Market makers play a central role in financial markets by quoting bid and ask prices that allow buyers and sellers to trade continuously. In sequential market making, each trade arrives independently according to price thresholds, which means the market maker must adjust prices dynamically to stay profitable while also managing inventory risk.
A new paper on arXiv addresses this challenge by framing it as a revenue maximization problem under inventory constraints. The authors propose an algorithm that achieves no-regret performance, meaning it can learn and adapt to market conditions without suffering systematic losses over time, even when inventory limits are binding.
The work is theoretical and does not provide empirical results in the abstract, but it offers a formal framework for designing market-making strategies that balance revenue and risk. For practitioners, this could inform more robust automated trading systems that operate under real-world capital and inventory restrictions.