THOMSON REUTERS

MARKETPSYCH
INDICES

USER GUIDE
MARKETPSYCH INDICES VERSION 3.3

Document Version: 1.0

Date of Issue: 26 April 2019
 About This Document

CONTENTS
About This Document 4
Intended Readership 4
In This Guide 4
Feedback 4

Chapter 1 Introduction 5
Overview 5
Contact Information 5
Documentation and Notifications 5

Chapter 2 Window Lengths and Update Frequencies 6

Overview 6
File Naming Convention 6
Available Combinations 6
Example 6
Daily Update Frequency Anomaly 6

Chapter 3 General Comments about Index Values 7

”NA” Index Values 7
Negative Values in Unipolar Indices 7

Chapter 4 Underlying Content Sources 8

Overview 8
News 8
Social Media 8

Chapter 5 Fields Generic to All TRMI Asset Classes 9

Id 9
Asset Code 9
Window Timestamp 9
Incoming Sources 9
System Version 10
Buzz 10

Chapter 6 Companies and Company Groups 11

Companies Assets 11
Company Groups Assets 11
Companies and Company Group TRMI Indices 16

Chapter 7 Currencies 18
Currencies Assets 18

MarketPsych Indices 3.3 – User Guide 2

Document Version 1.1
 About This Document

Currencies TRMI Indices 18

Chapter 8 Agricultural Commodities 20

Agricultural Commodities Assets 20
Agricultural Commodities TRMI Indices 20

Chapter 9 Energy & Material Commodities 22

Energy & Material Commodities Assets 22
Energy & Material Commodities TRMI Indices 22

Chapter 10 Countries 24
Countries Assets 24
Countries TRMI Indices 26

Chapter 11 Country Markets 29

Country Markets Assets 29
Country Markets TRMI Indices 29

Chapter 12 Cryptocurrencies 32
Cryptocurrency Assets 32
Cryptocurrency TRMI Indices 32

Chapter 13 Constructing TRMI with Greater Window Lengths 34

Background 34
Creating Custom TRMI Window Lengths 34
Comparing Constructed TRMI Data to Actual TRMI Data 34

Chapter 14 MRN FTP Site: Data and Reference Files 35

Overview 35
Access 35
Window Lengths and Update Frequencies 35
Live Files 35
Archive Files 37
Companies and Cryptocurrencies Reference Files 40
Trial Access Compared to Production Access 43

MarketPsych Indices 3.3 – User Guide 3

Document Version 1.1
 About This Document

ABOUT THIS DOCUMENT

INTENDED READERSHIP
This guide is primarily intended for individuals with responsibility to develop applications to consume Thomson
Reuters MarketPsych Indices (TRMI) messages. The information contained within this document is also of use to
anyone working with the associated archive files.

IN THIS GUIDE
This document describes the format of TRMI output images. It is applicable to both the live data and the archive files.
Except for cryptocurrencies, all assets referenced in this guide are generated using TRMI version 3.2. The
cryptocurrencies asset class is reflective of updates to TRMI version 3.3.

FEEDBACK
If you have any comments on this document, please contact the Thomson Reuters Machine Readable News team at
[email protected].

MarketPsych Indices 3.3 – User Guide 4

Document Version 1.1
 Introduction

CHAPTER 1 INTRODUCTION
OVERVIEW
Thomson Reuters MarketPsych Indices (TRMI) analyze news and social media in real-time to convert the volume and
variety of professional news and the internet into manageable information flows that drive sharper decisions. The
indices are delivered as real-time data series that can easily be incorporated into your investment and trading decision
processes – quantitative or qualitative.
Three types of indicators are provided:
• Emotional indicators such as Anger, Fear and Joy
• Macroeconomic metrics including Earnings Forecast, Interest Rate Forecast, Long vs. Short
• Buzz metrics on the asset level, i.e., Buzz, and on market-moving topics for that asset, such as Litigation,
Regulatory Crackdown, Mergers and Volatility
The indicators are updated every minute for companies, sectors, regions, countries, commodities and energy topics,
indices and currencies. They can be translated directly into spreadsheets or charts that can be monitored by traders,
risk managers or analysts – or they can be plugged straight into your algorithms for low frequency or longer term asset
allocation or sector rotation decisions.

CONTACT INFORMATION

For Production Customers

Production customers should direct their TRMI questions or support issues via one of the following:
 Phone: Call +1 877-814-3571 and choose option 2.
 Electronic: Visit the “Contact Us” or “RAISE A CASE” link on My Refinitiv. In filling out the form, type in
“Thomson Reuters MarketPsych Indices” as the product name.

For Trial Customers

Trial customers are assigned a dedicated Presales Technical Engineer to assist with technical questions during the trial.
During the trial, all questions should be sent directly to the Presales Technical Engineer.

DOCUMENTATION AND NOTIFICATIONS

Client-facing documentation may be found on the TRMI My Account page, located here. Other information may be
found on the MRN (Machine Readable News) FTP site, as described in Chapter 14.
Clients are advised to subscribe to service alerts regarding the live and archive TRMI services. To subscribe, go the
Service Alert-Edit Subscriptions page, here. Then check the box at Application -> Financial Information Applications ->
Enterprise Information Products ->Thomson Reuters Machine Readable News -> Thomson Reuters MarketPsych
Indices. Then click the button at bottom right.
For updates on releases of new versions of TRMI, clients should sign up for Product Client Notifications (PCNs). This
will send advance email notifications on relevant releases and updates. To sign up for PCNs on MarketPsych Indices,
go to the PCN subscription page, here. Check the box at All Products -> Reuters Enterprise Information -> Thomson
Reuters MarketPsych Indices. Toward the bottom, make sure that the box for “All Client impacts”, or at least “For
Action” under that, is checked. Then click the link in the lower or upper right to confirm your selection.

MarketPsych Indices 3.3 – User Guide 5

Document Version 1.1
 Window Lengths and Update Frequencies

CHAPTER 2 WINDOW LENGTHS AND UPDATE FREQUENCIES

OVERVIEW
TRMIs are asset-level scores on a collection of content. There are two time-related metrics that determine TRMI scores.
The window length determines what range of content is scored in generating a set of TRMIs. For live content, all
content collected by MarketPsych within that time period will be incorporated into the corresponding window. The
update frequency determines the time between consecutive TRMI scores. (See Chapter 5 for more information on
content types.)

FILE NAMING CONVENTION

For files on the FTP site, the convention for denoting the combination of window length and update frequency shall be

W[Period Length]_U[Period Length], where W denotes window length and U denotes update frequency.

Period lengths can be any of the following:

 01M: 1 minute
 HOU: hourly (60 minutes)
 DAI: daily (1440 minutes)

AVAILABLE COMBINATIONS
Window Length Update Frequency FTP File Abbreviation
1 minute 1 minute W01M_U01M
1440 minutes / 24 hours 1 hour WDAI_UHOU
1440 minutes / 24 hours Daily, at 3:30 Eastern time WDAI_UDAI

EXAMPLE
To illustrate with the WDAI_UHOU data, one series of TRMI scores occurs with data collected between December 22,
2016 15:00:00 UTC and December 23, 2016 15:00:00 UTC, a 1440-minute / 24-hour window length. The next set of
scores would occur with data collected between December 22, 2016 16:00:00 UTC and December 23, 2016 16:00:00
UTC, because of the 60-minute update frequency.

DAILY UPDATE FREQUENCY ANOMALY

Daily update frequency refers to TRMI windows that end at 3:30pm Eastern time. Because this time zone is subject to
daylight savings time adjustments, there is a biannual anomaly around the 24-hour / 1440-minute window length. When
a daylight savings hour is added or subtracted in the previous 24 hours, the time since the last update can be 25 or 23
hours, respectively. As a result, the window length also changes to 25 or 23 hours.

MarketPsych Indices 3.3 – User Guide 6

Document Version 1.1
 General Comments about Index Values

CHAPTER 3 GENERAL COMMENTS ABOUT INDEX VALUES

”NA” INDEX VALUES
As mentioned in Chapter 2, all TRMIs are based on relevant text collected over a window of content.

If over that window there was no relevant text identified for a particular index, then the correct index value is “NA”, not
zero, and the index will appear blank. For example, this could happen for the carryTrade index on the AUD (Australian
Dollar) in news if there were no discussions of the carry trade with the Australian dollar found in news-based text over
the content collected in that window.

NA differs in meaning from true zero in that true zero represents the presence of text corresponding to positive and
negative values that add up to zero. In other words, a zero value reflects that relevant text was found and its sentiment
implications net to zero. In contrast, NA represents the absence of any relevant text and of any resultant measurement.

Note that when the Buzz is zero, this means that no values were detected for any of the indices and thus all index
values necessarily will be NA. See Chapter 5 for more information on Buzz.

NEGATIVE VALUES IN UNIPOLAR INDICES

The indices are marked as ranging from either -1 to 1 or 0 to 1, corresponding to bipolar and unipolar indices,
respectively. In practice, those denoted as “unipolar” can in fact range below 0, although not below -1. This occurs
because unipolar indices reflect the orthogonal nature of many single emotions and topics. A negative comment such
as, “I don’t enjoy owning this stock” is not emotively equivalent to, “I am pessimistic about the stock’s prospects” or “I
am angry with the company’s management.” The initial statement is specifically one of negative Joy, which decreases
the overall Joy index for assets related to that company. When there are many such negative Joy comments for an
asset, the Joy index itself may show negative values.

Nonetheless, in practice unipolar indices are positive over 90% of the time, because language typically reflects positive
assertions.

