Our sources

We generate our news by enriching and combining data from multiple financial data sources

At the moment we focus on the Swedish markets, both regulated and MTF:s

Our coverage is continuously expanding. Please contact us if you have ideas about other sources we could cover for you.

Current data sources

Name

Type

Integration

Insynsregistret

Insider transactions

CSV parsing

Malta's PDMR announcements

Insider transactions

Web scraping

Blankningsregistret

Short positions

Excel parsing

Börsinformation

Press releases, Reports

Web scraping

MFN

Press releases

Websub (nice!)

Nasdaq Company News

Press releases

Email parsing

Flaggnings-email

Share holder changes

Email parsing

FINRA-RegSho

Short trading volumes

TXT parsing

Pensionsmyndigheten

Fund performance

Excel parsing

Yahoo! finance

Quote data

Unofficial API

GLEIF

Legal entity information

API

Possible future data sources

Name

Type

Integration

Investor pages

Company information

Web scraping

Bolagsverket

Company information

API

Wikipedia

Word explanations

API

Twitter

Event comments

API

SCB

Macro data

API

Fund allocations

Changes in fund allocations

Unknown (XML files?)

Insider transactions

Example article:

Svolders VD köper aktier för 144 000 SEK

Namn Namnsson betalade 112,4 SEK per aktie igår. Han köpte 1 200 aktier.

Features:

  • Aggregation of multiple transactions published by the same person and company during the same day

  • Based on name we try to figure out pronoun of person/company

  • Sanity check of self-declared price per share against quote data to avoid publishing erroneous entries

  • Disclaimer if transaction is part of internal shares program

  • PNG image of security quote chart annotated with insider sell and buy transactions

Possible improvements:

  • Handle incoming corrections in some way, perhaps just add their comments to the bottom of article

  • We already do mapping of self-declared title to a set of standardized occupation codes, use these to set title in snippet to avoid messy titles

  • Find image link of responsible person in company investors pages and add that to the news if transaction is above some threshold

  • If responsible is a company then look up that company in e.g Bolagsverket API and add summary of company in article

  • Aggregate different types of transactions by different persons within the same company

  • Add summary of responsible persons recent previous transactions

Short positions

Example article:

GSA CAPITAL PARTNERS LLP minskar sin korta position i OASMIA PHARMACEUTICAL AB

GSA CAPITAL PARTNERS LLP minskade sin korta position med 0.02 procentenheter till totalt 0.68 % igår.

Features:

  • Relative change of major short position holder

  • PNG image of changes in net short positions for security

Possible improvements:

  • Replace capitalized names with their proper representations

  • Aggregate if multiple position holders in the same security

  • Add summary of the position holder in some way

Press releases / Reports

Very basic at this point due to lack of integrations with most distributors

Example article:

Ökning av antalet aktier och röster i Balco Group

Bilagor: filename.pdf

Features:

  • Title of press release

  • External links to attachments

  • Varying accuracy in mapping press release to e.g related ISIN:s

  • Varying accuracy in mapping press release to category like e.g QuarterlyReport, Pressrelease

Possible improvements:

  • Integrations with Cision and Globe Newswire would make any feature development much easier

  • Extract some key quote from press release

  • "Impact score" idea: Generate text vector (word2vec method) from press release title and perhaps summary. This can then be combined with same-day quote movement to create labeled dataset for training. Goal would be to classify impact of press release.

Funds

Currently only days-delayed quote and category data from Pensionsmyndigheten with our own parser. This only covers funds in the PPM system though.

Possible improvements: