Electronic trading

Become a sponsor »

Electronic trading uses computer systems to bring buyers and sellers together in securities and derivatives markets. Orders are entered, routed, matched and executed with little or no manual intervention, allowing participants to establish prices and complete transactions.

History and background

Stocks

The U.S. Securities and Exchange Commission recommended a computerized, real-time substitute for printed over-the-counter quotation sheets in its 1963 Special Study of the Securities Markets. The National Association of Securities Dealers later built a centralized system, and the Nasdaq Stock Market (NASDAQ) quotation network went live in February 1971 with about 3,000 securities. It initially automated the distribution of dealer quotes rather than the entire execution process.

The Nasdaq Stock Market (NASDAQ) added execution functions in stages. Its Small Order Execution System automated trades against market-maker quotes, while SelectNet supported electronic negotiation and execution of larger orders; use of the small-order system became mandatory after the 1987 market break exposed weaknesses in telephone-based trading. These changes distinguished electronic quotation from automatic execution and helped move the dealer market toward an integrated screen-based workflow.

Electronic communications networks expanded competition among U.S. trading venues in the 1990s by displaying and matching subscriber orders outside traditional exchange floors. The SEC's 2005 Regulation NMS then required trading centers to maintain procedures intended to prevent executions at prices inferior to protected, immediately accessible quotations, imposed access standards and updated national market-data rules. The framework linked electronic venues through price-protection and data obligations while leaving execution distributed among competing markets.

Emergence of Electronic Communication Networks and the Disruption of Floor Monopolies

Before the widespread adoption of electronic matching engines, U.S. equity execution was split into two primary paradigms: the continuous auction floor and the over-the-counter dealer market. On the NYSE floor, a single specialist held exclusive custody of the manual order book for each assigned security, managing order execution while trading for proprietary dealer accounts. On Nasdaq, broker-dealers communicated via telephone to negotiate transactions, displaying indicative quotations across proprietary terminal screens without central limit order interaction2. This manual architecture resulted in wider transaction spreads, significant execution delays, and asymmetric informational advantages for floor members and over-the-counter dealers.

These structural inefficiencies were documented empirically in 1994 by William Christie and Paul Schultz, whose research revealed that Nasdaq market makers routinely avoided odd-eighth quotes (such as 1/8, 3/8, 5/8, and 7/8)2. This practice artificially inflated minimum bid-ask spreads to 25 cents ($0.25), pointing to tacit collusion that cost retail and institutional investors billions of dollars annually. ECNs challenged this paradigm by offering an automated, software-driven limit order book. In an ECN, buy and sell orders are routed into an electronic matching engine that displays transparent depth of book and matches incoming orders automatically based on price-time priority, removing the specialist and the manual market maker from the execution chain.

Instinet, founded in 1969, was the pioneer of off-exchange electronic execution. Originally conceived as a communication network allowing institutional fund managers to trade blocks of equity directly with one another without dealer intervention, Instinet evolved into an institutional platform that bypassed specialist markups1. However, Instinet operated primarily as an exclusive venue for institutions and broker-dealers, meaning the broader public remained isolated from its liquidity pool.

The democratization of automated order matching began in 1996 when Jeff Citron and Joshua Levine founded Island ECN at Datek Securities. Levine built Island using custom-engineered software designed for execution speed, determinism, and minimal overhead16. Island ran directly in random-access memory, processing orders in sub-second intervals and publicly broadcasting its entire limit order book via the internet. By providing direct execution to non-member day traders and active individual investors, Island bypassed traditional exchange specialists and captured substantial market share in actively traded Nasdaq stocks.

Archipelago, launched in 1996 by Gerald Putnam and Stuart Townsend, approached the market through automated smart order routing. Rather than relying solely on internally posted liquidity, Archipelago evaluated internal resting orders alongside the quotes of external ECNs and public exchanges, instantly routing outbound child orders whenever better prices appeared on outside books. Concurrently, proprietary consortia formed alternative platforms: Brut was launched in 1998 by a consortium of investment banks including Merrill Lynch, Morgan Stanley, and Goldman Sachs, while Knight Capital Group established Attain to automate over-the-counter order handling.

