CUP method

Example 1 — Coffee beans: Colombian, then Brazilian

Chapter II · CUP method · ¶2.24

OECD Guidelines 2022
At a glance
Unbranded Colombian coffee beans are sold between two associated enterprises, and an independent enterprise sells beans of a similar type, quality and quantity under similar conditions. If the only uncontrolled sale available were of unbranded Brazilian beans instead, how far could the CUP method still be relied on?
  1. One associated enterprise sells unbranded coffee beans from Colombia to another.
  2. Unbranded Colombian coffee beans are also sold by an independent enterprise; in type, quality and quantity they are similar to the beans in the controlled sales.
  3. The two sets of sales are assumed to coincide roughly in time, to sit at one and the same stage of the production and distribution chain, and to be made under similar conditions.
  4. In a variation, the only uncontrolled transaction available involves unbranded Brazilian coffee beans.
A

Unbranded Colombian coffee beans are what the independent enterprise sells and what is sold between the two associated enterprises, and the beans in the two sales are similar in type, quality and quantity. Timing is about the same, the stage in the chain is the same and the conditions are similar.

How reliable is the CUP method on these facts?

OECD analysis · CUP particularly reliable
For the Guidelines this is where the CUP method is particularly reliable: the very product that passes between the two associated enterprises is also sold by an independent enterprise. The beans are similar in type, quality and quantity, and the example assumes about the same timing, the same stage in the chain and similar conditions for both sets of sales. The uncontrolled sale in the example is one made by an independent enterprise.
B

The only uncontrolled transaction available now involves unbranded Brazilian coffee beans, whereas the beans sold between the associated enterprises are Colombian.

What should be done with the Brazilian sale?

OECD analysis · Inquire; adjust if material
The Guidelines neither discard the Brazilian sale nor accept it untested. The appropriate step is to ask whether the difference between the beans matters materially to price, for instance whether beans from one source generally fetch a premium, or sell at a discount, in the open market. Commodity markets may supply that information, or it may be worked out from dealer prices. Where the effect on price is material, the Guidelines consider some adjustment appropriate.
C

The difference between the Colombian and the Brazilian beans does have a material effect on price, and a reasonably accurate adjustment for it cannot be made.

What follows for the CUP method?

OECD analysis · Reliability reduced
Where a material difference cannot be adjusted for with reasonable accuracy, the Guidelines say the CUP method loses reliability. Another, less direct method might then have to be chosen. The wording is measured: reliability is reduced, and a change of method is a possibility, not an automatic consequence.
Key takeaway
Where the product sold between associated enterprises is also sold by an independent enterprise in similar circumstances, the CUP method is particularly reliable. Where the product differs, ask whether the difference materially affects price and adjust if it does; if no reasonably accurate adjustment can be made, the method is less reliable and another might have to be chosen.

Related: Part II, Section B.2 (examples of the CUP method) · ¶2.25 and ¶2.26 (differences in delivery terms and in volume)

Test yourself on this example, question by question.Open it in the interactive walkthrough →