Skip to content

No broad same-firm markup in deep repeated urgent markets

🟡 When the price comparison is taken within the same firm-buyer-item cell — holding fixed the supplier, the buyer, and the exact product — the administrative urgent coefficient is 0.035 (SE 0.041), indistinguishable from zero. In deep, repeatedly traded urgent markets, the same supplier selling the same item to the same buyer does not charge a broad same-firm markup for urgency (AN-003).

Economic intuition

Holding the firm, buyer, and item fixed turns a price comparison into a clean test of one story: does urgency let an incumbent charge its captive buyer more? In deep repeated markets the answer is no. A null is usually weak evidence, but here it discriminates — it shuts the markup door and redirects attention to scale and supplier matching. Where markets are thin, outside options vanish and the leverage story can revive; that boundary is itself informative.

This is a deep-market null, not a universal one. The within-firm-buyer-item contrast is most informative precisely where the same trading relationship recurs often enough to be estimated, which biases the estimable cells toward high-volume, established product lines. The headline coefficient should therefore be read as evidence that the urgency premium is not primarily a matter of incumbent suppliers raising their own prices in deep repeated markets — it shifts attention away from a same-firm pricing story and toward the sourcing margin (see Fragmented sourcing is the margin). The null is not universal, however: a residual within-firm gap persists in the earlier period (the quantity dimension instead reflects scale, not same-firm pricing — see The deep-market null is not universal).

Caveat. The near-zero within-firm-buyer-item coefficient applies to deep repeated urgent markets only and must not be generalized to all markets: the heterogeneity cuts in AN-003 show a residual within-firm gap in the earlier period (+0.117, administrative- dearer), while the below-median-quantity positive (+0.066) reflects scale — the within-triple log-quantity coefficient is −0.259 — rather than same-firm pricing, and the off-formulary positive (+0.101) is not statistically significant. The estimable cells are themselves selected toward deep relationships, so the null is informative about where same-firm pricing power is absent rather than a claim that it is absent in general. The reading is 🟡 because it is a single-source own-project estimate in São Paulo BEC.

Sources.