On November 21st, NYSE Arca reintroduced the Discretionary Pegged Order (DPO)1. The DPO rests non-displayed at the near side of the market (i.e., at the bid for buy orders and the offer for sell orders) and exercises the least amount of price discretion necessary to trade with contra-side orders up/down to the midpoint. This means a DPO order will trade at midpoint only when the contra-side order is a midpoint order.
The DPO offers discretion otherwise unavailable to market participants, by having knowledge of both orders’ price instructions when executing a match. DPO users benefit as they receive higher-quality midpoint fills when available and avoid providing midpoint price improvement to spread-crossing orders. The graphic below details how DPO determines whether to trade at midpoint or at another price.
Chart 1: NYSE Arca DPO Example: DPO Buy Order
To demonstrate the potential benefit from discretionary midpoint usage, we analyzed the contra-side order instruction to resting midpoint fills on current standard midpoint orders. We grouped all NYSE Arca midpoint fills2 based on whether the price instruction of the liquidity-removing order was at midpoint or more aggressive than midpoint. About 58% of current resting midpoint fills traded against incoming orders with midpoint price instructions.
To measure fill experience, we calculated the per-share fill price to midpoint change in dollars and the midpoint to midpoint returns in basis points at various short-term time horizons. Resting midpoint orders interacting with contra-side midpoint orders experience far better short-term execution quality than those that trade with non-midpoint contra-side orders.
Chart 2: Resting midpoint fills per-share PnL
Chart 3: Resting midpoint fills mid-to-mid markout (bps)
These same performance stats grouped by min quantity usage show both the limited use of min quantity and uncertain benefits associated with its usage. Across our sample, 84% of executed volume had no min quantity (97% of resting midpoint orders are entered with min quantity). Orders with no min quantity and a midpoint contra had far better performance than any other combination, while still averaging 180 shares per fill.
Table 1: Resting Midpoint Min Quantity Usage
% of Executed Volume | 100ms Per-Share PnL | 100ms Mid-to-Mid Markout (bps) | Avg Fill Size (shares) | Avg Fill Size ($) | ||
---|---|---|---|---|---|---|
Midpoint | No Min Qty | 56.8% | ($0.0005) | -1.0 | 180 | $8,142 |
Min Qty=100 | 5.9% | ($0.0018) | -0.9 | 149 | $4,556 | |
Min Qty>100 | 0.7% | ($0.0036) | -0.8 | 379 | $27,110 | |
Beyond Midpoint | No Min Qty | 27.6% | ($0.0039) | -0.2 | 94 | $12,290 |
Min Qty=100 | 7.0% | ($0.0072) | -0.4 | 148 | $3,866 | |
Min Qty>100 | 1.9% | ($0.0054) | -0.5 | 554 | $35,415 |
Finally, we compared midpoint to beyond midpoint interactions based on spread width at the time of trade. For this measurement we used the 100ms midpoint-to-midpoint return, divided by the spread at the time of trade, to get a return value in terms of spread. In this case, we see that the impact of more aggressive take orders is most pronounced in tighter-spread names, and that the benefit of interacting with midpoint orders holds across various spread scenarios.
Chart 4: Resting midpoint fills 100ms mid-to-mid markout (spreads)
1. An earlier version of DPO on NYSE Arca included a Quote Stability Indicator. This is no longer part of the NYSE Arca DPO.
2. Includes midpoint DAY orders executing during regular hours, September 1 - November 30, 2022. Orders sent with a midpoint instruction but executing at a price other than midpoint due to limit prices, and executions when the NBBO is locked, are excluded.