Thus, in the sections below we will mark this range as “0 to 1*”.

MarketPsych Indices 3.3 – User Guide 7

Document Version 1.1
 Underlying Content Sources

CHAPTER 4 UNDERLYING CONTENT SOURCES

OVERVIEW
TRMIs are evaluated on three different content sets: news, social media, and the combined content. History on all
content dates back to the beginning of 1998 for all assets except cryptocurrencies. Cryptocurrency history extends
back to 2009. Only English-language text is used.

NEWS
Reuters news is present in the entire historical news dataset, as are a host of mainstream news sources collected by
MarketPsych Data. During 2005, the archive began including Internet news content collected by LexisNexis. The
LexisNexis content is restricted to those from top international and business news sources, top regional news sources,
and leading industry sources. In 2017 MarketPsych Data added additional cryptocurrency specific sources to the
cryptocurrencies asset class feed.

SOCIAL MEDIA
The social media collection process is more diverse. It begins in 1998 with Internet forum and message board content.
Starting in late 2008, LexisNexis social media content was added. Starting in late 2009, tweets were included. Using
popularity ranks measured by incoming links, this includes generally the top 20% of blogs, microblogs, and other
financial social media content. MarketPsych Data also included content from hundreds of less-popular asset-specific
blogs and forums.
Note that the entire LexisNexis social media content set for top source ranks is included in the country (COU) social
media dataset, while other assets contain specifically selected social media content sources.

MarketPsych Indices 3.3 – User Guide 8

Document Version 1.1
 Fields Generic to All TRMI Asset Classes

CHAPTER 5 FIELDS GENERIC TO ALL TRMI ASSET CLASSES

ID
The id field identifies the data row. It is composed according to the following:
{feedFamilyCode}:{windowTimestamp in yyyy-mm-dd_hh.mm.ss}.{ dataType}.{assetClass}

Feed Family Code

The feedFamilyCode field describes the “feed” that delivers the raw content being scored by the TRMI system.
For TRMI it is “mp”, because MarketPsych Data centrally collects and filters the content scored by TRMI.

Other Components
See Incoming Sources section below on the dataType field and the Asset Code section for a list of assetClass values.

ASSET CODE
The assetCode field represents the code used to denote the asset scored, where the asset code type varies by Asset
Class.

Asset Class Asset Class Type of Asset Code Identifier

Abbreviation
Individual companies CMPNY Thomson Reuters organizational PermID
CMPNY_AMER
CMPNY_APAC
CMPNY_EMEA
Company groups CMPNY_GRP MarketPsych-originated code
Currencies CUR Three-character ISO-4217 code
Agricultural commodities COM_AGR
Thomson Reuters topic code
Energy & material COM_ENM
commodities
Countries COU Thomson Reuters topic code, referring to
Country markets COU_MKT the home country. Also corresponds to ISO
3166-1 code.
Cryptocurrencies CRYPTO The coinmarketcap.com ticker identifier

WINDOW TIMESTAMP
As mentioned in Chapter 2, TRMIs are based on content aggregated over a trailing window. The windowTimestamp
field represents the endpoint of that period.
It is expressed in ISO-8601 format: “yyyymmddThh:mm:ss.000Z”, where “000” represents the milliseconds and “Z”
represents the time zone. The time zone is GMT/UTC.

INCOMING SOURCES
The dataType field represents the type of content source(s) on which these TRMIs are based.

There are three possible values:

 “News” for news
 “Social” for social media
 “News_Social” for the combined content

See Chapter 4 for more information on these sources.

MarketPsych Indices 3.3 – User Guide 9

Document Version 1.1
 Fields Generic to All TRMI Asset Classes

SYSTEM VERSION
The systemVersion field represents an overall build for the MarketPsych Indices. Different versions may lead to different
scores as enhancements are made and bugs are fixed. It is thus used for version control.
It consists of a three-level version string preceded by an “MP” prefix denoting the MarketPsych scoring engine, e.g.,
“MP:3.2.09”.
The third digit of the System Version increases by one, for all asset classes, in conjunction with monthly updates to the
companies coverage list.

BUZZ
The buzz field represents a sum of entity-specific words and phrases used in TRMI computations. It can be non-integer
when any of the words/phrases are described with a “minimizer”, which reduces the intensity of the primary word or
phrase. For example, in the phrase “less concerned” the score of the word “concerned” is minimized by “less”.
Additionally, common words such as “new” may have a minor but significant contribution to the Innovation TRMI. As a
result, the scores of common words/phrases with minor TRMI contributions can be minimized.

MarketPsych Indices 3.3 – User Guide 10

Document Version 1.1
 Companies and Company Groups

CHAPTER 6 COMPANIES AND COMPANY GROUPS

COMPANIES ASSETS
TRMI Companies data covers more than 12,000 active companies from over 75 countries, across the feeds and
archives. In general, these companies were selected for inclusion based on domicile, listed market capitalization,
and/or economic/industry significance. Coverage includes all companies that are or were constituents of a company
groups asset. Coverage is updated monthly.

A list of the covered companies and monthly changes in coverage is available on the MRN FTP site. See Chapter 14 for
information on the files and how to access them.

COMPANY GROUPS ASSETS

TRMI Company Groups data covers 61 assets corresponding to groups of companies assembled according to domicile,
market capitalization, and/or business classification.

Many of these correspond to grouping in the hierarchical Thomson Reuters Business Classification (TRBC) system. In
descending order of hierarchy, the four levels are economic sector, business sector, industry group and industry. For
more information on TRBC, please see
http://thomsonreuters.com/products_services/financial/thomson_reuters_indices/trbc/. All TRBC codes below are extant
in the TRBC 2012 system.

The 61 assets can be characterized in a few ways. 26 are composed of US-based companies, while 17 are for non-US
companies and another 18 are global. 21 resemble equity indices and are filtered chiefly by market cap ranks, while the
other 40 are composed according to a combination of TRBC code, domicile, and market cap above $100 million USD.

Note that because these groups are calculated with approximate point-in-time composition, all the asset codes used
here were invented for this application. They are not RICs.

US-based Company Groups (26)

Index-oriented Assets (4)

All internal rankings are by market capitalization.

Asset Code Description Resembling Index

MPTRXUS30 Top 30 US-based companies Dow Jones Industrial Average
MPTRXUS500 Top 500 US-based companies S&P 500
MPTRXUSMID2000 Ranks 2001-3000 of US-based companies Russell 2000
MPTRXUSNAS100 Top 100 Nasdaq-based companies Nasdaq 100

TRBC Economic Sector-based Assets (10)

Asset Code TRBC Description TRBC Code
MPTRXUSENE Energy 50
MPTRXUSMAT Basic Materials 51
MPTRXUSIND Industrials 52
MPTRXUSYCY Cyclical Consumer Goods & Services 53
MPTRXUSNCY Non-Cyclical Consumer Goods & Services 54
MPTRXUSFIN Financials 55
MPTRXUSHLC Healthcare 56

MarketPsych Indices 3.3 – User Guide 11

Document Version 1.1
 Companies and Company Groups

MPTRXUSTEC Technology 57
MPTRXUSCOM Telecommunications Services 58
MPTRXUSUTL Utilities 59

TRBC Business Sector-based Assets (5)

Asset Code TRBC Description TRBC Code
MPTRXUSI4 Transportation 5240
MPTRXUSY3 Cyclical Consumer Services 5330
MPTRXUSY4 Retailers 5340
MPTRXUSN1 Food & Beverages 5410
MPTRXUSF4 Real Estate 5540

TRBC Industry Group-based Assets (2)

Asset Code TRBC Description TRBC Code
MPTRXUSOILS Oil & Gas Related Equipment and Services 501030
MPTRXUSAERO Aerospace & Defense 521010

TRBC Industry-based Assets (5)

Asset Code TRBC Description TRBC Code
MPTRXUSOILE Oil & Gas Exploration and Production 50102020
MPTRXUSWAST Environmental Services 52203010
MPTRXUSAIRL Airlines 52406010
MPTRXUSHBLD Homebuilding & Construction Supplies 53203010
MPTRXUSBANK Banks 55101010

Non-US- based Company Groups (17)

Index-oriented Assets (17)

All internal rankings are by market capitalization.

