Curated Measures

Political Risk and Sentiment

Measures of firm-level political risk (PRisk) and sentiment (PSentiment) similar to Hassan et al. (2019).

Updated Jun 26, 2026 9 series 1,910 viewers 4 downloaders

Sign in to download curated-measure data

Preview

You can browse series and charts without an account. Downloads become available after you sign in, based on your plan.

  • Plans at a glance: Free: selected Curated Measures included. Standard: time-series for every series, firm panels at 1 credit per series. Research: time-series and firm panels for every series.
Create free account Sign in Compare plans

Overview

This page is the home of the maintained, quarterly-updated continuation of the firm-level political risk measure introduced in Hassan, Hollander, van Lent, and Tahoun (2019, Quarterly Journal of Economics). The scoring follows closely what is done in the paper; the data extends it from 2002 through 2026-Q1 and is refreshed quarterly. NL Analytics is the authors' vehicle for ongoing updates and the original firmlevelrisk.com release points here for current data.

If you already know the paper, this page is written to answer the two questions you actually have:

  1. Is this the same series?
  2. How do I get the updated version for my own work?

If you arrived from firmlevelrisk.com

The free archive there ends at 2021q2. Everything from 2002 forward — including the years that archive already covers — is available here as a single, internally consistent panel, so your analysis runs on one continuous series rather than splicing a free file onto a paid one. The historical period is recomputed under the same methodology as the update, so there is no break in construction at the point where the free data stops.

How closely it matches the published series

Computed on the firm-quarters that overlap with the firmlevelrisk.com release (2002–2021q2):

Series Firm-quarter (gvkey × quarter) Aggregate (quarter)
PRisk 0.91 0.98
PSentiment 0.86 0.95
Topic-specific PRisk 0.81 – 0.91 0.95 – 0.99

The table shows the Pearson correlation with the published series. N=310,303 and 81 for the firm-quarter and aggregate series, respectively. Topic measures span Trade (lowest) to Economy (highest) at both levels.

The firm-quarter figures are relevant for cross-sectional and panel work; the aggregate figures show the two series tracking almost identically through time. We document remaining differences between them below; they reflect sample composition and tokenization at the call level — which average out when aggregated — not any difference in how the measure is computed. Correlation tables by year and by topic are available on request.

> Every number in this table is reproducible: the script in our public GitHub repository, https://github.com/nlanalyticsinc/prisk-correlations, regenerates the full set from two inputs — the full firm-level panel from this page, and the original published series, which you can download as a CSV from firmlevelrisk.com.

How to get it

Each series' firm-level panel — PRisk/PSentiment and the eight topic measures — can be unlocked individually with an export credit on the Free and Standard ($50/mo) plans, or comes included in full on Research ($2,000/quarter, student pricing on request), which includes panel exports across the whole Curated Measures library and your own custom searches. Time-series for this measure is included on Standard and Research

Non-commercial academic use of the historical archive remains free on firmlevelrisk.com. This page is for the updated, maintained, citable vintage.

Explore the data

PRisk and PSentiment (v2.1) over time

Date range: January 1, 2002 to June 15, 2026.

Time-series chart for PRisk and PSentiment (v2.1) over time.
Loading figure...

Explore series

Series

9 total Data through Jun 15, 2026

Browse the related series and open any series for its full query, filters, and methodology notes.

No series match your filter.

Methodology and data

What is the same, and what is different

The measure is computed as in the paper. Verified directly against the original implementation, the following are identical:

  • The scoring for PRisk, PSentiment, and the eight topic-based PRiskT series (Economy, Trade, Tax, Security, Institutions, Health, Environment, Technology).
  • The counting window: ±10 words around a risk term for PRisk (±20 words for the topic measures), with the center bigram's own tokens excluded — as in the paper.
  • The political bigram library and its tf·idf weights, the risk/uncertainty synonym set, and the sentiment dictionaries.

The differences that remain are peripheral. They affect which text enters the count and how scores are scaled and not how the measure is defined:

  • Tokenization. Word boundaries are drawn slightly differently (punctuation-split, digits retained). A small number of multi-word phrases do not match under the current tokenizer.
  • Sample composition. The mapping to GVKey differs from the original release. This is likely the main source of the remaining firm-level discrepancy — and it averages out under aggregation, which is why the aggregate series correlate far more tightly than the firm-quarter ones (below).
  • Minor filter and edge-case rules. No minimum-length filter on very short transcripts, and one boilerplate safe-harbor rule is handled slightly differently.
  • Denominator and normalization. Division by word count rather than bigram count, library-weight normalization for PRiskT, and multiplying the measures by 1/100,000, are scale-only: they do not affect correlations or firm rankings.

Data

Every row carries identifiers you may already use — gvkey, cusip, isin, ticker, company name, and headquarters country. Each row is one earnings call, carrying both the calendar date and start time, so you can aggregate to firm-quarter on whichever convention your panel uses and handle multiple calls per quarter explicitly. Exports are available as CSV.

References

Hassan, T. & Hollander, S. & van Lent, L. & Tahoun, A., "Firm-Level Political Risk: Measurement and Effects," The Quarterly Journal of Economics, 134(4) (2019)

NL Analytics. (2026). Political Risk and Sentiment [Data set]. NL Analytics. https://apps.nlanalytics.tech/curated-measures/political-risk/

Back to Curated Measures