Fractional Shares, Odd Lots, and the Overhaul of U.S. Market Data: What Data Teams Need to Know

If you work with real-time or historical U.S. equity data, including fractional shares, there are significant changes to plan for in 2026.

The SEC is in the middle of a multi-phase market structure reform that will change how trades are quoted, reported, and disseminated across the consolidated tape. Two of the three major components are already live, with the third going into effect in May.

At the center of it all is a concept that's become commonplace on retail brokerage platforms but has remained invisible in official market data: the fractional share.

What Is a Fractional Share, and Why Did It Get Complicated?

A fractional share is ownership of less than one full share of stock. Buying 0.5 shares of a $3,000 stock is economically straightforward. The plumbing behind reporting it accurately was not.

For most of market history, trade-reporting infrastructure was built around whole numbers. Trades were executed in clean round lots, reported in integers, and that was sufficient. As retail platforms began offering fractional investing (e.g., allowing investors to put $50 into a stock regardless of its price), actual transaction quantities like 1.257 shares or 0.083 shares became common. The official tape, however, kept rounding them down or truncating them entirely.

The result was a quiet but persistent distortion: Reported trading volumes in high-priced names were systematically understated, and a growing slice of real market activity was invisible to anyone analyzing the consolidated data.

The Three-Part Fix—And What's Already Changed

The SEC's reform addresses the distortion through three interconnected changes, each tackling a different layer of the problem.

1. Dynamic round lots (live since November 2025)

The traditional 100-share round lot was calibrated for a market where most stocks traded in single or double digits. Applied to a stock priced at $5,000 per share, a round lot represents $500,000 in notional value—and most real orders get classified as odd lots and excluded from the NBBO.

The new tiered system scales the round lot threshold to price:

• 100 shares for stocks under $250

• 40 shares for stocks priced $250–$999.99

• 10 shares for stocks priced $1,000–$9,999.99

• 1 share for stocks at $10,000 or above

For data teams, this means the composition of what counts as a "round-lot" trade (and therefore what feeds into NBBO calculations) has materially changed for high-priced equities. Liquidity metrics, spread calculations, and any model calibrated around the old 100-share threshold may need to be revisited.

2. Fractional share reporting to six decimal places (live since February 2026)

This is the core fractional shares mandate. The System Information Processors (SIP) that aggregate and disseminate consolidated market data are now required to publish fractional trade quantities up to six decimal places. A trade of 2.375418 shares is reported as exactly that, not as two shares.

For anyone consuming trade-level data, this has two immediate implications.

• First, volume figures for high-priced, heavily fractional-traded names are now more accurate. If your analytics rely on trade-reported volume as a proxy for participation or liquidity, the numbers just got better.

• Second, the data schema itself has changed; if your ingestion pipeline wasn't built to handle six-decimal quantity fields, this is a breaking change that needs attention now.

3. The best odd lot bid and offer—BOLO (going live May 2026)

Even after the round-lot changes, a substantial amount of trading continues to happen in odd-lot sizes, particularly from retail participants and algorithmic strategies executing precise fractional weightings. Until now, those quotes have been completely absent from the consolidated tape.

The BOLO changes that. It's a new data element—distinct from the NBBO —that surfaces:

• The highest-priced odd-lot bid sitting above the National Best Bid

• The lowest-priced odd-lot offer sitting below the National Best Offer

To be clear about what the BOLO is and isn't: It does not replace or alter the NBBO calculation. It runs alongside it as a separate signal. But for data teams building execution analytics, transaction cost analysis, or spread modeling, the BOLO will become an important input. It underlines a layer of competitive pricing that has always existed but was never formally captured in consolidated data.

The Data Infrastructure Ask

Collectively, these changes expand both the volume and precision of market data flowing through the tape:

• Six-decimal fractional quantities.

• A new BOLO data element requiring its own ingestion and normalization logic.

• Revised round-lot classifications that affect how you categorize and analyze historical and real-time quote data.

None of this is insurmountable, but it does require intentional preparation, particularly for teams whose data pipelines were built against older SIP specifications.

FactSet's real-time quotes and trade data is built to keep pace with these shifts by delivering fractional quantities at six-decimal precision and incorporating the BOLO as a distinct data element ahead of the May rollout. As the consolidated tape evolves, your analytics, execution models, and liquidity metrics stay grounded in accurate, complete data, not an increasingly outdated snapshot of how U.S. equity markets used to work.

This blog post is for informational purposes only. The information contained in this blog post is not legal, tax, or investment advice. FactSet does not endorse or recommend any investments and assumes no liability for any consequence relating directly or indirectly to any action or inaction taken based on the information contained in this article.