TL;DR
A Swiss National Bank working paper by Severin Joho and Enzo Rossi analyzes 1,063 speeches by senior SNB officials delivered from 1928 through 2024. The authors report that speeches became more frequent, shorter and more accessible over time, while inflation gained prominence in policy discussion during the flexible-exchange-rate era.
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, and that inflation took a more prominent place in policy rhetoric in the flexible-exchange-rate era.
Get business pricing on office and shipping supplies
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
The paper, published on the SNB website on October 23, 2026, examines changes in the annual frequency of speeches, their length and readability, and the themes officials addressed. The dataset covers nearly a century of public remarks by high-ranking central bank officials, giving the authors a basis for comparing communication across different monetary-policy periods.
According to the authors, the annual number of speeches followed an upward trend from the mid-1950s, with a more marked acceleration after Switzerland moved to flexible exchange rates. They also report a clear shift toward shorter speeches and increasing accessibility of language from the 1960s onward. In their thematic analysis, inflation became more prominent during the flexible-exchange-rate era, while cyclical economic development and unemployment received less emphasis.
The researchers compared traditional, Python-based natural-language processing with ChatGPT as a generative AI method for analyzing the texts. Their paper presents the approaches as complementary and discusses methodological lessons, including algorithmic pitfalls and open questions for future research. The source summary does not provide detailed numerical estimates for the reported shifts.
How SNB Communication Changed
The findings offer a long-run view of how a central bank’s public communication has changed alongside its policy environment. More frequent speeches can indicate a broader public-facing role, while shorter and more accessible language points to changes in how officials communicate policy. The paper reports those patterns across its dataset; it does not establish that any one policy change caused them.
The thematic shift also matters for understanding the SNB’s public messaging. The authors find that inflation gained prominence in the flexible-exchange-rate era, while unemployment and cyclical developments were less emphasized. These are findings about the content of speeches, not direct measures of policy priorities or the economic effects of communication.
The methodological comparison is relevant beyond this dataset. Central banks and researchers increasingly use automated tools to examine large collections of documents. The authors’ emphasis on combining deterministic Python methods with probabilistic language models underscores that AI-generated analysis should be checked against transparent methods and the underlying texts.
Popular right now

Bose QuietComfort Ultra Headphones Gen 1, Lunar Blue

Bose QuietComfort Ultra Wireless Noise-Cancelling Over-Ear Headphones with Spatial Audio, White Smoke

Ninja Foodi 10-Qt DualZone XL Air Fryer, 6-in-1, Grey (DZ401)
As an affiliate, we earn on qualifying purchases.
A Century of SNB Speeches
The study assembles speeches delivered by high-ranking Swiss National Bank officials between 1928 and 2024. Its long time span allows the authors to track changes in communication across major shifts in the monetary-policy setting, including the transition to flexible exchange rates cited in the paper.
The authors assess not only how many speeches were delivered but also their length, readability and subject matter. They then benchmarked conventional Python-based text mining against ChatGPT. The report’s stated aim is both to describe the evolution of SNB rhetoric and to identify practical considerations for using large language models to study central-bank communication.
“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, working paper summary
Limits of the Reported Findings
The source summary does not give exact annual speech counts, word-length measures, readability scores or topic shares, so the size of each change cannot be assessed from the summary alone. It also does not identify which specific speeches or officials account for the patterns, or provide detailed results comparing ChatGPT with the Python-based methods.
The reported timing of changes does not by itself show that the shift to flexible exchange rates caused more speeches or changed their subject matter. The authors describe correlations and thematic patterns in the speech record; the summary does not establish causal explanations. Further detail on the methods, model prompts and validation procedures would be needed to evaluate the comparison fully.
Further Research on AI Methods
The paper identifies open questions for future research on using large language models to analyze central-bank communication. The next step for readers assessing its claims is to consult the full working paper for the underlying measures, historical comparisons and validation details. The source material does not announce a follow-up study or a planned change to SNB communication policy.
Key Questions
What does the working paper examine?
It analyzes 1,063 speeches delivered by high-ranking Swiss National Bank officials between 1928 and 2024, examining frequency, length, readability and themes.
What changes did the authors report?
The authors report that speeches became more frequent from the mid-1950s, especially after the move to flexible exchange rates; they also became shorter and more accessible. Inflation became more prominent in the thematic discussion, while cyclical economic development and unemployment received less emphasis.
How did the researchers use ChatGPT?
They benchmarked ChatGPT against traditional Python-based natural-language processing to study the speech collection. The authors describe the methods as complementary and discuss algorithmic pitfalls; the provided summary does not include detailed comparative scores.
Does the paper show that flexible exchange rates caused these changes?
No causal conclusion is stated in the source summary. It reports that some speech trends accelerated or became more prominent after the transition, but timing alone does not establish that the policy change caused them.
Where can readers find the full report?
The working paper is listed in the research publications section of the Swiss National Bank website, under the 2026 working papers.
Source: primary