Asset Code Description Resembling Index

MPTRXAU500 Top 500 Australia-based companies ASX All Ordinaries
MPTRXBR50 Top 50 Brazil-based companies IBRX 50
MPTRXCA250 Top 250 Canada-based & Toronto-listed S&P/TSX Composite
companies
MPTRXCH20 Top 20 Switzerland-based companies Swiss Market
MPTRXCN300 Top 300 China-based companies CSI 300
MPTRXDE30 Top 30 Germany-based companies Deutsche Börse DAX 30
MPTRXEM50 Top 50 emerging markets companies MSCI 50
MPTRXES35 Top 35 Spain-based companies IBEX 35
MPTRXEU50 Top 50 pan-European companies EURO STOXX 50
MPTRXFR40 Top 40 France-based companies CAC 40
MPTRXGB100 Top 100 UK-based & LSE-listed companies FTSE 100
MPTRXGBMID250 Ranks 101-350 of UK-based & LSE-listed FTSE Mid 250
companies
MPTRXHK50 Top 50 Hong Kong-listed companies based in Hang Seng

MarketPsych Indices 3.3 – User Guide 12

Document Version 1.1
 Companies and Company Groups

Hong Kong and China

MPTRXIN50 Top 50 India-based companies Nifty 50
MPTRXJP225 Top 225 Japan-based companies Nikkei 225
MPTRXRU50 Top 50 Russia-based companies RTS
MPTRXSG30 Top 30 Singapore-based companies FTSE Straits Times

Global Company Groups (18)

TRBC Economic Sector-based Assets (10)

Asset Code TRBC Description TRBC Code
MPTRXENE Energy 50
MPTRXMAT Basic Materials 51
MPTRXIND Industrials 52
MPTRXYCY Cyclical Consumer Goods & Services 53
MPTRXNCY Non-Cyclical Consumer Goods & Services 54
MPTRXFIN Financials 55
MPTRXHLC Healthcare 56
MPTRXTEC Technology 57
MPTRXCOM Telecommunications Services 58
MPTRXUTL Utilities 59

TRBC Business Sector-based Assets (1)

Asset Code TRBC Description TRBC Code
MPTRXE14 Renewable Energy 5020

TRBC Industry Group-based Assets (3)

Asset Code TRBC Description TRBC Code
MPTRXCOAL Coal 501010
MPTRXBIOT Biotechnology & Medical Research 562020
MPTRXT11 Semiconductors & Semiconductor Equipment 571010

TRBC Industry-based Assets (2)

Asset Code TRBC Description TRBC Code
MPTRXGOLD Precious Metals & Minerals 51201010
MPTRXGAMI Casinos & Gaming 53301030

TRBC Hybrid-driven Assets (2)

Asset Code TRBC Description TRBC Code
MPTRXPKB Construction Materials 512020
Construction & Engineering 522010
Homebuilding & Construction Supplies 532030
MPTRXPEJ Leisure Products 532050
Hotels & Entertainment Services 533010

MarketPsych Indices 3.3 – User Guide 13

Document Version 1.1
 Companies and Company Groups

Constitution of AMER, APAC, and EMEA Regional Groups

The CMPNY asset class consists of companies domiciled in the following countries. Clients who subscribe to a
regional asset class receive companies domiciled in the respective region.

CMPNY_AMER
Companies included in the Americas (CMPNY_AMER) sub-grouping include those domiciled in the following
countries.

Asset
Country
Code
AR Argentina
BB Barbados
BR Brazil
CA Canada
CL Chile
CO Colombia
CW Curacao
EC Ecuador
JM Jamaica
KY Cayman Islands
MX Mexico
PA Panama
PE Peru
PR Puerto Rico
PY Paraguay
US United States
VE Venezuela
VG Virgin Islands (British)

CMPNY_APAC
Companies included in the Asia and Pacific (CMPNY_APAC) sub-grouping includes those domiciled in the following
countries.

Asset
Country
Code
AU Australia
AZ Azerbaijan
BD Bangladesh
CN China
FJ Fiji
GE Georgia
HK Hong Kong
ID Indonesia
IN India
JP Japan
KG Kyrgyzstan
KR South Korea

MarketPsych Indices 3.3 – User Guide 14

Document Version 1.1
 Companies and Company Groups

KZ Kazakhstan
LK Sri Lanka
MH Marshall Islands
MN Mongolia
MO Macau
MY Malaysia
NZ New Zealand
PG Papua New Guinea
PH Philippines
PK Pakistan
SG Singapore
TH Thailand
TW Taiwan
VN Vietnam

CMPNY_EMEA
Companies included in the Europe, Middle East, and Africa (CMPNY_EMEA) sub-grouping includes those domiciled
in the following countries.

Asset
Country
Code
AE United Arab Emirates
AT Austria
BE Belgium
BH Bahrain
BM Bermuda
CH Switzerland
CY Cyprus
CZ Czech Republic
DE Germany
DK Denmark
EG Egypt
ES Spain
FI Finland
FR France
GB United Kingdom
GG Guernsey
GI Gibraltar
GR Greece
HR Croatia
HU Hungary
IE Ireland
IL Israel
IM Isle of Man
IQ Iraq
IS Iceland
IT Italy
JE Jersey
KE Kenya

MarketPsych Indices 3.3 – User Guide 15

Document Version 1.1
 Companies and Company Groups

KW Kuwait
LB Lebanon
LU Luxembourg
MA Morocco
MC Monaco
MT Malta
NG Nigeria
NL Netherlands
NO Norway
OM Oman
PL Poland
PT Portugal
QA Qatar
RO Romania
RU Russia
SA Saudi Arabia
SE Sweden
SK Slovakia
TG Togo
TR Turkey
TZ Tanzania
UA Ukraine
ZA South Africa
ZM Zambia

Constitution of Company Groups Assets in the TRMI Archives and Feed

The constituent companies of these assets change over time. In the archives, the TRBC-based constituent lists are
refreshed quarterly starting on January 1, 2006. The remaining groups are refreshed quarterly since the beginning of
the archives, January 1, 1998. All groups have their constituents update monthly starting from September 2014.

In the live data, the systemVersion value will increase with each monthly update.

COMPANIES AND COMPANY GROUP TRMI INDICES

The 34 TRMI indices for the companies and company groups carry six significant digits past the decimal point. Negative
numbers have a leading minus (-) sign. The table below summarizes these fields. For more information on those with
range “0 to 1*”, see Chapter 3 above.

Index Description: references in news and social media to... Range

sentiment overall positive references, net of negative references -1 to 1
optimism optimism, net of references to pessimism -1 to 1
joy happiness and affection 0 to 1*
loveHate love, net of references to hate -1 to 1
trust trustworthiness, net of references connoting corruption -1 to 1
anger anger and disgust 0 to 1*
conflict disagreement and swearing net of agreement and conciliation -1 to 1

MarketPsych Indices 3.3 – User Guide 16

Document Version 1.1
 Companies and Company Groups

fear fear and anxiety 0 to 1*

gloom gloom and negative future outlook 0 to 1*
stress arousal and intensity, weighted towards distress 0 to 1*
surprise unexpected events and surprise 0 to 1*
timeUrgency urgency and timeliness, net of references to tardiness and -1 to 1
delays
uncertainty uncertainty and confusion 0 to 1*
violence violent crime, terrorism, and war 0 to 1*
emotionVsFact all emotional sentiments, net of all factual and topical -1 to 1
references
marketRisk positive emotionality and positive expectations net of negative -1 to 1
emotionality and negative expectations. Includes factors from
social media found characteristic of speculative bubbles –
higher values indicate greater bubble risk. Also known as the
“Bubbleometer.”
longShort buying, net of references to shorting or selling -1 to 1
longShortForecast forecasts of buying, net of references to forecasts of shorting -1 to 1
or selling
priceDirection price increases, net of references to price decreases -1 to 1
priceForecast forecasts of asset price rises, net of references to forecasts of -1 to 1
asset price drops
volatility volatility in market prices or business conditions 0 to 1*
analystRating upgrade activity, net of references to downgrade activity -1 to 1
cyberCrime Cyber attacks and data breaches 0 to 1*
debtDefault debt defaults and bankruptcies 0 to 1*
dividends dividends rising, net of references to dividends falling -1 to 1
innovation innovativeness 0 to 1*
earningsForecast expectations about improving earnings, less those of -1 to 1
worsening earnings
fundamentalStrength positivity about accounting fundamentals, net of references -1 to 1
to negativity about accounting fundamentals
laborDispute labor unrest and work stoppages 0 to 1*
layoffs staff reductions and layoffs 0 to 1*
litigation litigation and legal activity 0 to 1*
managementChange changes in a company's management team, net of references -1 to 1
to stability in the management team
managementTrust trust expressed in a company's management team, net of -1 to 1
references to reports of unethical behavior amongst the
management team
mergers merger or acquisition activity 0 to 1*

MarketPsych Indices 3.3 – User Guide 17

Document Version 1.1
 Currencies

CHAPTER 7 CURRENCIES
CURRENCIES ASSETS
There are 45 currencies, listed in the table below.

Asset Asset
Code Currency Code Currency
ARS Argentine Peso NOK Norwegian Krone
AUD Australian Dollar PKR Pakistani Rupee
BDT Bangladeshi Taka PEN Peruvian Sol
BTC Bitcoin PHP Philippine Peso
BRL Brazilian Real PLN Polish Zloty
CAD Canadian Dollar RON Romanian Leu
CLP Chilean Peso RUB Russian Ruble
CNY Chinese Yuan Renminbi SAR Saudi Riyal
COP Colombian Peso SGD Singapore Dollar
CZK Czech Koruna ZAR South African Rand
EGP Egyptian Pound KRW South Korean Won
EUR Euro SEK Swedish Krona
HKD Hong Kong Dollar CHF Swiss Franc
HUF Hungarian Forint TWD Taiwanese Dollar
INR Indian Rupee THB Thai Baht
IDR Indonesian Rupiah TRY Turkish Lira
IRR Iranian Rial USD U.S. Dollar
ILS Israeli Shekel UAH Ukrainian hryvnia
JPY Japanese Yen AED United Arab Emirates Dirham
MYR Malaysian Ringgit GBP United Kingdom Pound Sterling
MXN Mexican Peso VEF Venezuelan Bolívar
NZD New Zealand Dollar VND Vietnam Dong
NGN Nigerian Naira

CURRENCIES TRMI INDICES

The 24 TRMI indices for the currencies asset class carry six significant digits past the decimal point. Negative numbers
have a leading minus (-) sign. The table below summarizes these fields. For more information on those with range “0 to
1*”, see Chapter 3 above.

Index Description: references in news and social media to... Range

sentiment overall positive references, net of negative references -1 to 1
optimism optimism, net of references to pessimism -1 to 1
joy happiness and affection 0 to 1*
loveHate love, net of references to hate -1 to 1
trust trustworthiness, net of references connoting corruption -1 to 1
anger anger and disgust 0 to 1*

MarketPsych Indices 3.3 – User Guide 18

Document Version 1.1
 Currencies

conflict disagreement and swearing net of agreement and -1 to 1

conciliation
fear fear and anxiety 0 to 1*
gloom gloom and negative future outlook 0 to 1*
stress arousal and intensity, weighted towards distress 0 to 1*
surprise unexpected events and surprise 0 to 1*
timeUrgency urgency and timeliness, net of references to tardiness and -1 to 1
delays
uncertainty uncertainty and confusion 0 to 1*
violence violent crime, terrorism, and war 0 to 1*
emotionVsFact all emotional sentiments, net of all factual and topical -1 to 1
references
marketRisk positive emotionality and positive expectations net of negative -1 to 1
emotionality and negative expectations. Includes factors from
social media found characteristic of speculative bubbles –
higher values indicate greater bubble risk. Also known as the
“Bubbleometer.”
longShort buying, net of references to shorting or selling -1 to 1
longShortForecast forecasts of buying, net of references to forecasts of shorting -1 to 1
or selling
priceDirection price increases, net of references to price decreases -1 to 1
priceForecast forecasts of asset price rises, net of references to forecasts of -1 to 1
asset price drops
volatility volatility in market prices or business conditions 0 to 1*
carryTrade carry trade 0 to 1*
currencyPegInstability the instability of a currency peg, net of references to the -1 to 1
stability of a currency peg
priceMomentum currency price trend strength, net of references to trend -1 to 1
weakness

MarketPsych Indices 3.3 – User Guide 19

Document Version 1.1
 Agricultural Commodities

CHAPTER 8 AGRICULTURAL COMMODITIES

AGRICULTURAL COMMODITIES ASSETS
There are 14 agricultural commodities being scored. The asset codes correspond to Thomson Reuters topic codes
found in news stories.

Asset
Code Commodity
RAPOIL Canola
CTTL Cattle
COC Cocoa
COF Coffee
COR Corn
COT Cotton
HOGS Hogs
ORJ Orange Juice
POIL Palm Oil
RICE1 Rice
SOIL Soybean Oil
SOY1 Soybeans
SUG Sugar
WHT Wheat

AGRICULTURAL COMMODITIES TRMI INDICES

The 31 TRMI indices for the agricultural commodities asset class carry six significant digits past the decimal point.
Negative numbers have a leading minus (-) sign. The table below summarizes these fields. For more information on
those with range “0 to 1*”, see Chapter 3 above.

Index Description: references in news and social media to... Range

sentiment overall positive references, net of negative references -1 to 1
optimism optimism, net of references to pessimism -1 to 1
joy happiness and affection 0 to 1*
loveHate love, net of references to hate -1 to 1
trust trustworthiness, net of references connoting corruption -1 to 1
anger anger and disgust 0 to 1*
conflict disagreement and swearing net of agreement and -1 to 1
conciliation
fear fear and anxiety 0 to 1*
gloom gloom and negative future outlook 0 to 1*
stress arousal and intensity, weighted towards distress 0 to 1*
surprise unexpected events and surprise 0 to 1*
timeUrgency urgency and timeliness, net of references to tardiness and -1 to 1
delays

MarketPsych Indices 3.3 – User Guide 20

Document Version 1.1
 Agricultural Commodities

uncertainty uncertainty and confusion 0 to 1*

violence violent crime, terrorism, and war 0 to 1*
emotionVsFact all emotional sentiments, net of all factual and topical -1 to 1
references
longShort buying, net of references to shorting or selling -1 to 1
longShortForecast forecasts of buying, net of references to forecasts of shorting -1 to 1
or selling
priceDirection price increases, net of references to price decreases -1 to 1
priceForecast forecasts of asset price rises, net of references to forecasts -1 to 1
of asset price drops
volatility volatility in market prices or business conditions 0 to 1*
consumptionVolume Increasing, net of decreasing, commodity consumption -1 to 1
productionVolume Increasing, net of decreasing, commodity production -1 to 1
regulatoryIssues regulatory changes affecting the commodity 0 to 1*
supplyVsDemand surplus supply and lack of demand, net of references to -1 to 1
supply shortage and high demand
supplyVsDemandForecast expectations of supply outstripping demand, net of -1 to 1
references to expectations of demand outstripping supply
acreageCultivated increases in acreage and crop cultivation, net or references -1 to 1
to decreases in acreage and crop cultivation
agDisease commodity disease 0 to 1*
agStress production stress related to disease, water, or weather 0 to 1*
subsidies subsidies affecting commodity prices 0 to 1*
subsidiesSentiment increases in subsidies, net of references to decreases in -1 to 1
subsidies
weatherDamage commodity weather risk and damage 0 to 1*

MarketPsych Indices 3.3 – User Guide 21

Document Version 1.1
 Energy & Material Commodities

CHAPTER 9 ENERGY & MATERIAL COMMODITIES

ENERGY & MATERIAL COMMODITIES ASSETS
There are 22 energy & material commodities being scored. The asset codes correspond to Thomson Reuters topic
codes found in news stories.

Asset
Code Commodity
ALU Aluminum
BIOF Biofuels
COA Coal
CPPR Copper
CRU Crude Oil
BIOETH Ethanol
MOG Gasoline
GOL Gold
HOIL Heating Oil
IRN Iron
JET Jet Fuel
LNG Liquefied Natural Gas
NAP Naphtha
NGS Natural Gas
NKL Nickel
NSEA North Sea Oil
PALL Palladium
PLAT Platinum
RAREE Rare Earths
SLVR Silver
STEE Steel
URAN Uranium

ENERGY & MATERIAL COMMODITIES TRMI INDICES

The 27 TRMI indices for the energy & material commodities asset class carry six significant digits past the decimal
point. Negative numbers have a leading minus (-) sign. The table below summarizes these fields. For more information
on those with range “0 to 1*”, see Chapter 3 above.

Index Description: references in news and social media to... Range

sentiment overall positive references, net of negative references -1 to 1
optimism optimism, net of references to pessimism -1 to 1
joy happiness and affection 0 to 1*
loveHate love, net of references to hate -1 to 1
trust trustworthiness, net of references connoting corruption -1 to 1
anger anger and disgust 0 to 1*
conflict disagreement and swearing net of agreement and -1 to 1
conciliation

MarketPsych Indices 3.3 – User Guide 22

Document Version 1.1
 Energy & Material Commodities

fear fear and anxiety 0 to 1*

gloom gloom and negative future outlook 0 to 1*
stress arousal and intensity, weighted towards distress 0 to 1*
surprise unexpected events and surprise 0 to 1*
timeUrgency urgency and timeliness, net of references to tardiness and -1 to 1
delays
uncertainty uncertainty and confusion 0 to 1*
violence violent crime, terrorism, and war 0 to 1*
emotionVsFact all emotional sentiments, net of all factual and topical -1 to 1
references
longShort buying, net of references to shorting or selling -1 to 1
longShortForecast forecasts of buying, net of references to forecasts of shorting -1 to 1
or selling
priceDirection price increases, net of references to price decreases -1 to 1
priceForecast forecasts of asset price rises, net of references to forecasts -1 to 1
of asset price drops
volatility volatility in market prices or business conditions 0 to 1*
consumptionVolume Increasing, net of decreasing, commodity consumption -1 to 1
productionVolume Increasing, net of decreasing, commodity production -1 to 1
regulatoryIssues regulatory changes affecting the commodity 0 to 1*
supplyVsDemand surplus supply and lack of demand, net of references to -1 to 1
supply shortage and high demand
supplyVsDemandForecast expectations of supply outstripping demand, net of -1 to 1
references to expectations of demand outstripping supply
newExploration new ventures/exploration 0 to 1*
safetyAccident safety accidents 0 to 1*

MarketPsych Indices 3.3 – User Guide 23

Document Version 1.1
 Countries

CHAPTER 10 COUNTRIES
COUNTRIES ASSETS
There are 187 countries or regions being scored. Note that these scores are based on text not only of the
country/region, but also based on cities and other place names inside those geographies.
The asset codes correspond to Thomson Reuters topic codes for geopolitical units. The table below is sorted by
Country/Region.

Asset Code Country/Region Asset Code Country/Region

AF Afghanistan CI Côte d'Ivoire
AL Albania HR Croatia
DZ Algeria CU Cuba
AS American Samoa CY Cyprus
AD Andorra CZ Czech Republic
AO Angola DK Denmark
AR Argentina DJ Djibouti
AM Armenia DO Dominican Republic
AU Australia EC Ecuador
AT Austria EG Egypt
AZ Azerbaijan SV El Salvador
BH Bahrain GQ Equatorial Guinea
BD Bangladesh ER Eritrea
BY Belarus EE Estonia
BE Belgium ET Ethiopia
BJ Benin EZ Euro-zone
BT Bhutan FJ Fiji
BO Bolivia FI Finland
BA Bosnia and Herzegovina FR France
BW Botswana PF French Polynesia
BR Brazil GA Gabon
BG Bulgaria GM Gambia
BF Burkina Faso GE Georgia
BI Burundi DE Germany
KH Cambodia GH Ghana
CM Cameroon GR Greece
CA Canada GU Guam
CV Cape Verde GT Guatemala
KY Cayman Islands GN Guinea
CF Central African Republic GW Guinea-Bissau
TD Chad GY Guyana
CL Chile HT Haiti
CN China HN Honduras
CO Colombia HK Hong Kong
KM Comoros HU Hungary
CD Congo, Democratic Republic of the IS Iceland
CR Costa Rica IN India

MarketPsych Indices 3.3 – User Guide 24

Document Version 1.1
 Countries

Asset Code Country/Region Asset Code Country/Region

ID Indonesia NG Nigeria
IR Iran KP North Korea
IQ Iraq NO Norway
IE Ireland OM Oman
IL Israel PK Pakistan
IT Italy PW Palau
JP Japan PS Palestinian Territories
JO Jordan PA Panama
KZ Kazakhstan PG Papua New Guinea
KE Kenya PY Paraguay
KI Kiribati PE Peru
XK Kosovo PH Philippines
KW Kuwait PL Poland
KG Kyrgyzstan PT Portugal
LA Laos QA Qatar
LV Latvia CG Republic of the Congo
LB Lebanon RO Romania
LS Lesotho RU Russia
LR Liberia RW Rwanda
LY Libya WS Samoa
LT Lithuania ST Sao Tome And Principe
LU Luxembourg SA Saudi Arabia
MO Macau SN Senegal
MK Macedonia RS Serbia
MG Madagascar SC Seychelles
MW Malawi SL Sierra Leone
MY Malaysia SG Singapore
ML Mali SK Slovakia
MT Malta SI Slovenia
MH Marshall Islands SB Solomon Islands
MR Mauritania SO Somalia
MU Mauritius ZA South Africa
MX Mexico KR South Korea
FM Micronesia SS South Sudan
MD Moldova ES Spain
MN Mongolia LK Sri Lanka
ME Montenegro SD Sudan
MA Morocco SR Suriname
MZ Mozambique SZ Swaziland
MM Myanmar SE Sweden
NA Namibia CH Switzerland
NR Nauru SY Syria
NP Nepal TW Taiwan
NL Netherlands TJ Tajikistan
NZ New Zealand TZ Tanzania
NI Nicaragua TH Thailand
NE Niger TG Togo

MarketPsych Indices 3.3 – User Guide 25

Document Version 1.1
 Countries

Asset Code Country/Region Asset Code Country/Region

TO Tonga UY Uruguay
TN Tunisia UZ Uzbekistan
TR Turkey VU Vanuatu
TM Turkmenistan VA Vatican
TV Tuvalu VE Venezuela
UG Uganda VN Vietnam
UA Ukraine YE Yemen
AE United Arab Emirates ZM Zambia
GB United Kingdom ZW Zimbabwe
US United States

COUNTRIES TRMI INDICES

The 48 TRMI indices for the countries asset class carry six significant digits past the decimal point. Negative numbers
have a leading minus (-) sign. The table below summarizes these fields. For more information on those with range “0 to
1*”, see Chapter 3 above.

Index Description: references in news and social media to... Range

sentiment overall positive references, net of negative references -1 to 1
optimism optimism, net of references to pessimism -1 to 1
joy happiness and affection 0 to 1*
loveHate love, net of references to hate -1 to 1
trust trustworthiness, net of references connoting corruption -1 to 1
anger anger and disgust 0 to 1*
conflict disagreement and swearing net of agreement and -1 to 1
conciliation
fear fear and anxiety 0 to 1*
gloom gloom and negative future outlook 0 to 1*
stress arousal and intensity, weighted towards distress 0 to 1*
surprise unexpected events and surprise 0 to 1*
timeUrgency urgency and timeliness, net of references to tardiness and -1 to 1
delays
uncertainty uncertainty and confusion 0 to 1*
emotionVsFact all emotional sentiments, net of all factual and topical -1 to 1
references
marketRisk positive emotionality and positive expectations net of -1 to 1
negative emotionality and negative expectations. Includes
factors from social media found characteristic of speculative
bubbles – higher values indicate greater bubble risk. Also
known as the “Bubbleometer.”
budgetDeficit a budget deficit, net of references to a surplus -1 to 1
businessExpansion businesses expanding, net of references to contraction -1 to 1

MarketPsych Indices 3.3 – User Guide 26

Document Version 1.1
 Countries

commercialRealEstateSentiment positive references to commercial real estate, net of -1 to 1

negative references
consumerSentiment positive consumer sentiment, net of references to negative -1 to 1
consumer sentiment
creditEasyVsTight credit conditions being easy, net of references to credit -1 to 1
conditions being tight
economicGrowth increased business activity, net of references to decreased -1 to 1
business activity
economicUncertainty uncertainty about business climate net of confidence and -1 to 1
certainty
economicVolatility increasing economic volatility, net of economic stability -1 to 1
financialSystemInstability financial system instability, net of references to financial -1 to 1
system stability
fiscalPolicyLooseVsTight fiscal policy being loose, net of references to fiscal policy -1 to 1
being tight
inflation consumer price increases, net of references to consumer -1 to 1
price decreases
inflationForecast forecasts of consumer price increases, net of forecasts of -1 to 1
consumer price decreases (deflation)
infrastructureSentiment Positive net of negative references to institutions, facilities, -1 to 1
transportation, and industry
innovation innovativeness 0 to 1*
investmentFlows investment inflows, net of references to investment outflows -1 to 1
residentialRealEstateGrowth residential real estate expansion, net of references to -1 to 1
contraction
residentialRealEstateSales residential real estate sales rising, net of references to -1 to 1
sales decreasing
residentialRealEstateSentiment positive references to residential real estate, net of negative -1 to 1
references
residentialRealEstateValues residential real estate values rising, net of references to -1 to 1
declining values
tradeBalance exports, net of references to imports -1 to 1
unemployment unemployment rising, net of references to unemployment -1 to 1
falling
unemploymentForecast unemployment rising, net of references to unemployment -1 to 1
falling in the future tense
governmentAnger anger and disgust about government officials and 0 to 1*
departments
governmentCorruption Fraud, deceit, and corruption in government 0 to 1*
governmentInstability governmental instability, net of references to governmental -1 to 1
stability
politicalSentiment positive sentiment expressed about political parties and -1 to 1
institutions net of negative sentiment

MarketPsych Indices 3.3 – User Guide 27

Document Version 1.1
 Countries

regimeChange regime change 0 to 1*

sanctions sanctions or embargoes emanating from or against a 0 to 1*
country
socialInequality social inequality 0 to 1*
socialUnrest social unrest and calls for political change 0 to 1*
cyberCrime cyberattacks, data leaks, and electronic espionage 0 to 1*
terrorism terrorist threats and activities 0 to 1*
violentCrime criminal violence 0 to 1*
war war and militant activity 0 to 1*
agriculturalStress agricultural failures and production deficiencies 0 to 1*
coldWave unusually cold weather 0 to 1*
drought lack of rain with environmental consequences 0 to 1*
earthquake seismic activity 0 to 1*
famine human caloric deficiency 0 to 1*
fire forest, brush, and structural fires 0 to 1*
flood rising water levels with economic consequences 0 to 1*
heatwave unusually hot weather 0 to 1*
humanInfectiousDisease presence and outbreaks of contagious or epidemic 0 to 1*
diseases
tsunami tidal wave activity 0 to 1*
volcanicActivity volcanic eruption 0 to 1*
windStorm cyclonic storms and extreme wind weather events 0 to 1*

MarketPsych Indices 3.3 – User Guide 28

Document Version 1.1
 Country Markets

CHAPTER 11 COUNTRY MARKETS

COUNTRY MARKETS ASSETS
There are 62 countries or regions with the largest economic output being distributed. The asset codes correspond to
Thomson Reuters topic codes for geopolitical units.
The table below is sorted by Country/Region.

Asset Code Country/Region Asset Code Country/Region

DZ Algeria MA Morocco
AR Argentina NL Netherlands
AU Australia NZ New Zealand
AT Austria NG Nigeria
BD Bangladesh NO Norway
BE Belgium PK Pakistan
BR Brazil PE Peru
CA Canada PH Philippines
CL Chile PL Poland
CN China PT Portugal
CO Colombia QA Qatar
CZ Czech Republic RO Romania
DK Denmark RU Russia
EG Egypt SA Saudi Arabia
EZ Euro-zone SG Singapore
FI Finland ZA South Africa
FR France KR South Korea
DE Germany ES Spain
GR Greece SE Sweden
HK Hong Kong CH Switzerland
HU Hungary TW Taiwan
IN India TH Thailand
ID Indonesia TR Turkey
IR Iran UA Ukraine
IQ Iraq AE United Arab Emirates
IE Ireland GB United Kingdom
IL Israel US United States
IT Italy VE Venezuela
JP Japan VN Vietnam
KZ Kazakhstan
KW Kuwait
MY Malaysia
MX Mexico

COUNTRY MARKETS TRMI INDICES

The 30 TRMI indices for the country markets asset class carry six significant digits past the decimal point. Negative
numbers have a leading minus (-) sign. The table below summarizes these fields. For more information on those with
range “0 to 1*”, see Chapter 3 above.
The table below is broken into three sections: stocks, bonds, and rates. Each has its own buzz.

