Price Competition, Fluctuations, and Welfare Guarantees

Moshe Babaioff, Renato Paes Leme, Balasubramanian Sivan

Abstract: In various markets where sellers compete in price, price oscillations are observed rather than convergence to equilibrium. Such fluctuations have been empirically observed in the retail market for gasoline, in airline pricing and in the online sale of consumer goods. Motivated by this, we study a model of price competition in which an equilibrium rarely exists. We seek to analyze the welfare, despite the nonexistence of an equilibrium, and present welfare guarantees as a function of the market power of the sellers. We first study best response dynamics in markets with sellers that provide a homogeneous good, and show that except for a modest number of initial rounds, the welfare is guaranteed to be high. We consider two variations: in the first the sellers have full information about the valuation of the buyer. Here we show that if there are $n$ items available across all sellers and $n_{\max}$ is the maximum number of items controlled by any given seller, the ratio of the optimal welfare to the achieved welfare will be at most $\log(\frac{n}{n-n_{\max}+1})+1$. As the market power of the largest seller diminishes, the welfare becomes closer to optimal. In the second variation we consider an extended model where sellers have uncertainty about the buyer's valuation. Here we similarly show that the welfare improves as the market power of the largest seller decreases, yet with a worse ratio of $\frac{n}{n-n_{\max}+1}$. The exponential gap in welfare between the two variations quantifies the value of accurately learning the buyer valuation. Finally, we show that extending our results to heterogeneous goods in general is not possible. Even for the simple class of $k$-additive valuations, there exists a setting where the welfare approximates the optimal welfare within any non-zero factor only for $O(1/s)$ fraction of the time, where $s$ is the number of sellers.

Introduction

In classical economics, market analysis often centers on the "Nash Equilibrium"—a stable state where no seller has an incentive to change their price. However, in many real-world scenarios such as airline ticketing, gasoline retail, and online marketplaces, prices fluctuate constantly. In these models, a pure strategy equilibrium often simply does not exist.

This paper shifts the focus from static stability to the welfare guarantees of dynamic, fluctuating prices. The authors introduce the concept of the Eventual Welfare Guarantee (EWG). Rather than measuring efficiency at a static point, the EWG evaluates the ratio of optimal welfare to the welfare achieved by the market after a reasonable number of price-update rounds, regardless of whether the prices ever stabilize.

Price Dynamics and Welfare Guarantees

The study utilizes myopic best-response dynamics to model market behavior. In this setup, sellers set prices to maximize immediate revenue based on their competitors' current prices, while buyers follow a greedy algorithm to select bundles that maximize their utility (valuation minus price).

Price oscillations in a two-seller market under best-response dynamics.

As shown in the figure above, prices $p_1^t$ and $p_2^t$ often fail to converge, instead entering cycles of undercutting and resetting. Despite these persistent oscillations, the research proves that for homogeneous goods, the market remains remarkably efficient. A crucial insight in these proofs is the "Last Seller" logic: the seller who provides the final items to a greedy buyer effectively acts as a monopolist over the remaining demand, allowing for a bound on total welfare based on that seller's local optimization.

The Impact of Market Power and Information

The welfare guarantee is directly tied to $n_{max}$, the maximum number of items held by a single seller. As the market power of the largest seller diminishes, the eventual welfare approaches the social optimum. The paper highlights a significant "Value of Information" by comparing two scenarios for homogeneous goods:

  • Full Information: When sellers know the buyer's exact valuation, the welfare loss is logarithmic: $1 + \log\left(\frac{n}{n - n_{max} + 1}\right)$.
  • Uncertain Demand: When sellers must operate based on beliefs about the buyer's valuation, the welfare loss is linear: $\frac{n}{n - n_{max} + 1}$.

This exponential gap quantifies the efficiency lost when sellers must "learn" the market through observation rather than having perfect knowledge.

The Heterogeneity Paradox

While competition generally aids efficiency in homogeneous markets, the authors uncover a "Heterogeneity Paradox" for more complex goods. For $k$-additive valuations (where a buyer values a bundle based on the sum of the $k$ most valuable items), splitting a monopolist's inventory among multiple sellers can actually decrease total welfare. In some cases, the market may only achieve near-optimal welfare for a $1/s$ fraction of the time (where $s$ is the number of sellers), spending the majority of the time in low-welfare states.

Conclusion and Future Work

This work demonstrates that even in the absence of a stable equilibrium, competitive markets for homogeneous goods can provide strong welfare guarantees. However, the pessimistic results for heterogeneous goods suggest that finding specific classes of buyer valuations that yield high welfare in dynamic settings remains a significant open question. Future research may also consider non-myopic sellers who use long-term learning strategies or account for intrinsic wholesale costs.

Summary updated

Abstract

In various markets where sellers compete in price, price oscillations are observed rather than convergence to equilibrium. Such fluctuations have been empirically observed in the retail market for gasoline, in airline pricing and in the online sale of consumer goods. Motivated by this, we study a model of price competition in which an equilibrium rarely exists. We seek to analyze the welfare, despite the nonexistence of an equilibrium, and present welfare guarantees as a function of the market power of the sellers. We first study best response dynamics in markets with sellers that provide a homogeneous good, and show that except for a modest number of initial rounds, the welfare is guaranteed to be high. We consider two variations: in the first the sellers have full information about the valuation of the buyer. Here we show that if there are $n$ items available across all sellers and $n_{\max}$ is the maximum number of items controlled by any given seller, the ratio of the optimal welfare to the achieved welfare will be at most $\log(\frac{n}{n-n_{\max}+1})+1$. As the market power of the largest seller diminishes, the welfare becomes closer to optimal. In the second variation we consider an extended model where sellers have uncertainty about the buyer's valuation. Here we similarly show that the welfare improves as the market power of the largest seller decreases, yet with a worse ratio of $\frac{n}{n-n_{\max}+1}$. The exponential gap in welfare between the two variations quantifies the value of accurately learning the buyer valuation. Finally, we show that extending our results to heterogeneous goods in general is not possible. Even for the simple class of $k$-additive valuations, there exists a setting where the welfare approximates the optimal welfare within any non-zero factor only for $O(1/s)$ fraction of the time, where $s$ is the number of sellers.
BibTeX
@inproceedings{babaioff2015price,
  author = {Moshe Babaioff and Renato Paes Leme and Balasubramanian Sivan},
  title = {{Price Competition, Fluctuations, and Welfare Guarantees}},
  booktitle = {EC},
  year = {2015},
  url = {https://openreview.net/forum?id=1CydKCpf4cu},
}

Loading related content...

Figure