Does Overvaluation Predict Earnings Management? Evidence on the Agency Costs of Overvalued Equity in a Frontier Market

Most of the accounting literature on emerging markets asks whether capital markets reward high-quality earnings. This paper reverses the question and asks whether market valuation shapes subsequent reporting. Using 4981 firm-year observations for 619 non-financial Vietnamese listed firms, with misvaluation measured from 2016 to 2024 and earnings management from 2017 to 2025, and measuring misvaluation through the Rhodes-Kropf et al. decomposition alongside five Tobin’s Q-based measures, we find that overvaluation in one year predicts more aggressive earnings management in the next. The estimate is positive and significant at the one percent level for all six measures in specifications that absorb firm and year fixed effects, control for prior accrual behaviour, and use standard errors clustered by firm. A one within-firm standard deviation increase in the firm-specific error raises absolute discretionary accruals by 6.6 percent of their within-firm standard deviation. The accruals are directional, and the relation runs through magnitude rather than incidence: overvaluation predicts how large income-increasing accruals are once reported, but neither the probability that they are reported nor the probability of falling in the upper tail of the accrual distribution. The relation weakens as leverage rises, which is consistent with creditor monitoring; formal interaction tests do not support the corresponding predictions based on firm size or state ownership. A firm-block bootstrap that reproduces both first-stage models raises the standard error by 23 percent but leaves the confidence interval clear of zero, and randomisation inference across 1999 permutations places the estimate well outside the placebo distribution. The evidence is substantially stronger for within-firm variation than for cross-firm comparison: in a propensity-score matched sample the estimate is not distinguishable from zero and covariate balance is unsatisfactory. The design is observational and does not identify a causal effect. Because equity-based compensation and an active market for corporate control are largely absent in this setting, the findings indicate that the association does not depend on the channels through which price pressure is conventionally assumed to operate, although they do not identify the channel that does operate.