Cocoa costs barely affect what you pay for a Freddo. For that, be glad #NewsMarket

A recent MainFT piece dissecting the jumping price for Australia’s ‘Freddo frog’ chocolates — known simply as Freddos in the UK, because it’s obvious who Freddo is — prompted a valid response from readers of that austere organ:

Chocolate enthusiasts are presumably divided on this point. On the one hand, it is legally-speaking a chocolate — as opposed to something that is merely chocolate-flavoured. On the other hand it, like most milk chocolate, is mainly a lot of things that aren’t chocolate.

Cocoa is the main problem for chocolate makers, because its price has gone up. As a result, they’d really prefer to sell you as little chocolate as possible when you buy chocolate.

Bloomberg did an interesting piece about this back in February:

Facing limited room for additional price hikes, companies are shrinking packages, using automation to trim production costs and promoting products with less cocoa or other starring flavors…

The switch from cocoa butter to noncocoa substitutes works better with nonpremium chocolate applications, such as thin coatings on granola bars and fillings in bakery items, rather than more traditional chocolate bars. Noncocoa substitutes have long been more common in hot countries, where cocoa butter’s low melting point can pose a problem.

But we digress.

An obvious question is this: if my chocolate bar is only 20 per cent chocolate, then why does the cocoa price matter that much?

Well, it doesn’t. Cocoa prices are a tiny part of the picture when it comes to milk chocolate inflation, though (for obvious reasons) they’re one of the parts producers are more keen to talk about.

There’s a widely circulated estimate created by NGO Oxfam, which breaks down the actual cost of a bar thusly:

If that’s even roughly correct, the actual impact of cocoa prices should be utterly marginal, and (brace yourself) the corporations are playing you.

So what should we do, just not write an article? No, let’s be stupid: how much would a Freddo cost if its price was entirely derived from the cost of its key ingredients?

For whatever reason, Freddos have assumed a memetic status with regards to inflation. Everyone in Britain over the age of… 25?… remembers that they used to be 10 pence in the early 00s. Now, they seemingly cost 25p and up (35p in the FT’s local supermarket), plus they’ve been resized in both directions — from 17g to 20g in 2007, then back to 18g in 2011. Creating a price index is tricky, because current prices are wildly different.

Here’s the ingredient list for an 18g UK Freddo (which is basically the same for all Cadbury’s chocolate under the Dairy Milk label):

MILK**, sugar, cocoa butter, cocoa mass, vegetable fats (palm, shea), emulsifiers (E442, E476), flavourings.

Cadbury’s (part of Mondelez) says its Dairy Milk contains “The equivalent of 426ml of fresh liquid milk in every 227g of milk chocolate”.

Thinking about this hurt Alphaville’s brain, so we popped to Sainsbury’s to do some journalism.

Thankfully, the Freddo/Dairy Milk family packaging offers more details:

MILK SOLIDS 20% MINIMUM, ACTUAL 23%, COCOA SOLIDS 20% MINIMUM

Because ingredients are always listed from largest to smallest ingredient, this gives us most of what we need to deconstruct the bar. 23 per cent is powdered milk and 20 per cent minimum is cocoa, with sugar somewhere between, and about a third of the bar yet to be clearly accounted for.

Clearly, the amount of sugar must fall within a fairly narrow range. Each 100g Dairy Milk has 55g of sugar (10g in an 18g Freddo bar). If up to 23g of that is from the sugar itself, so where does the other 32g come from? The internet suggests milk powder is about 50 per cent sugar, so that gives us another 11.5g. The cocoa mass must have a bit more. The rest, search us. (Learnèd comments welcome).

Ugh. For the sake of progression, let’s assume cocoa is at the 20 per cent minimum and that just sugar is halfway between that and milk. So, 21 per cent.

That leaves 35.5 per cent of the bar left to account for. Flavourings are presumably minimal. The consistency-enhancing emulsifier E442, aka ammonium phosphatide, makes up 0.5 per cent of “all chocolate products” according to the European Food Information Council. Assuming that is correct, and that the descending quantity logic continues in paratheses, E476, aka polyglycerol polyricinoleate, must be less than that, but let’s say about the same.

That leaves perhaps 34.5 per cent palm oil and shea butter, with palm oil (the presence of which often makes people unhappy) taking the edge and neither individually above 20 per cent. Let’s just split the difference roughly, at 17.5 per cent palm and 17 per cent butter. Voilà, a rough chocolate bar:

Yummy.

Now, the commodity calculation.

First, a problem. The mix of ingredients in UK Dairy Milk has changed over the years — with the general consensus being that this has involved including more sugar and palm oil. Hence why it tastes worse than you remember back when you were a kid.

So let’s imagine two blatantly incorrect things for sake of a chart:

1) The sole driving factor in the price of Freddo is the cost of its ingredients
2) The Freddo recipe remained the same since the turn of the century.

We could construct a basket of Dairy Milk components far more easily, using a pretty random assortment of generic contracts for milk (milk powder didn’t go far enough back, but moves fairly similarly), sugar, cocoa, and palm oil (we couldn’t find a shea price on the Bloomberg Terminal).

We know, on a spiritual level, that Freddos cost 10p in the year 2000. So let’s rebase our basket to as close to then and weight the components, assuming:

— the price of 10p in 2000 was on its own terms correct.
— palm oil and shea have moved the same way as fellow fats — unlikely but needs must.
— the “others” can be ignored.
— that if something takes a long time to do it must be worth doing ¯\_(ツ)_/¯

Finally, we need to reweight by price, rather than proportion, otherwise we might end up with some weird skews.

Some spreadsheet work later, here’s the outcome:

What, if anything, can we learn from this?

— [Ed: oh]

Further reading
— How much does a pint of lager cost? An Alphaville investigation