MarketPsych Indices 3.3 – User Guide 29

Document Version 1.1
 Country Markets

Description: references in news and social media to the

country’s major stock indexes and shares traded in that
Index country... Range
stockIndexBuzz sum of all relevant references feeding into the TRMI
bondBuzz sum of all references to the country’s bonds and debt
(excluding corporate debt) in that country
ratesBuzz sum of all references underlying the centralBank,
debtDefault, interestRates, interestRatesForecast, and
monetaryPolicyLooseVsTight TRMI
stockIndexSentiment overall positive references, net of negative references -1 to 1
stockIndexOptimism optimism, net of references to pessimism -1 to 1
stockIndexTrust trustworthiness, net of references connoting mistrust -1 to 1
stockIndexFear fear and anxiety 0 to 1*
stockIndexStress arousal and intensity, weighted towards distress 0 to 1*
stockIndexSurprise unexpected events and surprise 0 to 1*
stockIndexUncertainty uncertainty and confusion 0 to 1*
stockIndexMarketRisk positive emotionality and positive expectations net of -1 to 1
negative emotionality and negative expectations. Includes
factors from social media found characteristic of speculative
bubbles – higher values indicate greater bubble risk. Also
known as the “Bubbleometer.”
stockIndexPriceDirection stock price increases, net of references to price decreases -1 to 1
stockIndexPriceForecast forecasts of stock price rises, net of references to forecasts -1 to 1
of asset price drops
stockIndexVolatility volatility in stock market prices 0 to 1*

bondSentiment overall positive references, net of negative references -1 to 1

bondOptimism optimism, net of references to pessimism -1 to 1
bondTrust trustworthiness, net of references connoting mistrust -1 to 1
bondFear fear and anxiety 0 to 1*
bondStress arousal and intensity, weighted towards distress 0 to 1*
bondSurprise unexpected events and surprise 0 to 1*
bondUncertainty uncertainty and confusion 0 to 1*
bondDefault debt defaults, late payments, and bankruptcy -1 to 1
bondPriceDirection bond price increase, net of references to price decrease -1 to 1
bondPriceForecast forecasts of bond price rises, net of references to forecasts -1 to 1
of asset price drops
bondVolatility volatility in bond and debt values 0 to 1*
centralBank country central bank references 0 to 1*
debtDefault debt defaults and bankruptcies in a country 0 to 1*

MarketPsych Indices 3.3 – User Guide 30

Document Version 1.1
 Country Markets

interestRates interest rates rising, net of references to rates falling -1 to 1

interestRatesForecast forecasts of interest rates rising, net of forecasts of rates -1 to 1
falling
monetaryPolicyLooseVsTight monetary policy being loose, net of references to monetary -1 to 1
policy being tight

MarketPsych Indices 3.3 – User Guide 31

Document Version 1.1
 Cryptocurrencies

CHAPTER 12 CRYPTOCURRENCIES
CRYPTOCURRENCY ASSETS
TRMI Cryptocurrencies data covers more than 150 active cryptocurrencies. Cryptocurrencies were selected for
inclusion based on listed market capitalization and/or technological significance. Coverage includes all cryptocurrencies
present in the coinmarketcap.com ranking of the top 100 cryptocurrencies by market capitalization starting in November
2017. Coverage is updated monthly to include new entrants from the top 100 list. Once covered, cryptocurrencies
remain in the feed until trading activity ceases.

A list of the covered cryptocurrencies and monthly changes in coverage is available on the MRN FTP site. See Chapter
14 for information on the files and how to access them.

CRYPTOCURRENCY TRMI INDICES

The 43 TRMI indices for the cryptocurrency asset class carry six significant digits past the decimal point. Negative
numbers have a leading minus (-) sign. The table below summarizes these fields. For more information on those with
range “0 to 1*”, see Chapter 3 above.
Index Description: references in news and social media to... Range
sentiment overall positive references, net of negative references -1 to 1
optimism optimism, net of references to pessimism -1 to 1
joy happiness and affection 0 to 1*
loveHate love, net of references to hate -1 to 1
trust trustworthiness, net of references connoting corruption -1 to 1
anger anger and disgust 0 to 1*
conflict disagreement and swearing net of agreement and conciliation -1 to 1
fear fear and anxiety 0 to 1*
gloom gloom and negative future outlook 0 to 1*
stress arousal and intensity, weighted towards distress 0 to 1*
surprise unexpected events and surprise 0 to 1*
timeUrgency urgency and timeliness, net of references to tardiness and -1 to 1
delays
uncertainty uncertainty and confusion 0 to 1*
violence violent crime, terrorism, and war 0 to 1*
emotionVsFact all emotional sentiments, net of all factual and topical -1 to 1
references
marketRisk positive emotionality and positive expectations net of negative -1 to 1
emotionality and negative expectations. Includes factors from
social media found characteristic of speculative bubbles –
higher values indicate greater bubble risk. Also known as the
“Bubbleometer.”
longShort buying, net of references to shorting or selling -1 to 1
longShortForecast forecasts of buying, net of references to forecasts of shorting -1 to 1
or selling

MarketPsych Indices 3.3 – User Guide 32

Document Version 1.1
 Cryptocurrencies

priceDirection price increases, net of references to price decreases -1 to 1

priceForecast forecasts of asset price rises, net of references to forecasts of -1 to 1
asset price drops
volatility volatility in market prices or business conditions 0 to 1*
priceMomentum currency price trend strength, net of references to trend -1 to 1
weakness
adoption acceptance and widespread adoption 0 to 1*
adoptionForecast forecasts of acceptance and widespread adoption 0 to 1*
anonymity user anonymity in transactions 0 to 1*
attack hacks and other cyberattacks related to the coin 0 to 1*
codeSentiment overall positive references to the code base, net of negative -1 to 1
references
codeUpgrade planned code upgrades and enhancements 0 to 1*
criminalActivity criminal activity involving the cryptocurrency or community 0 to 1*
developerSentiment overall positive references to the development team, net of 0 to 1*
negative references
fOMO fear of missing out 0 to 1*
fork fork in the blockchain 0 to 1*
forkForecast forecasts of a fork in the blockchain 0 to 1*
hodl holding on for dear life 0 to 1*
innovation innovativeness 0 to 1*
litigation litigation and lawsuits involving the cryptocurrency or 0 to 1*
community
majorityAttack majority of miners attacking the blockchain 0 to 1*
noobs amateur and naive investor activity 0 to 1*
regulatoryCrackdown regulatory crackdown on cryptocurrency-related activities 0 to 1*
regulatoryIssues tax or securities authorities intervention or legislation 0 to 1*
scam fraudulent activity or references to scams associated with the 0 to 1*
cryptocurrency
transactionSpeed fast versus slow transactions speeds for the cryptocurrency -1 to 1
vulnerability vulnerabilities or weaknesses in the code 0 to 1*

MarketPsych Indices 3.3 – User Guide 33

Document Version 1.1
 Constructing TRMI with Greater Window Lengths

CHAPTER 13 CONSTRUCTING TRMI WITH GREATER WINDOW LENGTHS

BACKGROUND
TRMIs are constructed internally as Buzz-weighted averages across various news and social media content sources.
Custom TRMI on greater window lengths can be constructed based on TRMI of shorter window lengths.

CREATING CUSTOM TRMI WINDOW LENGTHS

Users may construct custom TRMIs of varying window lengths from Buzz-weighted averages of minutely TRMI data.

For example, for a given company (assetCode), content source (dataType) and datetime (windowTimestamp), let Buzz0,
Buzz-1, ..., Buzz-(N-1) and TRMI0, TRMI-1, …, TRMI-(N-1) represent the corresponding Buzz and TRMI minutely data over
the trailing N minutes. Then the Buzz-weighted average TRMI over the trailing N-minute window length may be explicitly
calculated as:

(Buzz0*TRMI0 + Buzz-1*TRMI-1 + … + Buzz-(N-1)*TRMI-(N-1)) / (Buzz0 + Buzz-1 + … + Buzz-(N-1))

COMPARING CONSTRUCTED TRMI DATA TO ACTUAL TRMI DATA

Note that the daily (WDAI_UDAI) and hourly (WDAI_UHOU) TRMI data, which use a 24-hour/1440-minute trailing
1
window , cannot be reproduced perfectly using minutely TRMI data (W01M_U01M). This is because textual content is
aggregated in the daily and hourly feeds based in part on their online availability (publication timestamp). Articles are
excluded from the daily and hourly feeds 24 hours after publication, whereas content in the minutely feed is aggregated
solely by its time of acquisition by MarketPsych crawlers (crawl timestamp). See “Timing of Social Media Articles in
Archive” for more information about timestamps. The greater the delay between the publication and crawl timestamps,
the fewer 1440-minute TRMI windows will include that content. As a result, the Buzz of the artificially reconstructed 24-
hour TRMI will always match or exceed that of the actual 24-hour TRMI.

1
See “Daily Update Frequency Anomaly” section above for exceptions to when WDAI_UDAI data does not cover a
24-hour window.

MarketPsych Indices 3.3 – User Guide 34

Document Version 1.1
 MRN FTP Site: Data and Reference Files

CHAPTER 14 MRN FTP SITE: DATA AND REFERENCE FILES

OVERVIEW
Thomson Reuters provides live and archive TRMI data over FTP.
Data is formatted as tab-delimited text files. Archive files are compressed, in .zip format.
Production clients can access the full history, while trial clients may access a more limited period. Files for the two client
types are stored in different directories.
The FTP site also contains useful reference files for the Companies and Cryptocurrencies offerings:
 Listings of live and obsolete assets
 Mappings to common identifiers, including CUSIP, ISIN, and SEDOL for companies and coinmarketcap.com
tickers and name identifiers for cryptocurrencies.
 Changes in coverage

ACCESS
Production and most trial users are granted access to the FTP site. Trial users who convert to paid customers should
have their credentials upgraded to full-history access. Users should contact their sales specialist or account manager to
obtain login credentials.
Clients may use their credentials to connect via plain FTP and also secure FTP (FTPS) via explicit FTP over TLS, using
TLS version 1.2.
The FTP site is available at ftp://mrn-ftp.thomsonreuters.com, or 54.243.148.106. The site is accessible via FTP client
only. Please set your FTP client to passive mode, although in some cases active mode will work instead. FTP/S is
available on port 21, and data ports are 60022-60100.

