Institutional trust is understood as the attitude formed in public perceptions toward government policies and performance and, according to the OECD (2017) framework, is defined along two core dimensions: alignment of government actions with the public interest and competence in service delivery. This study examines institutional trust as reflected among Telegram users in response to government policies in Iran’s capital market during the period 2020–2022, a timeframe marked by severe volatility and structural shocks in the market.

Using deep learning–based sentiment analysis, a Balanced Institutional Trust Sentiment Index (BITSI) is constructed. Subsequently, the asynchronous relationship between capital market fluctuations and digital institutional trust is analyzed using time-series data and the Autoregressive Distributed Lag (ARDL) model.

The findings indicate that the BITSI index, derived from the analysis of 2.5 million Telegram posts, remained predominantly at negative levels, reflecting persistent distrust toward economic institutions and policymakers. ARDL results confirm that changes in the stock market index and exchange rate volatility exerted statistically significant and relatively stable effects on users’ digital institutional trust. In contrast, the net inflow and outflow of retail investors’ capital did not demonstrate a stable or significant effect, which can be attributed to the technical nature of this variable and the limited public understanding of such financial flows.

Accordingly, the study concludes that rebuilding institutional trust requires more than short-term support for the stock market; it depends on transparency, regulatory stability, and, in particular, continuous monitoring of digital institutional trust to prevent the reproduction of cycles of distrust.