Bufetico
Inditex published its results for the six months to July 31 this morning. Sales of 19.8 billion euros, up 7.6%.
Two lines later, in the same document: up 9.2% in constant currency.
The two figures measure two different things, and the argument about a company like this one usually comes down to which of the two you decide to quote.
What the two numbers are
The 7.6% is what happened. Money came in, got converted into euros at whatever the exchange rates were that day, and landed on a set of accounts an auditor signed.
The 9.2% is a reconstruction. It asks what growth would have been if exchange rates hadn’t moved at all since last year, which is a thing that didn’t happen. To build it, someone has to decide which rates to hold still, over which period, applied to which pile of sales. That someone works at the company.
Constant currency is a pretty standard disclosure, and a business selling in dozens of countries needs it. But one of those numbers was audited and the other one is management’s arithmetic on assumptions management chose, and that’s worth keeping in mind while you read.
The number that gets skipped
Both figures get repeated everywhere. The gap between them almost never does.
1.6 percentage points. That’s the size of the currency effect on this company’s sales this half, stated by the company itself, and it’s the only part of the disclosure that tells you something you couldn’t have guessed.
If that gap is small, currency is noise and you can read the reported number and move on. If it’s large, the reported growth is partly a story about the dollar, the peso and the yen, and reading it as a story about clothes will mislead you. The gap tells you how much to trust the rest.
Get in the habit of writing it down. It’s one subtraction, it takes ten seconds, and it’s worth more than either number on its own.
The second place currency shows up
There’s a second place where currency decides a lot, and it’s the one that made me pay attention to the first.
To value a company you have to bring its future cash back to today, and the rate you use for that should be the one that belongs to the currency the company reports in. Inditex reports in euros, so it gets the euro rate. A Japanese company reporting in yen gets the Japanese one. I went through what happens when you get this wrong a few days ago, with Toyota (/blog/sovereign-curve-discount-rate): the gap between what Japan and the US paid on their bonds was enough on its own to make a Japanese company look like something it isn’t.
So when a company hands me two growth numbers in two currencies, I’m already thinking about which currency the whole valuation is standing on. For a company that collects in dozens of currencies and reports in one, even the reporting currency is a simplification. A defensible one, though not a perfect one.
One more line from this morning
Sales grew 7.6%. Cash generated by the operations grew 11%, to 4.1 billion euros. The net cash position finished the half at 10.4 billion, 4% higher than a year ago.
Cash growing faster than sales is worth noticing, because the two can come apart for boring reasons and for serious ones. It can mean the business is collecting faster or holding less stock. It can also mean a payment landed on one side of a date instead of the other.
It doesn’t settle anything by itself. I’ll keep it as a question for the next set of accounts.
Where this breaks
Everything above pushes you toward the constant-currency figure as the cleaner read of the business. Over one half-year, it usually is. Over many years, it can be a lie you tell yourself every time.
A company earning in a currency that keeps weakening will report a currency drag every single period, and stripping it out every single time describes a business that doesn’t exist. The euros that reach the accounts are the euros you own. If a company needs its results adjusted for currency to look like it’s growing, then over a long enough stretch it isn’t growing, whatever the adjusted line says.
My own rule has the same flaw pointed the other way. Using the rate of the reporting currency is right for a company whose whole economy lives in that currency, and only more or less right for one that sells everywhere and reports in one place. I use it because the alternatives are worse and because I can apply it the same way to every company on the list, which matters more to me than being a little more accurate on some of them.
Both figures were true this morning. I wrote down the 1.6.