To solve the liquidity bootstrapping challenge common to nascent electronic books, Island introduced an asymmetric pricing mechanism that altered the financial foundations of order routing: the maker-taker fee model. Invented by Joshua Levine in 1997, the system paid a cash rebate to market participants who entered non-marketable passive limit orders that rested on the book and provided liquidity ("makers"). To fund this rebate, the platform charged a slightly higher access fee to participants submitting marketable orders that immediately executed against resting interest ("takers"). The difference between the taker fee and the maker rebate provided operational margin to the platform11. By subsidizing liquidity provision, maker-taker pricing incentivized proprietary quantitative trading firms to continuously quote narrow, two-sided limit orders, creating an environment that undermined the economics of human floor market making.

Structural Evolution of Equity Electronic Trading, Alternative Trading Systems, and Exchange Consolidation

The transformation of United States equity markets from centralized, physical trading floors to a dispersed network of high-speed electronic venues represents the most consequential structural realignment in modern capital market history. For over two centuries, equity execution relied on manual dealer intermediation—specialists on the New York Stock Exchange (NYSE) exercising localized monopoly control over manual order books, and syndicated telephone dealers on Nasdaq maintaining wide bid-ask spreads. The modern era of equity execution emerged when independent Electronic Communication Networks (ECNs) introduced automated central limit order books (CLOBs) that matched orders deterministically based on strict price-time priority, bypassing human intermediaries and dismantling physical floor cartels.

Federal securities regulation catalyzed and codified this technological shift. In 1997, the Securities and Exchange Commission (SEC) enacted the Order Handling Rules to eliminate non-public pricing tiers, forcing market makers to integrate public customer limit orders and ECN quotes directly into the national market. In 1998, Regulation ATS established a regulatory structure that lets electronic venues operate as broker-dealer execution systems without the regulatory burdens of national securities exchanges. The structural inflection point arrived with Regulation National Market System (Reg NMS) in 2005: Rule 611 (the Order Protection Rule) granted trade-through protection exclusively to automated "fast quotes," stripping manual floor specialist quotes of regulatory shelter and compelling legacy exchanges to either automate their workflows or lose order flow.

Faced with rapid market-share losses to low-latency platforms that used innovative maker-taker rebate schedules, legacy exchange operators demutualized, transitioned to for-profit corporations, and systematically acquired major ECNs. This dynamic sparked horizontal consolidation across the industry: the NYSE acquired Archipelago to launch NYSE Arca, Nasdaq absorbed Brut and the INET platform from Instinet, and Bats Global Markets acquired Direct Edge before ultimately merging into Cboe Global Markets. To retain their listing franchises and handle complex block volume during this migration, floor exchanges pioneered hybrid trading architectures—such as the NYSE Hybrid Market and the Cboe Hybrid Trading System—that paired automated electronic order books with designated market makers and floor broker reserve interest.

Emergence of Electronic Communication Networks and the Disruption of Floor Monopolies

Before the widespread adoption of electronic matching engines, U.S. equity execution was split into two primary paradigms: the continuous auction floor and the over-the-counter dealer market. On the NYSE floor, a single specialist held exclusive custody of the manual order book for each assigned security, managing order execution while trading for proprietary dealer accounts. On Nasdaq, broker-dealers communicated via telephone to negotiate transactions, displaying indicative quotations across proprietary terminal screens without central limit order interaction. This manual architecture resulted in wider transaction spreads, significant execution delays, and asymmetric informational advantages for floor members and over-the-counter dealers.

William Christie and Paul Schultz documented these structural inefficiencies in 1994, and their research revealed that Nasdaq market makers routinely avoided odd-eighth quotes (such as 1/8, 3/8, 5/8, and 7/8). This practice artificially inflated minimum bid-ask spreads to 25 cents ($0.25), pointing to tacit collusion that cost retail and institutional investors billions of dollars annually. ECNs challenged this paradigm by offering an automated, software-driven limit order book. In an ECN, buy and sell orders are routed into an electronic matching engine that displays transparent depth of book and matches incoming orders automatically based on price-time priority, removing the specialist and the manual market maker from the execution chain.

