TL;DR
Get business pricing on office and shipping supplies
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
A Swiss National Bank working paper by Severin Joho and Enzo Rossi examines 1,063 speeches by senior SNB officials from 1928 through 2024. It reports rising speech frequency, shorter and more accessible language, and greater emphasis on inflation in the flexible exchange-rate era; the authors also compare conventional Python text analysis with ChatGPT.
Severin Joho and Enzo Rossi have published a Swiss National Bank working paper analyzing 1,063 speeches by senior SNB officials from 1928 to 2024. The report finds that speeches became more frequent, shorter and more accessible over time, while inflation gained prominence in policy discussion during the era of flexible exchange rates.
The authors say they assembled and processed a new dataset spanning 96 years, then examined speech frequency, length, readability and subject matter. Their findings indicate that annual speech frequency rose from the mid-1950s onward, with a sharper increase after Switzerland moved to flexible exchange rates. The report does not provide a specific speech count for each year in the supplied summary.
The analysis also finds a long-term shift toward shorter speeches and language that has become more accessible since the 1960s. On subject matter, the authors report that inflation moved to the foreground in the flexible-rate period, while cyclical economic developments and unemployment received less emphasis. These are findings about the themes and language in the speech collection, not claims that the underlying economic conditions changed in the same way.
As part of the study, Joho and Rossi compared traditional Python-based text mining with analysis using ChatGPT. The authors describe the approaches as complementary and discuss methodological lessons, algorithmic pitfalls and open questions for future research. The supplied report summary does not give detailed performance scores or a direct measure of how often the two methods reached different conclusions.
How SNB Communication Changed
The findings offer a long-run view of how a central bank’s public communication has changed in both form and focus. More frequent speeches can mean officials address audiences more often, while shorter, more accessible language may alter how policy explanations reach the public. The paper documents patterns in the speeches; it does not establish that these changes caused shifts in public understanding or market behavior.
The thematic results may also help readers interpret the SNB’s institutional communication across different monetary regimes. The reported prominence of inflation in the flexible-rate era gives researchers a way to examine how policy narratives correspond to changing frameworks. The comparison of ChatGPT and Python methods is relevant to that work: choices about text-analysis tools can affect how researchers identify themes and describe changes over time.
central banking monetary policy history book
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
A Century of SNB Speeches
The dataset covers speeches by high-ranking Swiss National Bank officials from 1928 through 2024. That span includes major changes in Switzerland’s monetary framework, including the transition to flexible exchange rates, which the paper identifies as a point after which speech frequency rose more sharply. The supplied material does not specify the exact date or define the transition in further detail.
The authors frame the project as both a historical analysis and a methods comparison. Conventional Python text-mining tools apply defined, repeatable procedures, while generative AI systems can interpret language probabilistically. The working paper says these approaches can complement each other, while warning that LLM analysis has algorithmic pitfalls and that important methodological questions remain open.
“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, in the working paper summary
natural language processing Python textbook
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Limits of the Reported Findings
The supplied summary does not include the paper’s detailed tables, statistical tests, or exact year-by-year measures for speech frequency, length and readability. It also does not explain how the authors define accessibility or quantify the prominence of each topic. Those details would be needed to assess the size and robustness of the reported trends.
The summary says the study benchmarks Python-based methods against ChatGPT but does not provide the model version, prompts, evaluation criteria or comparative results. It is therefore unclear from the available information how closely the systems’ outputs aligned, which method performed better on particular tasks, or how the authors addressed possible errors in generated analysis.
text analysis and computational linguistics book
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Further Testing of Language Models
The working paper identifies open questions for future research on using large language models to study central bank communication. Further work could test whether the reported patterns hold under different model settings and text-analysis procedures, and publish more detail on how the findings were measured. The supplied source does not announce a follow-up study or set a timetable for additional research.
Readers can consult the full paper through the SNB’s working-paper publication page for its methods and supporting results. Until those details are examined, the headline findings should be read as the authors’ analysis of this specific speech collection and period.
As an affiliate, we earn on qualifying purchases.
Key Questions
What did the SNB working paper study?
It analyzed 1,063 speeches by high-ranking Swiss National Bank officials delivered from 1928 to 2024, examining frequency, length, readability and themes.
What changes did the authors report?
They report that speeches became more frequent from the mid-1950s, with a faster rise after the shift to flexible exchange rates; speeches also became shorter and more accessible. Inflation received greater thematic emphasis in the flexible-rate era.
Did the study compare ChatGPT with other methods?
Yes. The authors benchmarked ChatGPT against Python-based text-mining methods and describe the approaches as complementary. The supplied summary does not include detailed comparative scores.
Does the report show that SNB speeches changed public or market behavior?
The summary reports patterns in speech frequency, language and themes. It does not establish that those patterns caused changes in public understanding, market behavior or economic outcomes.
Source: primary
Fall Picks
fall essentials
As an affiliate, we earn on qualifying purchases.
