The data beneath the surface: what the first month of Reg NMS changes actually revealed

First published by Rebellion Research, 7 January 2026

Dr Elliot Banks, Chief Product Officer, BMLL

The US equity market is undergoing some of the largest structural upheavals in decades, with changes to market data, regulations, and trading methods. I recently met with my colleague Rob Laible, Head of Americas, BMLL, to review the state of the market one month after the latest SEC changes went live, and to discuss other major changes to the market, including 24-hour trading and hidden liquidity. As we head into 2026, understanding the changing liquidity landscape will be important for all market participants, whether managing risk, trading efficiently, or generating alpha.

A year of change

November 2025 introduced the first of several major regulatory changes for US equity markets, impacting how participants trade and view market data for all listed equities - one of these being changes to lot sizes. Rather than almost all stocks having 100 shares corresponding to a round lot, lot sizes will change based on traded price, with 4 different lot sizes (1, 10, 40, 100).

In addition to the November changes, there are a series of updates over the next twelve months, all of which involve significant structural adjustments.

The next change arrives in February 2026, when FINRA trade reporting facilities (TRFs) begin reporting fractional shares. Until now, if a trade was 1.257 shares, the TRF would round it down to the nearest whole number. This created a gap in the data, where we could not accurately see the granularity of fractional shares. From February, TRFs will report fractional shares with up to six decimal places. This may seem minor, but firms need to ensure they are properly equipped to deal with shares no longer being whole numbers, and be able to properly account for that when analysing market structure.

Following that, the SIP will begin reporting the Odd Lot NBBO in May 2026, alongside the official NBBO. Finally, there are major tick size changes coming into place in November 2026 (which was delayed from November 2025). This will transition the US equity market to a more dynamic tick size regime, where tick sizes can change based on spread. We have seen dynamic tick size regimes in other regions, most notably Japan and Europe.

Round lot changes

We first took a look at the impact of the round lot size changes. Under the new lot size rules, the size of a round lot will change from the standard 100 (or 1 for a very small number of stocks), based on the average closing price of the security:

  • < $250 - 100 shares

  • $250 - $1000 - 40 shares

  • $1000 - $10000 - 10 shares

  • > $10000 - 1 share

While the actual number of stocks affected by this change is small (c. 250), stocks affected include highly liquid equities, such as Tesla, Meta, and Taiwan Semiconductor.

Following the change, we examined its impact on liquidity. We see that the percentage of trading executed as odd lots has dropped, from approximately 20% to between 15% and 16%, as shown in Figure 1. This makes sense - orders that were previously classified as odd lots are now round lots.

It is important to note that this is not the only factor that can affect levels of odd lot trades - the dip in September is due to a ‘triple witching’ event, where a large percentage of volume is traded through auctions.

Figure 1: Percentage of Odd Lot trading for US equities

Next, we turn to the bid-ask spread. For those 40-share lot size stocks, we find the NBBO spread from around 80 cents to 40 cents.

Interestingly, however, the spread when including odd lots remained unchanged, as shown in Figure 2. This suggests the changes have not affected the way that market makers and liquidity providers provide liquidity to the stock. Instead, the mechanics of a 40-share NBBO are making SIP liquidity appear to be more.

Figure 2: Impact of Lot size changes on the bid-ask spread

For quants running backtests or best execution analysis using the NBBO, understanding this is important. Otherwise, you might falsely interpret November 2025 as a moment where liquidity has changed. It did not. The underlying liquidity remained constant, but the reporting metric changed. If your models do not account for this structural shift, your historical backtests will be flawed.

The mechanics of overnight liquidity

One of the major stories of 2025 has been the growth of overnight trading. This has been seen through both exchanges (which offer extended hours trading but not full 24 hours due to technical constraints) and overnight ATSs such as Blue Ocean ATS, OTC Markets’ MOON ATS, and Bruce Markets’ BOSS ATS. Looking at FINRA data, we can see that, for the ATS venues, Blue Ocean, as the first mover, is the current market leader, illustrated in Figure 3. These trading volumes are driven by Asian trading accessing US stocks, and in November, South Korea opened up retail trading to all 3 ATSs, following a period in which they could not trade via these ATSs. It will be important to watch this growth of overnight trading, in order to determine the liquidity impact.

Figure 3: Shares traded on MOON, BOSS & BLUE

As a fully transparent ATS, we can dive into the trade-by-trade details of Blue Ocean. Firstly, we see a clear growth in trading volumes following the opening of South Korea, as shown in Figure 4. Furthermore, we see a clear ‘volume smile’ between 8:00 pm and 4:00 am (the hours when Blue Ocean is open). This curve has solidified significantly since the Korean retail market reopened access to these venues. The correlation suggests that this liquidity is structural and driven by specific geographic retail engagement rather than sporadic trading.

Figure 4: USD traded on Blue Ocean
Figure 5: Blue Ocean Volume Smile

Interestingly, we see heavy overnight trading in ADRs, even when the underlying stock for that ADR is trading in the local market. It shows that traders are still looking to trade in the US markets and access that liquidity pool. Understanding this further, and the drivers of this liquidity, is vital for anyone looking to access these markets

Hidden liquidity and execution quality

Finally, we utilised our Trades Plus dataset to examine execution styles. By melding SIP data with direct feed depth data, we can classify trades beyond simple reporting codes. One particularly important area is understanding how much on-exchange liquidity is traded via hidden orders, rather than pre-trade transparent orders. We see a clear variation between venues, suggesting that firms should leverage this information to better tune routing strategies.

Figure 6: Percentage of exchange volume that is hidden

Looking ahead to 2026

As we move through the next year, the complexity of US market structure will only increase. The introduction of six-decimal reporting and dynamic tick sizes will require a more granular approach to data analysis. We cannot simply rely on top-of-book data anymore. The story is in the depth, the odd lots, and the overnight curves.

As these structural changes take hold, a more granular approach to data will be essential. Leveraging Level 3 data enables market participants to understand the underlying mechanics of how tick sizes, odd lots and market data evolve. This level of detail supports more accurate analysis, whether for alpha generation, trading or risk management, and enables a smoother transition into the new environment of dynamic round lots and varying tick sizes.

Watch the full webinar recording for deeper insight into how Level 3 data supports a smoother transition through these changes.

Contact our team to schedule a demo and discuss your needs in more detail.