NYSE Closing Auction: Timing Shifts and Marketability Trends
Author
Sophia Lichoulas
Senior Data Scientist, NYSE
Published
November 18, 2025
On August 29, 2025, NYSE Floor brokers began exclusively using third-party Order Management System (OMS) technology that automates D-Order entry. This article reviews the closing auction environment, focusing on Market on Close (MOC) and Limit on Close (LOC) orders and D-Orders. It then examines how D-Order entry timing has shifted with the expanded use of third-party OMS technology and explores trends in order marketability.
Three main order types participate in the NYSE closing auction. MOC and LOC orders may be entered until 3:50:00 p.m., unless a significant imbalance extends offsetting order entry to 4:00:00 p.m. D-Orders may be entered until 3:59:50 p.m. - 10 seconds before the 4:00:00 p.m. auction. To illustrate order volume and timing, we examine average auction interest (defined as open, not canceled positions) between 3:00:00 p.m. and the closing auction during September 2025 (Figure 1). Leading into the close, D-Orders account for roughly 60% of volume, while MOC and LOC orders each represent about 20%. Not all limit orders and D-Orders are marketable - a point we return to later. We also see a significant uptick of D-Order volume beginning around 3:57:30.
Figure 1: Total Auction Volume by Order Type
Figure 1: Includes MOCs, LOCs, and D-Orders entered at 3:00:00 p.m. or later in September 2025. Only S&P 500 securities listed on NYSE are considered. Canceled orders are removed from volume at the time of cancellation. Imbalance orders are grouped with LOCs but represent a small fraction of total volume. Volumes are aggregated by second.
Figure 2 normalizes Figure 1 to show the timing of each order type, with each scaled to 100% of its total volume by the close. This view highlights that over 60% of D-Orders are submitted after 3:57:30 p.m.
Figure 2: Total Auction Volume by Order Type, Scaled Individually
Figure 2: Displays the same dataset as Figure 1, normalized to show the timing distribution of each order type as a percentage of its total volume.
D-Orders, due to their added flexibility, tend to be entered later than MOCs and LOCs. Between August 2024 and September 2025, as the share of D-Orders entered by Floor brokers using third-party OMS technology increased from 69.3% to 100%, the percentage submitted in the final 10 seconds of trading rose from 4.45% to 13.4%. This trend underscores the success of third-party OMSs in reducing operational risk and facilitating later D-Order submission.
Figure 3: Third Party OMS Usage and Late Order Entry
Figure 3: Shows the percentage of D-Orders entered after 3:00:00 p.m. that were submitted between 3:59:40 p.m. and 3:59:50 p.m. for each month. Only S&P 500 securities listed on the NYSE are included. Canceled orders are excluded from volume at the time of cancellation.
Next, we shift from analyzing order timing to examining marketability, a function of both price and time. We consider how the relationship between price and marketability changes as the close approaches. We introduce the Threshold Auction Probability (TAP) score, defined as the probability that the NYSE closing auction price will meet or exceed a specified threshold, expressed as a percentage of the current market reference price. TAP is modeled as a function of two variables: (1) the threshold percentage relative to the reference price, and (2) the seconds remaining until the auction. It is calculated using historical data to capture the proportion of instances in which the market reached or exceeded the threshold.
The advantage of TAP is its ability to quantify the likelihood that the closing auction price will be above or below a given limit price for closing auction limit orders (D-Orders and LOCs), treating that limit price as the threshold. Because TAP can be modeled in advance, it can be applied to an order at the time of entry. Another advantage of TAP is its simplicity: it relies only on reference price, time, and the closing auction price but not trade-level data, making it a clean, unbiased measure. A limitation is that it does not account for order priority. While TAP is not the exact probability that an order will execute in the auction, it serves as a close proxy.
Using this approach, each order receives a TAP score between 0% and 100%, based on three factors: (1) time of entry, (2) the ratio of its limit price to the current reference price, and (3) order side. Buy orders take the TAP score directly, while sell orders use one minus the TAP score. All MOC orders are assigned a TAP score of 100%. Historical probabilities are calculated using data from all S&P 500 securities listed on the NYSE over the past year.
Below, we present examples of TAP scores to illustrate how the metric captures marketability. For instance, an order with a limit price at 99.8% of the reference price at 3:00 p.m. has a 79% probability that the closing auction price will meet or exceed that level. As time passes, an order at a 99.8% limit becomes increasingly likely to fall below the final auction price. TAP provides a clear, quantitative way to express how price and time jointly determine marketability in the closing auction.
Figure 4: Threshold Auction Probability
Figure 4: This metric uses NYSE-listed S&P 500 securities from September 2024 through August 2025. At each second, for each security and trading day, we evaluate reference price thresholds ranging from 99% to 101%. For each threshold, we calculate the proportion of instances the closing auction price meets or exceeds it. The chart shows the average of these proportions across all days and securities, illustrating the probability that a given limit price will be met at the close.
This analysis evaluates order marketability at the time of entry, excluding outlier activity but without adjusting for cancellations. For D-Orders, the limit price is expressed as a percentage of the reference price and serves as the TAP threshold. Buy orders use the TAP score directly, while sell orders use one minus the TAP score. Overall, D-Orders tend to be submitted at more marketable prices than LOC orders.
Figure 5: Threshold Auction Probability by Order Type
Figure 5: The dataset includes orders from NYSE-listed S&P 500 securities submitted between October 2024 and September 2025. TAP scores are calculated using data available at the time of order entry. As a result, the dataset includes orders that were later canceled.
The final chart shows how order marketability trends during the last 10 minutes of trading. The light blue line reflects average marketability at each moment, while the black line captures cumulative marketability, weighted by volume. Both lines indicate a decline, suggesting that traders submitting orders later in the session tend to use less marketable prices. This pattern may reflect efforts to minimize market impact by delaying order entry.
Figure 6: Threshold Auction Probability in the 10 Minutes Before the Closing Auction
Figure 6: This chart uses the same dataset as Figure 5, based on NYSE-listed S&P 500 securities from October 2024 through September 2025. The light blue line shows a rolling 15-second average TAP score by second, with each second’s value volume-weighted and all 15 seconds equally weighted. The black line represents cumulative TAP, weighted by order volume. Together, the lines illustrate how marketability declines as the close approaches, with later orders generally submitted at less marketable prices.
We examined the NYSE closing auction through the lens of order timing and marketability. Following the widespread adoption of third-party OMS technology by Floor brokers earlier this year, we observed a notable increase in very late D-Order submissions. To assess marketability at the time of entry, we introduced the TAP score to capture how price and timing jointly influence execution likelihood. Our analysis shows that D-Orders are generally more marketable than LOCs, though marketability tends to decline in the final minutes of trading as participants delay entry to manage price impact.