Futures

Electronic futures markets replaced floor bids and offers with orders entered at terminals and matched by a host computer. The Chicago Mercantile Exchange (CME) Globex system began operating in 1992, New York Mercantile Exchange (NYMEX) ACCESS followed in 1993, and Chicago Board of Trade (CBOT) Project A began operating in 1994. Early systems extended trading hours and broadened access while operating alongside open-outcry sessions.

By fiscal 1997, exchanges were using electronic systems for more than overnight access. The Chicago Mercantile Exchange (CME) routed small orders in E-mini S&P 500 futures and option contracts exclusively through Globex, while larger orders remained subject to open-outcry procedures. The Commodity Futures Trading Commission required audit trails and recordkeeping controls when it approved related electronic-market arrangements.

Competition accelerated the European transition. After the electronic Deutsche Terminbörse gained share in the Bund contract, the London International Financial Futures and Options Exchange (LIFFE) developed Liffe Connect in 1998 and allowed third-party trading software to connect to it. Electronic access helped LIFFE rebuild its euro money-market business, and the exchange closed its open-outcry pits in November 2000.

Wagner patent and INTEX

U.S. Patent No. 4,903,201, titled "Automated futures trading exchange," names Susan W. Wagner as the inventor. World Energy Exchange Corp. filed the application on November 3, 1983, and the U.S. Patent and Trademark Office issued the patent on February 20, 1990. It described remote order-entry terminals connected to a central processor that matched bids and offers, confirmed executions and supported clearing, compliance and surveillance functions.

Electronic Trading Systems Corp. brought the principal infringement action on May 5, 1999, in the U.S. District Court for the Northern District of Texas, Dallas Division. It named the Chicago Mercantile Exchange (CME), Chicago Board of Trade (CBOT), New York Mercantile Exchange (NYMEX) and Cantor Fitzgerald L.P. as defendants and alleged that their electronic futures systems infringed the Wagner patent. A CME filing later said Cantor Fitzgerald had settled with Electronic Trading Systems for undisclosed consideration.

In April 2001, eSpeed, a Cantor Fitzgerald affiliate, purchased exclusive rights to the patent from Electronic Trading Systems for $1.75 million in cash and 24,334 shares of eSpeed Class A common stock valued at $500,000. The agreement also required contingent payments when the patent generated revenue. Electronic Trading Systems remained a party to later settlements and received agreed shares of licensing proceeds.

CME and CBOT resolved their claims in August 2002 without admitting liability. Each exchange agreed to pay eSpeed $15 million over five years, consisting of $5 million in 2002 and five annual payments of $2 million, for a total of $30 million. The settlements granted each exchange a paid-up patent license; the CME license was worldwide, nonexclusive and nontransferable for the life of the patent and generally barred CME from processing trades for an unlicensed third-party futures exchange. The agreement also covered OneChicago LLC, the single-stock-futures venture of CME, CBOT and the Chicago Board Options Exchange (CBOE).

The pending claim crossed CME's transition to public ownership. CME demutualized on November 13, 2000, while the case was pending, settled on August 26, 2002, and disclosed the settlement in its 2002 initial-public-offering prospectus. The filing recorded the $15 million obligation at a present value of $13.7 million, described the license's third-party-processing restriction as a risk and said the settlement did not admit infringement. It did not state that the offering caused the settlement.

CME also invoked an indemnification provision in its software license with Euronext-Paris, whose NSC software powered Globex. After disputing responsibility for the defense and settlement costs, the companies signed a December 23, 2002, agreement under which Euronext-Paris paid CME $7.5 million, equal to half of CME's Wagner settlement. That agreement resolved an indemnification dispute; it did not make Euronext-Paris a defendant in the patent action.

CBOT also resolved the patent claim before its proposed public offering. Its May 2005 Form S-1 said the exchange had paid $5 million in September 2002 and $2 million in each of September 2003 and September 2004, with three annual $2 million payments remaining. The same filing described CBOT's April 22, 2005, demutualization and proposed initial public offering and warned generally that patent claims could require licenses, redesigns or discontinued services.

