company. 

Divest at 27% 

“As a strawman, I propose by January 2024, any food retailer with more than 27% market share in any region of NZ (eg Canterbury) shall be forced to divest stores until they reach 27%” market share", a level chosen by reference to European and North American regional peers.

Foodstuffs and Countdown wouldn't be able to sit down in a room together and make a deal requiring each to sell off some stores if it ever crossed a market-share threshold. But it's being pitched as an antitrust solution. 

I expect that they are far from perfect Bertrand competitors. But banning them from going past some market share threshold tells them, when at the threshold, to stop competing. That doesn't seem like any kind of good idea. 

I still think all of this is stupid when the regulatory barriers to new entrants are the first-order problem that haven't been dealt with.