Information quality and the divergence of IPO returns: evidence from Thailand

Abstract

This paper examines whether prospectus information quality is associated with IPO underpricing and long-run stock performance among 325 Thai IPOs listed on the Stock Exchange of Thailand and the Market for Alternative Investment from 2012 to 2022. Information quality is measured using LLM-assisted scores for use-of-proceeds specificity and risk-factor disclosure specificity, together with a financial reporting quality proxy based on discretionary accruals, and the tests use OLS regressions with control variables, industry fixed effects, and year fixed effects. For underpricing, the evidence is narrow: Legal Risk Number is associated with a 9.87 percentage-point lower first-day return at the 1 percent level, while other risk categories, use-of-proceeds measures, and financial reporting quality do not provide broad support for H1 at the 1 percent or 5 percent levels. For long-run performance, firms above the idiosyncratic-risk specificity threshold exhibit three-year abnormal returns 25.89 percentage points higher than firms below the threshold at the 1 percent level, and Idiosyncratic Risk Percentage is also positive and significant at the 5 percent level. H2 is therefore only partially supported, mainly through idiosyncratic risk disclosure at the three-year horizon, while tax risk is negative and significant at BHAR36 and financial reporting quality is not significant across the BHAR models. The moderation hypotheses H3a, H3b, and H3c are not supported as broad claims because the interaction results vary by category, measurement variant, and horizon; these estimates are conditional associations and should not be interpreted as causal effects.