WINDOW LENGTHS AND UPDATE FREQUENCIES

Chapter 2 explained that TRMI are produced over varying window lengths and update frequencies. The excerpted table
below shows which combinations are available.

Window Length Update Frequency FTP File Abbreviation

1 minute 1 minute W01M_U01M
1440 minutes / 24 hours 1 hour WDAI_UHOU
1440 minutes / 24 hours Daily, at 3:30 Eastern time WDAI_UDAI

LIVE FILES
Each live file contains all the scores for an asset class, for a certain window length and update frequency.

Non-zero Buzz Records and Files

Only rows with non-zero Buzz appear in these files. Moreover, if there are no non-zero Buzz rows to include in a file,
then that file will not be published.

Directory Structure
/TRMI_LIVE/{Asset Class}/{Time Abbreviation}/[LastTwoHours/LastTwoDays/Older]
 {Asset Class} can take one of the following values:
o CMPNY: all individual companies

MarketPsych Indices 3.3 – User Guide 35

Document Version 1.1
 MRN FTP Site: Data and Reference Files

o CMPNY_AMER: individual companies domiciled in the Americas

o CMPNY_EMEA: individual companies domiciled in Europe, the Middle East, and Africa
o CMPNY_APAC: individual companies domiciled in Asia and the Pacific
o CMPNY_GRP: company groups
o COM_AGR: agricultural commodities
o COM_ENM: energy & material commodities
o COU: countries
o COU_MKT: country markets
o CRYPTO: cryptocurrencies
o CUR: currencies
 {Time Abbreviation} describes the combination of window length and update frequency, as per the table above
o W01M_U01M
o WDAI_UHOU
o WDAI_UDAI
 [LastTwoHours/LastTwoDays/Older]
o LastTwoHours contains files that were published since the beginning of the previous calendar hour. For
example, if the current time is 18:32 UTC, then it will contain files published since 17:00 UTC. Unless
older files are needed, FTP clients may concentrate only on this subdirectory.
o LastTwoDays contains files, not present in LastTwoHours, that were published since the beginning of
the previous calendar day in UTC. For example, if the current date and time is 2016-03-04 18:32 UTC,
then it will contain files published since 2016-03-03 00:00 UTC.
o Older contains less recent files, not present in LastTwoHours or LastTwoDays, that were published
since the last monthly archive update. At each monthly archive update, all corresponding files in
/TRMI_LIVE are deleted.

File Naming
TRMI.Live.{Asset Class}.{Time Abbreviation}.{Time Period}.{System Version, converted}.txt, such that

 {Asset Class} as per above

 {Time Abbreviation} as per above
 {Time Period} will be in “yyyymmdd-hhmm” format. Note that boundaries are according to UTC and mark the
end of the content window.
 {System Version, converted}. The systemVersion value, “MP:3.0.0” will be converted to “0300”. Note that this
string’s length reaches five characters when the third number in the systemVersion exceeds 9, e.g., “3.0.10” will
be converted to “03010”.

Fields
See Chapter 5 for information on fields that are generic to all TRMI asset classes.

 id
 assetCode
 ticker: only for Companies. Note that ticker is generally not point-in-time, although it can vary across System
Version values.
 windowTimestamp
 dataType: “News”, “Social”, or “News_Social”
 systemVersion, excluding its prefix
 buzz

MarketPsych Indices 3.3 – User Guide 36

Document Version 1.1
 MRN FTP Site: Data and Reference Files

 non-generic indices: sentiment, optimism, et al.

Sort Order
 assetCode
 dataType

Accessing Old Live Files and Directory Cleanup

Live files remain available until a corresponding archive file has been uploaded. Thus there is typically between 1-2
months of data in the /TRMI_LIVE subdirectories.

ARCHIVE FILES
Each live file contains all the scores for an asset class, for a certain window length and update frequency.

Timing of New Archive Files

New archive files for the previous month are typically uploaded to FTP between 10 and 15 days after the close of that
month.

Timing of Social Media Articles in Archive

As background, there are three timestamps used in assigning timestamps to content.
 Publication time: a timestamp from the data itself that may be assumed to represent when the content was
published
 Acquisition time: MarketPsych-assigned timestamp, for when MarketPsych obtained the data. For some content
sources, the difference between the publication time and acquisition time is consistently ~1 second or less, but
for certain significant social media sources this can average in the tens of minutes. This large difference is
especially problematic for certain sources that MarketPsych did not collect live.
 Aggregation time: MarketPsych-assigned timestamp describing when the component TRMI scores from content
are finalized. For aligning live feed data and pro forma archives, aggregation time showed the best performance.
For all asset classes except cryptocurrencies, over different periods, the timestamp used for archive calculation
changed.
 1998 - February 2013: publication time
 March 2013 - August 2015: acquisition time. For TRMI data that uses a 1440-minute/24-hour window length,
the live feed and all archives are aggregated from article acquisition time to publication time plus 24 hours.
 September 2015 - present: aggregation time.
For cryptocurrencies, the publication time was used until April 1, 2018, at which time aggregation time is used for
archive calculation.

Using aggregation time, although single-minute differences between the live feed and archive buzz remain, they affect
well under 1% of buzz scores in the minute-window data.

Archive Backfills
As might be expected for data derived from numerous textual sources, the processing of MarketPsych Indices
undergoes intermittent changes to its content inputs and filtering algorithms. These changes generally are introduced in
the live data and in monthly archive updates, on a pro forma basis. Full-history TRMI archives are backfilled to reflect
these content changes on an occasional basis, often in tandem with a major version upgrade.

MarketPsych Indices 3.3 – User Guide 37

Document Version 1.1
 MRN FTP Site: Data and Reference Files

Non-zero Buzz Records

Only rows with non-zero Buzz appear in these files, except for daily updating files (WDAI_UDAI) for non-Companies
data.

Standard and Bulk Archive Files

Archive files come in two types. Standard archive files were designed so that they that can be viewed in Microsoft Excel
2007, which supports 1,048,576 rows. These files are stored in the “Historical” and “Recent” subdirectories described in
the next section.

For users who prefer to work with fewer files, archive files are also being provided in a small number of large files,
according to the capacity of what can be zipped. These files are stored in the “Bulk” subdirectory.

Bulk files contain scores on an entire asset class over a given time period. Standard files may cover only a single asset,
depending on the window length and update frequency.

Directory Structure
Because of the large number of Companies assets, the directory structure for Companies adds two or three extra levels
for organization.
/TRMI/{Asset Class}/{Time Abbreviation}/[Historical/Recent/Bulk] [[/{Country}/{Economic Sector code}] /{Asset Code }]

 {Asset Class} can take one of the following values:

o CMPNY: all individual companies
o CMPNY_AMER: individual companies domiciled in the Americas
o CMPNY_EMEA: individual companies domiciled in Europe, the Middle East, and Africa
o CMPNY_APAC: individual companies domiciled in Asia and the Pacific
o CMPNY_GRP: company groups
o COM_AGR: agricultural commodities
o COM_ENM: energy & material commodities
o COU: countries
o COU_MKT: country markets
o CRYPTO: cryptocurrencies
o CUR: currencies
 {Time Abbreviation} describes the combination of window length and update frequency, as per the table above
o W01M_U01M
o WDAI_UHOU
o WDAI_UDAI
 [Historical/Recent/Bulk]
o Historical contains all files from the previous calendar year, generally. Files are annual in duration.
o Recent contains files from the current year. Files are monthly in duration.
o Exception to the previous/current year rule: Recent will continue to hold files from the previous year that
just passed, until that year’s archives have been regenerated based on the most recent TRMI build.
o Bulk is an option for most asset class / time abbreviation combinations, as per the table below. This
directory contains larger files than the Historical and Recent directories. As a result, it has none of the
subdirectories listed below, and the files are different from the standard archive files. Note that 2017
files shall be standalone just as 2016 files are.

