Market Making without Adverse Selection: Evidence from Retail Savings Plans

Strong data-driven microstructure paper on retail execution quality, showing predictable uninformed ETF flow receives slightly worse prices.

Paper Metadata

• Publication Date: 2026-04-30

• Source: SSRN

• Link: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=6681861

Keywords

• Market Making

• Retail Order Flow

• payment for order flow (PFOF)

Notes for Review

The setting of this paper is particularly valuable: retail savings plans are pre-scheduled and therefore largely remove the usual adverse-selection channel that complicates spread decomposition. By isolating this flow, the paper can ask whether execution costs persist even when informed trading risk is absent. The empirical design is also strong. The authors use proprietary transaction-level data from LS Exchange, a major European retail-oriented venue, and benchmark executions against high-frequency Xetra order book and trade data. This makes the paper clearly data-driven and grounded in a real major-market institutional setting. The core finding is important: although execution quality is generally high, the single market maker appears to extract small but systematic rents from savings-plan trades. For ETFs, which are the dominant savings-plan asset class, executions are worse than comparable non-savings-plan trades by roughly one basis point, despite the fact that these trades should have minimal adverse selection. This is economically modest but conceptually significant. It challenges the classic view that predictable uninformed flow should receive especially favorable pricing, and it suggests that market segmentation, broker routing, and PFOF-like arrangements can allow market makers to monetize captive retail flow. The paper does not appear to offer a direct alpha signal or deployable trading strategy, so it is not a top read from a pure alpha-generation perspective. However, it is very useful for understanding execution costs, retail venue design, quote quality, and the economics of market making. Its relevance is strongest for researchers interested in best execution, retail internalization, spread components, ETF execution, and regulatory market structure.

---

Abstract

We study how market makers price predictable, uninformed order flow, which allows the isolation of inventory costs from adverse selection. Retail savings plans (SPs) provide a unique laboratory for this analysis, as their pre-determined nature eliminates adverse selection risk. Using proprietary data from LS Exchange, a major European retail venue with a single market maker, we develop a novel methodology to identify SP trades and benchmark their execution prices against the reference market Xetra. While overall execution quality is high, we find that the market maker extracts small, systematic rents. For ETFs, the dominant SP asset class, execution is paradoxically worse than for similar in size non-SP trades, with the market maker charging a size-independent premium of approximately one basis point. Contrary to theory, this premium cannot be meaningfully explained by inventory risk. Our results show that even in the absence of adverse selection, market makers price predictable order flow distinctly worse than discretionary trades, challenging classic models of spread components and further motivating regulatory discussions regarding payment for order flow.

More from the paper..

So what they’re doing is actually pretty simple, but the setup is unusually clean.

They take this very specific type of retail flow — savings plans — where people just automatically buy every month. No decision, no timing, no alpha. Just “buy X euros of ETF on the 2nd”. From a microstructure point of view, this is about as close as you get to pure uninformed flow.

The nice trick is: because this flow is so structured, you can actually find it in raw trade data.