Wednesday, 11 December 2013

Optimisation fetish strikes again

In an ironic coincidence, the day after I wished that optimised replacement cost (ORC) valuations would die in telecommunications, its effects show up in cowsheds across the country.

The gDT index of dairy commodity prices is at record levels and Fonterra watchers have been tipping that the farmgate milk price would be increased in the scheduled December announcement. Here are prices for the last five years for whole milk powder.

You can see why people were excited. But instead, farmers were told today that while the milk price manual justified a price of $9.00/kgms, it was holding its forecast payout at $8.30.

Confusion ensued: if the milk price manual says $9, why restrict the payout to $8.30? Answer: the manual is a fiction, and Fonterra can't actually afford to pay $9.

The manual/model starts from the world price of a set of commodities X, which are the ones Fonterra & its rivals compete over. Fine. But from that point onwards, the manual/model is assumption-central. Three assumptions are particularly important, and none of them is correct.

  • All of Fonterra's NZ milk is made into a product from a set we'll call X
  • Every site has a standard-sized equipment installed to make X products
  • Every month the world prices are used to decide how to divide the milk between the products in X

The manual's price is the difference between the revenue and costs arising from these assumptions.

It is the first assumption that has tripped Fonterra up. According to today's announcements, this assumption is 30% wrong - ie, 30% of Fonterra's NZ milk does not get made into products from X. It goes into less profitable stuff like cheese and casein (this is new information btw).

Normally, Fonterra likes this fiction, because by pretending to not make that less-profitable stuff it can pay higher milk prices, which makes life a bit harder for its rival processors. But there are limits to how much it can afford to artificially support the milk price, as we have seen today.

In related news, Fonterra also cut its projected divided by 22c to 10c. Predictably, the share price tanked and no doubt we will shortly hear complaints from external investors in those shares. The things to remember here are that

  • those investors supplied around 6% of Fonterra's capital, which was only issued to provide liquidity;
  • other things being equal, Fonterra would prefer a low share price (ie a low dividend) because that
    • limits the cash its farmers will get if they sell their shares and jump ship to an investor-owned processor; and
    • makes it easier for its farmers to buy the shares required to back increased production.
I think Fonterra supplier/shareholders will be happy, if they understand what's going on. But I wonder whether a less optimised milk price model might be better.

No comments:

Post a Comment