Rebutting a time-series comparison
Check the assumptions it rests on.
By examining the assumptions it rests on. The time-series comparison sets weekly goods input against weekly revenue and infers the whole period from the highest gross mark-up reached. That only carries if purchase and sale fall within the same period, if range and prices stay largely constant and if no stock is shifted. If one of those assumptions does not hold, the method does not measure the mark-up but the shift.
This is first a question of arithmetic, not of law
The Bundesfinanzhof tied the time-series comparison to narrow preconditions in its judgment of 25 March 2015 (X R 20/13) and held it unsuitable in particular where the ratio of goods input to revenue is not largely constant in the business. The objection works only if it is evidenced, however: the assertion that the range changed is worthless as long as it does not follow from the purchase data.
The assumptions and how to test them
| Assumption of the method | Where it can break | What shows it |
|---|---|---|
| Purchase and sale fall in the same period | Stockholding, bulk packs, promotional buying | The incoming invoices with quantity and date |
| The range stays the same | Seasonal menu, change of range, new products | The item master data across the period |
| Prices stay the same | Price adjustments, promotions, lunch menu | The price change history of the till |
| The gross mark-up is the same across products | Beverages and food carry very different mark-ups | The recipes and the sales statistics |
| There is no free consumption | Staff meals, own consumption, shrinkage, breakage | The operational records kept for this |
| The comparison period is representative | Refurbishment, business holidays, exceptional events | The documented course of business |
The more effective route does not run through criticism
A time-series comparison is a Verprobung — a plausibility test that infers a result from a few variables. Criticising it means staying inside its logic. Presenting the complete calculation instead makes it moot: under § 162 Abs. 1 Satz 1 AO estimation applies only in so far as the bases of taxation cannot be established or calculated. What has been calculated does not need to be tested for plausibility.
What we deliberately do not publish
How we build such a calculation in an individual case is not on this page. The point of attack in a method is public — it appears in every relevant judgment. The calculation with which we replace it is not. Those who need the method will find it in our seminars; those who need the result will find it with us.
What we contribute
We record the complete goods input, store the actual recipes and recalculate the quantity produced and sold. Every figure leads back to a document, every step is documented so that a third party can follow it.
The legal provisions cited here are given for information and do not replace legal advice.