Hedging when you do not know your volumes: the layered method
The most common situation in a mid-market company is also the least covered: a real but non-contractual exposure. How to hedge an imperfect forecast without creating a speculative position.
Textbooks describe hedging an invoice: known amount, known date, forward contract. In the reality of a mid-market company, most of the exposure looks nothing like that. It looks like: « we do around 4 million dollars a year, mostly in the second half, but last year we did 3.1 and the year before 4.8 ».
Stating the problem properly
Hedging a forecast means arbitrating between two errors of opposite nature:
- Under-hedging: the flow materialises, the currency has moved against you, the margin is cut.
- Over-hedging: the flow does not materialise, you hold a firm commitment with no underlying, and you unwind it at the market rate.
The first error costs a margin. The second costs a margin and takes the transaction outside the scope of hedge accounting. They are not symmetrical, and the method has to reflect that.
The layered method
Rather than a single decision on a single amount, the exposure is cut into layers of decreasing probability, each handled with a suitable instrument.
| Layer | Definition | Typical share | Instrument |
|---|---|---|---|
| Core | The minimum achieved over the last three years, month by month | 50 – 60% | Firm forward contract |
| Intermediate | Between the core and the year's budget | 25 – 35% | Partial forward or call/put option |
| Upper | Above budget, commercial upside | 10 – 20% | Not hedged, or option only |
The core is not a forecast, it is a history. It is computed from your own past realisations, with no commercial assumption. That is what makes it defensible to an auditor.
Computing the core, in practice
- 1Take three years of realised flows by currency and by month
Actual receipts, not invoices issued.
- 2Keep the lowest monthly figure observed
For each calendar month, the lowest of the three years. That is prudent by construction.
- 3Apply a safety haircut
10 to 20 % depending on the stability of your customer base. A customer representing 40 % of the flow justifies a larger haircut.
- 4Recalibrate each quarter
The core moves with the business. A quarterly recalibration is enough and avoids jolts.
The limit-order alternative
On the intermediate layer, a limit order lets you set a target level without an immediate commitment: if the market reaches your rate, the hedge triggers automatically. It is the right tool when you know the rate you want to be hedged at but not whether the flow will exist.
The limitation is explicit: if the market never reaches the level, nothing happens. A limit order does not replace a policy, it executes one.
What not to do
- Hedging the commercial budget at 100 %: that is hedging an ambition, not an exposure.
- Waiting for the firm order before hedging: on a six-month sales cycle, that amounts to never hedging.
- Using a barrier product to « improve the rate » on an uncertain exposure: you then stack two uncertainties, the flow's and the barrier's.
See the effect on your margin
The exposure simulator projects the impact of an FX scenario on your margin, layer by layer, over twelve months.
Simulate my layersNot ready to open the pricer yet?
Get the one-page FX policy template — hedge ratios by horizon, tolerance band, delegation table. The one your auditor will ask for.
- One field only
- One-click unsubscribe
- No data resold
Frequently asked questions
The questions finance departments ask most often on this topic
Can you apply hedge accounting to a forecast flow?
Yes, provided the flow qualifies as « highly probable » and that qualification is documented before the hedge is put in place, generally through a track record of realisation. That is exactly what the core method produces: a statistical justification predating the transaction.
What if the forecast flow ultimately does not materialise?
The hedge has to be unwound at the market rate, and the result of that unwind goes to financial income instead of adjusting the commercial margin. If it keeps happening, it is the core that is badly calibrated, not the market that is hostile: it needs reducing.
How often should the core be recalibrated?
Quarterly in most cases. Monthly if your business is highly seasonal or if one customer represents more than a third of the flow in a currency. Beyond that, the administrative cost of recalibration exceeds the gain in accuracy.