SIGNUM
SIGNUM is Exegy's set of machine learning trading signals, sold to broker dealers, market makers, quantitative funds and other firms that trade US equities on low latency market data. Exegy, the St. Louis market data and trading technology company, generates the signals in its hardware accelerated ticker plant appliances in under a microsecond and appends them to the normalized quotes and order book updates its clients already receive.
Quote Vector gives the probability that the next national best bid or offer price will move up or down, on every quote for every US listed equity. Quote Fuse gives the probability that the next price change will come within 50 milliseconds. Liquidity Lamp flags reserve, or iceberg, orders resting at a displayed price on a given exchange. Firms set their own probability thresholds and use the signals to cut slippage in execution algorithms, route orders and tighten market making quotes. Exegy's own team built a predictive VWAP algorithm on them and publishes estimates of the price improvement the signals make possible.
Live signals arrive through the Exegy Client API, and firms can test them over the internet, by data center cross connect or with an appliance in their own test environment. Liquidity Lamp Summary data comes intraday or at the end of the day and is also distributed by CloudQuant, Exchange Data International and a Signum Bot on Symphony. Exegy offers ten days of sample data and quotes each service.
Capability Axes
Capability grades
15 of 15 axes rated · 9 graded A or B
The signals are what SIGNUM sells, and machine learning produces every one of them: direction and timing probabilities for the next quote change and a flag for reserve orders. The appliance that also normalizes and delivers market data is Exegy's separate core business. Remove the models and there is no SIGNUM product left.
SIGNUM informs and never trades. Quote Vector and Quote Fuse return probabilities for both bid and offer, Liquidity Lamp a flag, and Exegy tells clients to select the probability thresholds at which their own systems act. Applications subscribe to the symbols they want, starting from a snapshot. Order handling, risk limits and supervision of any algorithm that uses the signals stay with the trading firm.
Each output is precisely defined, down to the 50 millisecond window behind Quote Fuse, and Exegy has published analysis of its price change and liquidity forecasts. Firms can test the signals against observed market data with the ten day sample. Training methods and validation results for each model release are not published.
Exegy's own analysis estimates the net price improvement achievable with Quote Vector and measures how well Quote Fuse forecasts quote stability, figures it reports on a performance website. Those are Exegy's estimates rather than a named desk's realized results. CloudQuant, Exchange Data International and Symphony distribute the summary data, which shows how the product reaches buyers rather than what it earned them.
The models run on market data inside Exegy's appliances and send their output to every subscriber alike, so no single firm's trading data is described as an input. Whether a client's usage or telemetry feeds retraining, which history the models learned from and how long connection data is kept are not stated.
The signals are built from public market quotes and order book events, not from customer or consumer data. Exegy's corporate privacy policy covers the personal information it collects from contacts and clients. Terms for subscription records, usage data or client system metadata tied to a firm's signal feed sit in the service contract.
Exegy's material covers continuity rather than certification: its managed service team monitors every signal enabled appliance through the trading day, and delivery runs over cross connects, internet links or an appliance at the client. A security certification or attestation covering SIGNUM or Exegy's appliance operations is not named.
Exegy supplies market data and model signals as a technology service, separate from its order entry and execution products. The output is a probability, a flag or a historical summary, and trading policies, order handling and market obligations stay with the broker, exchange member or fund that runs the strategy.
Each signal targets a distinct market behavior, and sample files let firms compare predictions with what the market did. Exegy monitors signal delivery through the trading day. A formal model governance process, or a breakdown of accuracy by venue, instrument type or market condition, has not been described, on a product whose bias surface is market behavior rather than people.
A firm acts on the signals in its own strategy, and Exegy's legal and FAQ pages say nothing about who bears a loss when a forecast is wrong or how a disputed signal could be challenged. Sample data lets a buyer judge accuracy before signing, but any service remedy comes only from the individual order.
Exegy's own data science team built the models, and they run inside Exegy's appliances on normalized market data, so no outside model provider sits in the chain. CloudQuant, Symphony and Exchange Data International only distribute the output. Model versions and the full list of feed dependencies are not published.
Live signals arrive as fields in the normalized market data events of the Exegy Client API, so a trading application that already consumes Exegy data subscribes to symbols, takes a snapshot and then processes updates with new values on every tick. Liquidity Lamp Summary data reaches research and collaboration tools through CloudQuant, Exchange Data International and the Signum Bot on Symphony.
The signals run on Exegy's hardware accelerated appliances, which a firm can place in its own test environment or reach by data center cross connect or internet, and production delivery is monitored by Exegy's managed service team. Summary data comes through third party channels or as a file by download or cloud transfer, so the firm chooses where the signals land.
Exegy sells the real time signals directly and offers ten days of sample data, by download or cloud transfer, and free sample data with whitepapers after registration. Firms can arrange a live trial over the internet, by cross connect or on an appliance. Summary data is also bought through CloudQuant and other distributors. No rate, metering unit or package price is published for the signals.
The signals serve firms that trade US equities algorithmically: broker dealers tuning execution algorithms and order routing, market makers tightening quotes and improving queue position, and quantitative and buy side teams researching on the summary data. Research desks can take delayed Liquidity Lamp summaries through CloudQuant or Symphony. Named firms using the live signals in production are not given.
Alternatives to SIGNUM
Other products that do the same kind of work as SIGNUM, and what each one does.
Qraft Technologies is a Seoul company, founded in 2016, that builds AI investment models for asset managers, insurers, securities firms and banks.
Stockpulse turns news and online discussions into market sentiment data for analysts, quantitative funds, data platforms and exchange surveillance teams.
bondIT sells AI fixed income software for portfolio construction, credit analytics and trading to insurers, wealth managers, financial institutions, asset owners and asset managers.
Transparently.AI detects accounting manipulation and fraud across more than 85,000 listed companies worldwide, serving portfolio managers, risk professionals, auditors, banks, exchanges and sovereign investors.
IVM Markets sells structured product idea generation and optimization to the distribution side of the equity derivatives market, meaning brokers, private banks, asset managers, wealth advisers and insurance companies rather than the banks that issue the products.
AgentSmyth sells an agent platform for capital markets desks at banks, brokers and asset managers. Its Trading Harness reads a question, decides which of six specialist agents it needs and runs them in parallel.
Pricing
Vendor-published figures are labeled as such. Figures labeled “Estimated” are derived from third-party sources and have not been confirmed by the vendor.
| Entry Price | Pricing Basis | Data Protection Terms | Implementation | Source |
|---|---|---|---|---|
|
Quoted by Exegy. Ten days of sample data are free.
|
SIGNUM is sold as a real time signal service by quote, with summary datasets also sold through distributors. | Data terms are set in the service contract, and the signals are built from public market data. | Trials run over the internet, by cross connect or on an appliance in the firm's test environment, with terms set in the order. | Vendor Published |
Exegy supplies real time signals through its managed market data infrastructure and provides ten days of sample data for evaluation. Liquidity Lamp Summary data is available intraday or at the end of the day, directly or through CloudQuant, Exchange Data International and Symphony. Prices vary with the dataset and delivery method.