Punitive policies to combat misreporting in dynamic supply chains

TL;DR

The paper addresses misreporting of market potential in a dynamic, information-asymmetric supply chain by casting the interaction as a dynamic Stackelberg game with a leader (supplier) and follower (manufacturer). It proposes two punitive pricing policies that align incentives for truthful reporting under quasi-static market fluctuations and proves truth elicitation for greedy misreporting, with a second micro-observation policy providing convergence guarantees against strategic misreporting; both yield per-state Stackelberg equilibria $p_i^*,q_i^*$ that scale with the true market potential. The key results include explicit SBE expressions, a penalty-based truth-revelation mechanism $q_{\pi}(\hat{\phi})$, and stochastic-approximation proofs of convergence under micro-observation policies. Numerical Monte Carlo studies validate the approach, showing that large penalties deter misreporting and the micro-observation policy achieves robustness against strategic behavior, highlighting a path toward near full-information efficiency in asymmetric SCs.

Abstract

Wholesale price contracts are known to be associated with double marginalization effects, which prevents supply chains from achieving their true market share. In a dynamic setting under information asymmetry, these inefficiencies manifest in the form of misreporting of the market potential by the manufacturer to the supplier, again leading to the loss of market share. We pose the dynamics of interaction between the supplier and manufacturer as the Stackelberg game and develop theoretical results for optimal punitive strategies that the supplier can implement to ensure that the manufacturer truthfully reveals the market potential in the single-stage setting. Later, we validate these results through the randomly generated, Monte-Carlo simulation based numerical examples.

Figures (2)

• Figure 1: Supply Chain: flow of materials, prices and information
• Figure 2: Left: Impact of punitive policies on various misreporting policies Right: Time averaged utilities as a function of $T$ for different ${\bm r}$ and $\pi$.