What algorithmic trading did to the retail investor, and what the next decade of paper trading asks of regulators

When the New York Stock Exchange opened for business on the morning of 14 July 2026, the order book on the floor was, as it has been for years, almost entirely silent. What matters now happens a few feet away in the data centre in Mahwah, New Jersey, where servers trade against each other in microseconds, and at terminals in Kansas City, London and Singapore, where retail investors tap a phone screen to place orders routed through a chain of intermediaries that none of them ever sees. Three decades into the transition from open-outcry to electronic markets, a new peer-reviewed survey of that shift lands at a moment when regulators in Washington and Brussels are again arguing about how much of the price a retail investor pays is being set by algorithms they did not hire and cannot audit.

The paper, published in Phys.org's coverage of an academic literature review on 14 July 2026, traces the structural migration from human-centered trading floors to electronically-mediated markets and asks a deceptively simple question: what has the change actually done to the way ordinary investors participate in, and access, financial markets? The answer, in plain terms, is that the cost of trading has collapsed, the speed of execution has multiplied by orders of magnitude, and the share of investor activity taking place away from any physical exchange has grown to the point where the venue is now an abstraction even for professionals.

What the literature says actually changed

The headline finding is the one most readers will already suspect: speed, scale and access have all widened. Electronic trading has stripped out the friction that used to sit between an investor's decision and the market's response, and it has done so on a curve that bends every quarter. But the study is more useful for what it pushes back on. It notes that the dominant story of "democratisation", that electronic trading gave retail investors a seat at the table they did not previously have, understates two things at once. It understates how much of the price formation on which retail orders now execute is performed by professional algorithms with no retail client in the loop, and it understates how much of the gain retail investors have experienced has been siphoned off in the form of payment for order flow, latency arbitrage, and the widening of the spread between quoted and executed prices.

The other non-trivial finding is about access. The authors document a generational widening of who can trade, with U.S. retail brokerage account numbers rising into the tens of millions over the period under review, and a parallel deepening of what those accounts can do: options, fractional shares, crypto-linked products and 24-hour markets that did not exist in the human-centred era. The same infrastructure that widened the door also widened the menu of ways a self-directed investor can lose money in circumstances they were never trained to read.

The counter-narrative the brokers tell

The industry line, repeated in lobby filings and earnings calls by retail brokerages and exchange operators, is that the migration has been a net good: tighter spreads, faster fills, lower explicit commissions, and an explosion of educational content that lets newcomers learn before they put capital at risk. That is not wrong, exactly. It is also incomplete. The brokers have benefited from the same migration, and the data on payment for order flow, the practice by which retail brokerages route customer orders to wholesale market makers in exchange for a share of the spread, shows that the firm capturing the residual in the trade is rarely the firm whose customer initiated the trade.

The plausible counter-read is that the next decade will look more like the last decade of search-engine regulation than like the last decade of bank regulation. The technology is not going to be reined in; it is going to be audited. That implies less time spent arguing about whether algorithmic trading should exist and more time spent arguing about what an investor is owed when the price they receive is, in some measurable sense, set by software they did not choose and cannot inspect.

A structural read in plain prose

What the literature is documenting is a transfer of pricing power from a layer that was visible, if slow (the human specialist on the exchange floor), to a layer that is invisible, fast and owned by a small number of firms. The retail investor did not lose access; they lost adjacency to the price-setting process. They are still permitted to trade. They are merely no longer present, in any operational sense, where the price is made. That is a structural change of the kind that typically takes a generation to register as a political fact rather than a technical one.

The same dynamic is visible in other sectors where a technological layer accreted between the user and the underlying asset. Cloud computing moved the customer's data out of their own server room. Social platforms moved the audience out of the publisher's owned-and-operated site. Electronic trading moved the price out of the room where the customer could see it being made. Each transition delivered real efficiency gains and real consumer surplus on the headline metric. Each also produced a slower-moving debate, still ongoing in each case, about what the customer is owed when they can no longer inspect the layer that determines their outcome.

What to watch before the next decade turns

Three policy dates are more informative than the next quarter's earnings. In Washington, the Securities and Exchange Commission's pending review of best-execution standards and the post-2020 rules on order competition will set the outer boundary of how much routing discretion a brokerage may exercise on a retail customer's order. In Brussels, the Markets in Crypto-Assets regulation and the related retail-trading disclosures will determine whether the same friction-stripping that happened in U.S. equities repeats in tokenised products with even fewer legacy safeguards. In both jurisdictions, the auditor's question, what was the realised price, and what would it have been on a competing venue, is the one that will define the next decade of retail investor protection.

What the literature does not settle, and what the sources do not yet permit a confident read on, is whether the widening of the gap between quoted and executed price is a transitional artefact of an electronic market still learning to price itself or a structural feature of a market whose pricing layer is now owned by a handful of firms. The next round of regulatory filings and academic replication will tell. Until then, the retail investor is faster, cheaper and more informed than at any point in the history of the trade, and is also, by the same shift, further from the price.

Desk note: this article reads the new study as a structural document, not a market call. Monexus frames the transition as a transfer of pricing power rather than a debate over whether electronic trading should exist.