The NYMEX branch of the case moved from Texas to the Southern District of New York after the court granted NYMEX's severance-and-transfer motion on September 15, 2000. NYMEX demutualized on November 17, 2000, while the claim remained pending, and the court added eSpeed as a plaintiff on August 1, 2001. After an April 2002 claim-construction hearing, the court interpreted several patent elements more broadly than NYMEX had proposed, which NYMEX told investors could narrow its noninfringement arguments. The parties remained in discovery at June 30, 2002, and no final ruling on infringement or validity followed before settlement.

eSpeed and Electronic Trading Systems settled with NYMEX in December 2003 without an admission of liability. NYMEX agreed to four annual payments of $2 million, totaling $8 million. Its November 2006 initial public offering registration statement disclosed the resolved action, the three payments already made, and the final $2 million payment due in December 2006; it said the obligation was fully reserved and paired the disclosure with a general risk factor on third-party intellectual-property claims. The filing did not say that the offering caused the settlement.

eSpeed also licensed the Wagner patent to Intercontinental Exchange (ICE) on March 29, 2002. ICE agreed to a $2 million annual royalty, of which 12 percent went to Electronic Trading Systems, plus either 10 cents for each contract submitted for trading or 20 cents for each contract in a matched trade. The agreement ran through the patent's life.

The Wagner claims ended through licenses and settlements rather than a reported merits judgment. The patent expired on February 20, 2007, after which eSpeed said it would receive no further Wagner-patent revenue. eSpeed's separate suit against BrokerTec USA LLC and related entities concerned U.S. Patent No. 6,560,580, not the Wagner patent; BrokerTec therefore was not a verified Wagner-patent defendant or licensee.

Commodity Futures Trading Commission records identify the Bermuda venue as the International Futures Exchange (Bermuda) Ltd. (INTEX). The regulator issued a no-action letter on September 5, 1985, covering the exchange's Financial News Composite Index futures contract.

INTEX commenced operations on October 25, 1984. Its system aligned bids and offers, verified execution prices and displayed the order book to traders and brokers, who could enter orders from remote terminals and print transaction records in their offices or homes.

Merrill Lynch had sponsored the venture-capital group that financed INTEX's formation. The exchange used time/price priority and a central limit order file, and it attempted to trade gold contracts after the Commodity Exchange (COMEX) had closed. At its peak, 30 terminals in Chicago, New York and London traded 200 to 300 contracts a day. INTEX did not establish sufficient liquidity and ceased operating in 1988 after selling its software to other start-up ventures.

Options

The SEC approved the International Securities Exchange (ISE) as a registered exchange in February 2000, the first new U.S. securities exchange registration since the Chicago Board Options Exchange (CBOE) in 1973. ISE was designed as a fully electronic market for U.S. equity options, combining screen-based trading with auction-market principles and without a physical trading floor.

The Chicago Board Options Exchange (CBOE) pursued a hybrid model rather than an immediate replacement of its floor. Its proposed Hybrid System combined automatic executions of eligible electronic orders with open-outcry trading and auction-market principles, allowing floor participants to submit electronic quotes while preserving in-person negotiation. The SEC published the proposal in 2003 as incumbent options exchanges responded to competition from all-electronic venues.

Cash bonds

Cash bond trading adopted electronic execution later than exchange-listed equities and derivatives. U.S. Treasury trades had traditionally been arranged by telephone, but electronic communication networks began gaining volume in the late 1990s; electronic systems handled about 0.5 percent of Treasury volume in 1998 and 2 percent to 4 percent by fall 1999. The number of systems offering electronic transaction services for U.S. government and private debt rose from 11 in 1997 to 39 by November 1999.

Fixed-income platforms developed around both interdealer trading and dealer-to-customer execution. In 1999, an SEC review reported that electronic systems accounted for a much smaller share of bond activity than equity activity, while the number of fixed-income systems was increasing rapidly. Electronic trading reduced some operating and transaction costs but also raised questions about fragmentation, access and regulatory treatment.

