TL;DR
Get your next haul delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
A Swiss National Bank working paper by Severin Joho and Enzo Rossi analyzes 1,063 speeches by senior bank officials from 1928 through 2024. It reports that speeches became more frequent, shorter and more accessible, while inflation gained prominence in policy discussion during the flexible exchange-rate era.
Swiss National Bank researchers Severin Joho and Enzo Rossi have published a working paper examining how the bank’s public language changed across a century, using a dataset of 1,063 speeches delivered from 1928 to 2024. The analysis finds that speeches became more frequent, shorter and more accessible over time, while inflation moved to the forefront of policy discussion in the flexible exchange-rate era.
The paper, published by the SNB on October 23, 2026, examines the length, readability and themes of speeches by high-ranking central bank officials. The researchers describe the dataset as a comprehensive collection spanning 1928–2024. Their findings trace a long-run rise in annual speech frequency beginning in the mid-1950s, with that trend accelerating after the shift to flexible exchange rates.
The authors also report a clear movement toward shorter speeches and language that has become more accessible since the 1960s. In their thematic analysis, inflation receives greater emphasis in the flexible-rate period, while cyclical economic development and unemployment receive less. The paper describes these as patterns in the bank’s communication; they do not, by themselves, establish why the topics or style changed.
To assess how text-analysis methods perform on central-bank material, Joho and Rossi compare traditional Python-based text mining with analysis using ChatGPT. Their methodological discussion presents probabilistic language models and deterministic Python frameworks as complementary approaches, while drawing attention to algorithmic pitfalls and questions for further research.
How SNB Communication Changed
The findings offer a long historical record of how a central bank presents policy and economic issues to the public. More frequent speeches mean officials’ public remarks became a more regular part of the SNB’s communication over the period studied; shorter and more accessible language point to changes in the form of that communication. The paper documents these trends rather than claiming that any single change caused them.
The shift in topics also matters because public speeches can show which economic concerns receive attention in different policy settings. The study reports that inflation became more prominent after the move to flexible exchange rates, while economic cycles and unemployment received less emphasis. That is a finding about the content of speeches, not a direct measure of policy decisions or the importance of those issues to the Swiss economy.
The methods comparison is relevant beyond this dataset. Researchers and institutions increasingly use language models to process large collections of documents. The authors’ emphasis on combining ChatGPT with established Python methods—and on recognizing each approach’s limitations—speaks to how such analyses can be tested and interpreted.
natural language processing with Python book NLTK spaCy
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
A Century of SNB Speeches
The study covers speeches by high-ranking Swiss National Bank officials over 96 calendar years, from 1928 through 2024. Rather than focusing on one recent policy address, it tracks broad changes in speech frequency, length, readability and subject matter across several monetary-policy periods.
One dividing point highlighted in the report is the transition to flexible exchange rates. The authors say annual speech frequency rose from the mid-1950s and increased more sharply after that transition. The supplied report summary does not give the exact date of the change, year-by-year speech counts, or numerical readability measures, so those details cannot be specified here.
The working paper is identified by the SNB as Working Paper 2026/13. Its scope includes both historical findings and guidance on applying language-processing tools to central bank communication, linking the study’s empirical record to a broader methodological question.
“The frequency of annual speeches has followed an upward trajectory since the mid-1950s, accelerating notably after the transition to flexible exchange rates.”
— Severin Joho and Enzo Rossi, authors of the SNB working paper
text mining and computational text analysis textbook
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Limits of the Findings
The supplied summary does not provide the full statistical results, specific counts by year, or the measures used to quantify readability and thematic emphasis. It also does not give a detailed account of how closely ChatGPT and Python-based methods agreed, or which method performed better for particular tasks. The paper’s conclusions should therefore be read as the authors’ reported trends, not as independently verified numerical comparisons in this article.
The findings concern the content and form of speeches, not whether public communication changed policy outcomes or how audiences understood the remarks. The summary also does not identify the precise date of the flexible-exchange-rate transition used to mark the comparison, or explain the causes of the observed changes in speech frequency and topic selection.
central banking monetary policy economics book
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Further Work on Language Models
The paper points to open questions for future research on using large language models to analyze central bank communication. The source material does not identify a specific follow-up study, release date or planned SNB project. Readers seeking the detailed methods and results can consult the bank’s published Working Paper 2026/13.
Further evaluation would need to clarify how the different text-analysis approaches compare across tasks and how researchers can account for algorithmic pitfalls. Until those details are assessed, the paper’s methodological message is that generative models can complement conventional text-processing frameworks rather than simply replace them.
Python programming for data analysis reference book
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
What did the SNB researchers study?
They analyzed 1,063 speeches by high-ranking Swiss National Bank officials delivered from 1928 to 2024, examining frequency, length, readability and themes.
What changed in the speeches over time?
The authors report that annual speeches became more frequent, speeches grew shorter, and language became more accessible from the 1960s onward.
How did the topics change?
According to the paper, inflation gained prominence in the flexible-exchange-rate era, while cyclical economic development and unemployment received less emphasis.
How did the study use ChatGPT?
The researchers benchmarked ChatGPT against traditional Python-based text-mining methods to examine how the approaches could be used to analyze central bank communication.
Does the paper show that speech changes affected policy?
No such causal finding is included in the supplied summary. The paper reports patterns in speeches and discusses analytical methods; it does not establish that changes in wording or frequency caused policy outcomes.
Source: primary
Halloween Picks
halloween
As an affiliate, we earn on qualifying purchases.
