How the rise of off-exchange trading is a feature – not a bug – of our current equity market structure.
Market structure watchers have witnessed the share of off-exchange trading steadily increasing over the past few years. Whether a function of meme-stock related spikes, permanently higher levels of retail participation in our markets, the warping effects of very low (or high) priced shares, or increasing innovation in the ATS space, the ratio of off-exchange volume to total has continued it’s march ever higher. Headlines were made early this year when off-exchange trading surpassed on-exchange not just for a day but for a full month. Have we truly crossed over to the “dark side”?
Reasonable concerns have been raised as to whether this reflects some deeper issue with our markets. After all, if most volume is occurring off exchange, do the prices displayed on exchange really reflect the true levels of trading interest in the marketplace? Is the price really … the price?
As an aside, it is worth noting that the share of off-exchange volume varies significantly by symbol. Within the S&P 500 for example, off-exchange volume in January represented anywhere from 27% (NWS) to 61% (F) of total. Aggregate off-exchange volume also remains below 50% if you look only at popular indexes such as the S&P 500 or Russell 3000, or if you exclude sub-dollar stocks.[1] Whether this trend “feels” impactful to you probably depends on which stocks you have been trading…
Displayed quotations are critically important as they form the basis for all realized prices. Indeed, in our National Market System, the prices represented by the national best bid and offer (NBBO) reflect not just indications of prices where traders *might* trade, they set the pricing parameters for where trades *must* occur. So it is worth considering whether the NBBO is a true reflection of supply and demand in the marketplace, and whether trading that does not result from interaction with a lit quote impedes the price discovery process.
Displayed quotations also define spreads. Spreads are a key determinant of trading costs for liquidity takers and are also used as a scaling factor in measuring benchmark slippage across trading instruments (a benchmark for the benchmark, if you will). If spreads are artificially wide, then we get a skewed sense of what liquidity actually costs. Investors, brokers, exchanges, ATSs, and market participants of all types have spent fortunes, calories, and PhD theses to optimize performance within and relative to the spread. An NBBO that accurately reflects *true* prices and spreads is critical to our marketplace, and nothing I say here is intended to argue otherwise.
At some level, however, this ignores the limitations of what the NBBO or any displayed quote regime could hope to address. The NBBO does not and never did represent a full accounting of investor interest, especially for institutional investors. A “dark secret” of institutional trading is that the vast majority of “liquidity” is non-displayed – by design. Institutional investors do not necessarily provide their whole picture to a broker or exchange. Brokers do not broadcast full parent orders to the whole street (that is, if they want to receive the next order). Exchanges themselves provide hidden and pegged order types that do not directly contribute to the quote, acknowledging there are good reasons for traders not to reveal their full intent. The whole process of trading is essentially about balancing the risk of showing your hand with the risk of not getting done.
Market microstructure debates sometimes miss this point. Our marketplace incentivizes narrow spreads and accepts smaller quoted size as a tradeoff. Because of this, institutional traders in particular have and will continue to use a variety of “alternative” methods to communicate and discover liquidity: blotter-scraping; broker IOIs; dark pool pinging; a trusted high-touch sales trader (who still, by the way, handle a meaningful portion of order flow) Traders of all types will continue to use technical analysis to determine the “real” levels where size supply and demand might rest. The bids, offers and levels that many traders react to aren’t necessarily what they see “in the box”. That is not a symptom of a broken market structure but simply acknowledgement that market data is the answer to a different question.
Counterintuitively, the rise of off exchange trading might indicate that the NBBO is doing its job. Anyone who trades off exchange has accepted the prices set on the lit markets. (There is, perhaps, a parallel here with passive investing.) If those prices are truly out of line with investors’ views then investors can choose not to trade, or they can step in and display their intent and accept the risk that this will generate market impact.
Non-displayed trading exists to fill a need that interacting with lit quotes alone cannot meet. It’s worth observing that off exchange trading is driven by *both* ends of the investor size spectrum; retail investors, who tend to execute against wholesalers willing to improve the best published quote, and institutional investors who are often seeking greater size than may trade in an entire day (let alone displayed on the touch). “Fixing” off-exchange trading for one side risks making it worse for the other; we decry a “one-size-fits-all” market structure but then complain when one market structure doesn’t fit all sizes.
Ultimately, market structure enthusiasts will continue to debate how much off-exchange trading is too much and whether we have gone too far. Time will tell whether the share of off-exchange trading will continue its upward march. In the meantime, we will all do our best to continue navigating a fragmented and complex marketplace, while remembering that divergent opinions and mindsets are what makes a market in the first place.
Jack Miller is Head of Global Execution Services at Baird.
[1] January 2025. Source: Bloomberg
