Mission Statement

‘Be the Leader in Unstructured Financial Data’

SMA has built Intellectual Property in four Major FinTech Areas:

Natural Language Processing, Textual Parsing, Topic Modeling, and Source Accuracy

Products

US Equities
Futures
FX
ETF
LSE: FTSE
Crypto Currencies

Alternative Data Products

SMA was one of the first firms to normalize Social Media into Predictive JSON API data feeds. Twitter and StockTwits are processed into two separate APIs. Out-of-sample point-in-time historical data for back testing goes back to December 1, 2011

Traditional Data Products

SMA is changing the way Traditional data sets are ingested by turning any document into machine readable textual data with quantitative metric overlays. SMA delivers Real Time APIs and provides historical baselines for quantitative analysis go back as far as 1996

Social Media APIs
SMA supports Quantitative Trading across Long Only and L/S, Market Making, Algorithmic Execution, Asset Management, TCA, Risk, Research, and Market Intelligence
SMA S-Factor API Feed
SMA S-Factor API Feed aggregates Tweets and makes a comparison to baseline to deliver 15 metrics every minute. Snapshots can be requested at different frequencies based on a 24H lookback with a 20D baseline with decay
SMA Activity API Feed
SMA Activity API Feed isolates and aggregates information every minute and allows clients to create their own baselines and metrics. Ideal for both HFT for Market Making and Algo Execution and longer baselines for both trading Weekly, Monthly, and Quarterly Long Only and L/S
SMA Tweet Level API Feed
SMA Tweet Level API Feed delivers SMA metrics at the Tweet level from our Natural Language Processing Sentiment Scoring and SMA’s Account Rating Algorithm. Delivered in 300 milliseconds after receiving the Tweet from Twitter or StockTwits
SMA Lexical API Feed
SMA Lexical API Feed is an enhanced version of the Tweet Level feed. The Lexical Feed also contains a phrase level break down of each Tweet. The data is delivered through the API in 300 milliseconds after receiving the Tweet from Twitter or StockTwits
SMA Dashboard
SMA’s Dashboard is highly customizable across all asset classes. The Dashboard supports a Watchlist with a Tweet stream, live Market Sentiment Snapshot, Daily Pre-Open and Pre-Close Reports, Sentiment and Tweet Volume Charts at the Security, Sector, and Index Level
Alerts and Accuracy Feed
SMA delivers highly Customized Alerts 24/7 to SMS and e-mail on U.S. and LSE Equities, ETFs, Commodity and Financial Futures, FX Spot and Futures Pairs, and Crypto Currencies. SMA curates the most impactful Tweets driving the Predictive Sentiment change in SMS text and e-mail
Consulting and Customized Products
Not all Clients can ingest data via a JSON API. SMA supports Macro, Credit, and Event Driven Funds through consulting and customized products, reports, and programmatic delivery of data visuals

Machine Readable Filings
(in Partnership with S&P Global Market Intelligence)
SEC Edgar Filings and International Reports Parsed Machine-Readable text released in near Real Time with a historical baseline going back to 1996 with Word Counts and Sentiment across Parts, Items, Subsections, and Footnotes.
Private Company Data
In addition to Listed Securities, SMA covers large Multi-National Private Companies to very Early Stage Start-ups. SMA’s Topic Model Engine produces Machine Learning ‘Inclusions’ and ‘Exclusions’. Downstream, SMA Sentiment NLP, Volume, and other metrics are applied
Advertising and Market Intelligence
SMA Data supports Advertising and Market Intelligence. SMA’s platform was built for analyzing unstructured data on the Topic Model Engine to provide pinpoint Advertising and drilled-down Market Intelligence
Indices
Custom Indices across Cboe, Long Only, L/S, ESG, Block Chain, CC Index, Regulatory Filings, Sectors, Market Caps
XBRL and iXBRL Filings
Parsed tabular data using SEC Edgar’s new format for tagging Financial Filing Data

Technology

SMA Sentiment NLP

SMA started with Stanford Tokenization and discovered that we would have to build our own Machine Learning Natural Language Processing platform. SMA uses a multi-pass approach combining techniques like stemming and lemmatization. SMA’s environment is supervised and unsupervised, and uses programmatic regression analysis

SMA Parsing Engine

SMA has built a general document parser to produce structured data in the form of machine-readable JSON. Algorithms and historical baselines can be built to generate Alpha and actionable Market Intelligence

SMA Topic Engine

To capture information related to a Security, Private Company, or Custom Topic, SMA has built a Real Time Machine Learning Topic Model Engine that uses Collocation to identify Inclusions and Exclusions. Topics are constructed using a rules-based system grounded in qualitative knowledge.

SMA Source Rating

It is not just what is being said in Traditional and Social Media, it is who is saying it. SMA has developed a twelve-factor algorithm to score each source before inclusions in sentiment and volume metrics. The ability to eliminate spam, bad bots, bad actors produces a pure data set.

Patents

Granted Patent

#9,104,734 has been granted. Extracting, Cleansing and Publication of metrics on social media data Patent
#10,031,909 has been granted. These claims added Real Time Alerts
Canada and Australia Cases Granted

Pending

U.S. Application Ser. No. 16,018,953 has been filed

In Process

SMA General Document Parser
SMA Source Rating