Post-trade transparency expanded through the Trade Reporting and Compliance Engine, which requires Financial Industry Regulatory Authority member broker-dealers to report eligible over-the-counter fixed-income transactions under SEC-approved rules. TRACE is a reporting and data system rather than an execution venue, but its transaction records give market participants and regulators information about activity in markets that historically relied on bilateral dealer quotes.

Market operation and usage

An electronic venue accepts orders through participant interfaces or application programming connections, validates their terms and passes eligible orders to a match engine. Most exchange algorithms rank better-priced orders first and then use time/price priority for orders at the same price, although some venues use pro rata or parity models. Market, limit, hidden and reserve instructions let participants trade immediately, rest interest in a book or reduce displayed size, while automated routing can seek protected prices across venues.

Electronic fixed-income markets use several protocols because debt instruments differ in issue size, trading frequency and dealer participation. Request-for-quote systems solicit prices from selected dealers, while electronic communication networks display executable interest and permit direct matching; the protocols trade off immediacy and anonymity. Research on corporate-bond alternative trading systems found that dealer participation was associated with lower customer transaction costs, but electronic activity remained concentrated in particular bonds and trade sizes.

Electronic trading supports continuous or extended-hours access, rapid order modification and systematic execution strategies. It can strengthen price discovery by incorporating orders and trades quickly, but its effect on liquidity depends on market design, participant behavior and conditions; a Federal Reserve study of the EBS foreign-exchange venue found that limit orders gained information share over a decade as price discovery became faster.

Key users

Applicable users include retail investors, institutional investors, market makers and broker-dealers. Electronic communication networks let subscribers enter limit orders, view available interest, negotiate anonymously in some systems and use reserve-size instructions; access may be direct for professional firms or intermediated through a broker for customers.

In Treasury markets, banks, proprietary trading firms, hedge funds and asset managers use electronic venues for dealing, automated market making, execution and hedging. Interdealer cash Treasury trading had become almost fully electronic by 2015, and automated firms had gained direct platform access, while a substantial part of dealer-to-customer activity still occurred by telephone.

Risk management

Electronic-market risk management begins before an order reaches a venue. SEC Rule 15c3-5 requires broker-dealers with direct or sponsored market access to maintain documented controls and supervisory procedures designed to limit financial exposure and enforce regulatory requirements. Common controls include credit and capital thresholds, order-size and price checks, permissioning, message-rate limits, duplicate-order detection, cancel-on-disconnect functions and the ability to stop trading.

Operational resilience also depends on redundant infrastructure, controlled software releases, capacity testing, cybersecurity, incident response and recovery procedures. Failures at technology providers can spread across firms: a January 2023 ransomware attack on Ion Markets disrupted critical processes at derivatives-market participants for several weeks. Dependence on common vendors and cloud services can therefore concentrate operational risk even when trading venues and users are geographically dispersed.

Controversies

The May 6, 2010 flash crash demonstrated how quickly stress could cross electronic futures and equity markets. The E-mini S&P 500 contract and the SPDR S&P 500 exchange-traded fund fell about 5 percent in five minutes and then recovered, while hundreds of securities briefly traded at extreme prices. Regulatory analysis found that displayed buy-side depth in the E-mini had already eroded sharply before the fastest decline, and the episode prompted scrutiny of automated order flow, fragmented routing and market-wide volatility controls.

The joint regulatory review did not identify one culprit. It described high volatility, declining liquidity, a large algorithmic sale of 75,000 E-mini contracts and the transmission of disruption among linked venues; regulators subsequently emphasized circuit breakers, limit-up and limit-down mechanisms and procedures intended to restore orderly trading. The event remains a central example of the debate over whether speed and interconnection can amplify a liquidity shock.

Knight Capital's August 1, 2012 incident showed the consequences of defective deployment and weak controls at a market participant. The SEC said the firm sent millions of erroneous orders after activating old code during a software release and lacked adequate safeguards and review procedures; Knight agreed to pay $12 million to settle the SEC's first enforcement action under the market-access rule. The case focused the controversy on governance and testing rather than on electronic execution alone.

See also

Related subjects include:

• [[futures]
• options
• price discovery
• liquidity
• risk management
• settlement
• hedging
• benchmark
• interest rate.