Drawback Matching

Every eligible import line can back only one export claim, so building a claim is an assignment problem: which imports pair with which exports to recover the most. Across thousands of lines and years of history, the difference between a good and a poor set of matches can be large. Two companies with identical import and export data can recover very different amounts purely because of how the lines were matched.

First-in-first-out (FIFO) and greedy first-fit matching take the earliest or the first available eligible pairing and move on. It is simple and fast, but it is not trying to maximize recovery. Optimized (global) matching instead solves across all valid permutations at once to maximize total refund, while respecting the five-year window, substitution rules, and 8-digit HTS constraints. It is a combinatorial optimization problem, not a lookup, which is why it recovers more from the same data.

The merchandise processing fee is capped per entry, so how a claim spreads recovery across entries changes how much fee you get back. First-fit matching tends to over-allocate MPF to the wrong lines and strand recoverable fee; optimized matching allocates to capture the most eligible MPF drawback. This is one of the clearest places where the matching engine, not the underlying data, decides the outcome.

No. FIFO is a common default in older drawback systems, not a legal requirement. Substitution and inventory-accounting rules govern which pairings are eligible, and within those rules you are free to choose the matches that recover the most. Choosing FIFO by default simply leaves that optimization on the table.