Under one venue’s Retail Price Improvement (RPI) rules, customers using its website or app can see resting quotes unavailable to traders connected through its API, creating different views of available liquidity.
API-based depth and slippage estimates show what traders using that connection can reach; they do not include every quote displayed to website and app users. That gap can change estimates of how much trading can occur before prices shift.
RPI orders wait on the book. At an equal price, ordinary orders take priority, and only customers using the exchange’s website or app can trade against RPI. The restriction also bars API-originated orders, including automated arbitrage and high-frequency strategies.
The eligibility limit reduces the number of trades a maker might receive. A maker could accept that constraint to avoid executions from API-connected strategies.
In one venue’s example, an API taker clears the regular best ask (Ask 1), prompting the engine to shift an RPI quote without filling it. The mechanics differ across venues; one also sets separate RPI maker fees or rebates.
The stated retail aim is greater visible depth and narrower spreads. Blocking some API flow may also spare makers trades against traders who act before prices adjust, reducing adverse-selection risk. Displayed size alone cannot show whether a maker earned a return.
A markout records price movement after a trade. The proposed comparison would check the market midpoint 100 milliseconds, one second and five seconds after RPI fills against the same intervals after regular maker trades, to see whether post-trade movement differed.
