Why is there still currency noise in the results after you hedged?
· 5 min read
Somebody asks it at the close. Sometimes the CFO, sometimes an audit committee member, sometimes a business unit head who has just been handed a variance. The wording moves around. The question doesn’t. We hedge. Why is there still a currency number in here?
It is almost never an accusation, and the person asking is right to be puzzled: a program can be running exactly as designed and the results can still carry a currency number. Explaining why, in the time you get, is the hard part.
Three things make it hard. None of them is anyone’s oversight.
01The two sides are recorded in different places
Where cover is held against balances that get revalued every period, both sides land in the income statement in the same period, usually in the same line, and they largely take each other out there. Where the protection is held back until the sale or purchase it was placed for actually happens, it arrives later, and in the operating line, in revenue or cost of sales, rather than in the currency line.
Fig 1 · Where each side lands
The same program, recorded in two places.
A company can be well covered and still show a currency line that looks unprotected.
Before Anyone Asks
Someone reading down the income statement for the answer will not find it, and will reasonably conclude the program did not work.
02The two sides are recorded on different dates
A rate gets set for planning in one month. Cover goes on across several months, in pieces. The exposure is booked when it is booked. Results are translated at each month’s average.
Fig 2 · Four dates
Four dates, one comparison.
Both numbers are correct, and neither measures the program.
Before Anyone Asks
Once four different dates are involved, the arithmetic guarantees a difference. A position built in layers has a blended rate, and a blended rate will not equal any single snapshot rate except by coincidence. That difference is not a result. It is what you get for comparing two numbers that were never the same measurement.
This is the one most often mistaken for performance, because it is the comparison the business actually makes: our plan said this, our results say that. Both numbers are correct. Neither measures the program.
03Some of what gets called currency noise is not what the cover was pointed at
Nobody in treasury needs telling that a subsidiary’s results converted at the period’s average rate move reported revenue and margin, and that cover placed against balances does not hold that still. Say it first anyway: it is frequently the largest part of what the person asking means by “currency noise,” and the part the program was never asked to address. So the first move is a question back. Which number are we looking at? A translation question answered with a hedging answer is how an easy conversation turns into a long one.
Fig 3 · A question back
Which number are we looking at?
A translation question answered with a hedging answer is how an easy conversation turns into a long one.
Before Anyone Asks
What actually cancels, and what doesn’t
Where the two sides do land in the same line, what cancels is the spot move. That is the design working: the currency went one way on the balance and the other way on the cover.
What does not cancel is everything that was priced into the contract on the day it was placed. The forward points, a cost on some pairs and a pickup on others, and a published number either way at the moment of trade. The distance between the rate that was dealt and the market’s mid at that moment. How many times a position was renewed, and what each renewal cost.
None of that is market movement. All of it was in dollars, on the screen, on the day.
One qualification, because it decides whether the point holds. The points fixed at inception are what accrues over the life of a contract, so for a position that settled inside the period, that is the number. A position still open at the period end is marked using today’s rate differential rather than the one it was struck at, and that movement was not knowable either. It washes out by maturity; at a quarter-end snapshot it does not, and it belongs on the other side of the line.
Which means the residue in that line, once the two sides have largely offset, is not the market. The market has cancelled itself. What is left standing is what the program cost to run.
Why the separation is the point
Fig 4 · The line
What was knowable when the decision was made, and what wasn’t.
The test is whether it could have been known when the decision was made, not whether it could have been changed.
Before Anyone Asks
Leave the number undivided and what gets discussed is the net outcome, which has a rate path inside it. A view gets formed about the program, and through it about the people running it, on the strength of something nobody in the building could have known in advance. Nobody intends that. It is what happens to a number that arrives whole.
Draw the line and the subject changes. What is left to discuss is what was on the screen when each decision was made: the points, the rate dealt against the mid, the tenors actually available, the size and timing of the position. Those were in dollars, on somebody’s desk, at the time, and they can be taken into account when the strategy is set next time, whoever sets it. A rate path cannot.
Be precise about the test that sorts them, because it is not the obvious one. Not whether something could have been changed. Whether it could have been known when the decision was made. Nobody moves an interest rate differential; carry is not anyone’s to control. It was still published, in dollars, on the day you traded. The two come apart much further than people expect, and most of the priced-in number sits in the gap. That is worth an issue of its own.
Saying it out loud
The version that survives the room is short.
The market moved, and the program took most of it out. What is left in that line is what the program cost to run, and that part was priced in on the day we traded, not decided by the market afterwards. We checked that the cover stayed close to the balances it was placed against, which is what has to be true for the first two sentences to hold.
Three sentences. The first two answer the question. The third is what makes the first two believable, and it is the one that usually gets left off.
Because the question at the close is not really did the program work. It is can you tell me what this number is. Only one of those has an answer, and it happens to be the one being asked.