MarketPsych Indices 3.3 – User Guide 38

Document Version 1.1
 MRN FTP Site: Data and Reference Files

Bulk File Details

Asset Class Minutely: W01M_U01M Hourly: WDAI_UHOU Daily: WDAI_UDAI
1998-2004, 2005-2009, 2010-
Annual with monthly YTD 2011, 2013-2015, 2016, 2017
CMPNY updates Quarterly with monthly YTD updates with monthly YTD updates
1998-2000, 2001-2002, 2003-2004, 2005-
2006, 2007-2008, 2009-2010, 2011-2012, 1998-2012, 2013-2015,2016,
CMPNY_GRP None [ keep] 2013, 2014-2015, 2016, 2017 2017
1998-2012, 2013-2015,2016, 1998-2012, 2013-2015,2016,
COM_AGR 2017 1998-2012, 2013-2015,2016, 2017 2017
1998-2012, 2013-2015,2016, 1998-2012, 2013-2015,2016,
COM_ENM 2017 1998-2012, 2013-2015,2016, 2017 2017
1998-2012, 2013-2015,2016,
COU None None [keep] 2017
1998-2012, 2013-2015,2016, 1998-2012, 2013-2015,2016,
COU_MKT 2017 1998-2012, 2013-2015,2016, 2017 2017
CRYPTO 2009-2017 2009-2017 2009-2017
1998-2012, 2013-2015,2016, 1998-2012, 2013-2015,2016,
CUR 2017 1998-2012, 2013-2015,2016, 2017 2017

 Two intermediate subdirectory levels apply only to Companies non-bulk data

o {Country} is the two-letter ISO 3166-1 country code
o {Economic Sector} is a three-letter code corresponding to the TRBC Economic Sector.
 ENE: Energy
 MAT: Basic Materials
 IND: Industrials
 YCY: Cyclical Consumer Goods & Services
 NCY: Non-Cyclical Consumer Goods & Services
 FIN: Financials
 HLC: Healthcare
 TEC: Technology
 COM: Telecommunications Services
 UTL: Utilities
 {Asset Code }. This level only applies to Companies non-bulk data.
o This level is used for the minutely data only, due to the amount of data

File Naming
TRMI.Archive.{Asset Class}.[[.{Country}.{Economic Sector code}] /.{Asset Code }]{Time Abbreviation}.{Time
Period}.{System Version, converted}.zip/txt, such that
 {Asset Class} as per above
 {Country}.{Economic Sector code} as per above. This applies only to CMPNY WDAI_UHOU and WDAI_UDAI
Historical/Recent files.
 {Asset Code }. This level is only present for CMPNY W01M_U01M non-Bulk.
 {Time Abbreviation} as per above
 {Time Period} can take one of the following values, depending on its directory:
o Bulk: “yyyy-yyyy” for multi-year data, as per “Bulk File Details” table above
o Historical:
 “yyyy” format for annual data.

MarketPsych Indices 3.3 – User Guide 39

Document Version 1.1
 MRN FTP Site: Data and Reference Files

 “yyyyQq” format for quarterly data. This applies only to hourly (WDAI_UHOU) data.
o Recent: “yyyymm” format for monthly data
 {System Version, converted}. The systemVersion value, “MP:3.0.0” will be converted to “0300”. Note that this
string’s length reaches five characters when the third number in the systemVersion exceeds 9, e.g., “3.0.10” will
be converted to “03010”.

Archive File Structure

Each .zip file holds one .txt file that contains relevant archive data for that asset and time period, across all three Data
Types (News, Social, and combined).

Fields
See Chapter 5 for information on fields that are generic to all TRMI asset classes.
 id
 assetCode
 ticker: only for Companies. Note that ticker is generally not point-in-time, although it can vary across System
Version values.
 windowTimestamp
 dataType: “News”, “Social”, or “News_Social”
 systemVersion, excluding its prefix
 buzz
 non-generic indices: sentiment, optimism, et al.

Sort Order, Single-asset Files

 windowTimestamp
 dataType

Sort Order, Multi-asset Files

 windowTimestamp
 assetCode
 dataType

VERSION UPDATES
The file system version is updated in one of three digits. The first digit of the version is updated (e.g., from version 2 to
3) approximately every three years due to advancements in the NLP and overhauls in source constituents. Such major
versions are run in production in parallel while customers upgrade over 12 to 18 months. For the most recent major
versions, version 2 NLP was frozen in March 2014, and version 3 NLP was frozen in March 2017.

The second digits in the version number are updated if an urgent patch is made. For example, between versions
3.0/3.1 and 3.2/3.3 a source-level change affecting media acquisition led to excessive buzz in some assets. The
change was identified, its effect was quantified, and a minor version update restores the feed to regain consistency from
that source. The second digits are also updated when a forked sub-feed is launched. For example, the cryptocurrency
sentiment feed was forked off of v3.0 and launched as v3.1 due to minor NLP changes to account for cryptocurrency
source-specific linguistic differences.

The third digits of the version number are updated monthly to reflect the inclusion of IPOs, company additions,
corporate name changes, mergers, and buyout activity. These minor version updates include additional corporate

MarketPsych Indices 3.3 – User Guide 40

Document Version 1.1
 MRN FTP Site: Data and Reference Files

aliases and changes in corporate aliasing. Updates reflected in second and third version number digits do not reflect
NLP or text analysis alterations.

COMPANIES AND CRYPTOCURRENCIES REFERENCE FILES

Companies and cryptocurrencies reference files serve several purposes. The files list companies and cryptocurrencies
that can be found in the live or archive scores. They add supporting information on the company or cryptocurrency and
on its primary quote, In addition, for companies only, third-party identifiers – CUSIP, ISIN, and SEDOL – are also
available. CUSIP and ISIN require a license with Standard & Poor’s for CUSIPs. ISINs are included because some
ISINs are CUSIP-based. SEDOL access requires a license with the London Stock Exchange.

Please contact your Account Manager or Sales Specialist if you are interested in viewing these third-party identifiers for
companies and have a requisite license. Thomson Reuters will contact the identifier issuer(s) to verify the license(s).

Directories

Companies
/TRMI/CMPNY/[BASIC/CUSIPISIN/SEDOL/CUSIPISINSEDOL]/

Notes:
 The third level directory is permissioned according to user’s combination of licenses for third-party identifiers.
Access will be given to exactly one such directory. By default, users are granted access to the BASIC directory.

Cryptocurrencies
/TRMI/CRYPTO/[BASIC]/

Asset Lists and Mappings Files

Companies
TRMI.Companies.[BASIC/CUSIPISIN/SEDOL/CUSIPISINSEDOL].{System Version, converted}.txt

Notes:
 {System Version, converted}. The systemVersion value, “MP:3.0.0” will be converted to “0300”. Note that this
string’s length reaches five characters when the third number in the systemVersion exceeds 9, e.g., “3.0.10”
will be converted to “03010”.

Fields:
 PermID: Thomson Reuters organizational identifier.
 companyName
 countryOfDomicile: two-character ISO 3166-1 country code
 TRBCEconomicSector: plain-text description of Thomson Reuters Business Classification (TRBC) economic
sector
 status: “active” if the PermID may be scored in a live feed. Otherwise, “inactive”.
 RIC: main RIC for this company.
 ticker: Because ticker is in the score files, it shall be populated for all companies. Private companies shall
have a MarketPsych-designated ticker beginning with “PVT-“.
 marketMIC: ISO 10383 code for market or exchange identification. Value may differ from similar value
maintained by London Stock Exchange.
 CUSIP: only available in files with “CUSIPISIN” in the file name
 ISIN: only available in files with “CUSIPISIN” in the file name

MarketPsych Indices 3.3 – User Guide 41

Document Version 1.1
 MRN FTP Site: Data and Reference Files

 SEDOL: only available in files with “SEDOL” in the file name

Values on inactive companies shall attempt to represent their most recent values, including after delisting. Some values
may be blank.

Cryptocurrencies
TRMI.Crypto.[BASIC].{System Version, converted}.txt

Notes:
 {System Version, converted}. The systemVersion value, “MP:3.0.0” will be converted to “0300”. Note that this
string’s length reaches five characters when the third number in the systemVersion exceeds 9, e.g., “3.0.10”
will be converted to “03010”.
 For cryptocurrencies, the System Version begins at MP:3.1.05.

Fields:
 assetCode
 cryptocurrencyName
 status: “active” if the cryptocurrency may be scored in a live feed; otherwise “inactive”.

Changes File
The second is a file of changes since the previous System Version.

Companies
TRMI.Companies.Changes.{System Version, converted}.txt, such that

 {System Version, converted}. The systemVersion value, “MP:3.0.0” will be converted to “0300”. Note that this
string’s length reaches five characters when the third number in the systemVersion exceeds 9, e.g., “3.0.10” will
be converted to “03010”.

File contents:
 First line: “TRMI Companies list for systemVersion [systemVersion] was created on [yyyy-mm-dd].”
 Further lines:
o “PermID [xxxxxxxxxxx], [companyName], became active.”
o “PermID [xxxxxxxxxxx], [companyName], became inactive.”

Cryptocurrencies
TRMI.Crypto.Changes.{System Version, converted}.txt, such that

 {System Version, converted}. The systemVersion value, “MP:3.0.0” will be converted to “0300”. Note that this
string’s length reaches five characters when the third number in the systemVersion exceeds 9, e.g., “3.0.10” will
be converted to “03010”.

File contents:
 First line: “TRMI Cryptocurrencies list for systemVersion [systemVersion] was created on [yyyy-mm-dd].”
 Further lines:
o “assetCode [xxxxxxxxxxx], [cryptocurrencyName], became active.”
o “assetCode [xxxxxxxxxxx], [cryptocurrencyName], became inactive.”

MarketPsych Indices 3.3 – User Guide 42

Document Version 1.1
 MRN FTP Site: Data and Reference Files

TRIAL ACCESS COMPARED TO PRODUCTION ACCESS

As mentioned in the overview to this chapter, trial clients can access less data than can production clients, and from a
different directory. Thus the files are the same, but the time duration is less.
Following are the key differences in directory structure for trial clients compared to production clients:
 Root directory is /TRMI_TRIAL, instead of /TRMI
 The third-level directory denotes the amount of data
 The directory level of Recent and Historical is absent

MarketPsych Indices 3.3 – User Guide 43

Document Version 1.1
© 2019 Refinitiv All rights reserved. For product information or to make a sales
Republication or redistribution of Refinitiv content, enquiry, visit the Refinitiv website:
including by framing or similar means, is prohibited
without the prior written consent of Refinitiv. 'Refinitiv' and https://www.refinitiv.com/en/contact-us
the Refinitiv logo are registered trademarks and
trademarks of Refinitiv and its affiliated companies.
Read more about our products at:
https://www.refinitiv.com/en/products/world-
news-data

Document Version 1.0

Date of Issue: 23 May 2018