Exploring the impact of customer concentration on stock price crash risk

The study examines how customer concentration—reliance on a small number of customers for a significant portion of a firm’s revenue—relates to stock price crash risk in the context of Iran’s developing capital market. Prior findings on whether customer concentration increases or decreases crash risk are mixed. Grounded in agency theory, the paper argues that concentrated customer bases can alter managerial incentives, information disclosure, and risk, thereby affecting the propensity for bad news hoarding and eventual stock price crashes. Building on Lee et al. (2020), the paper extends the inquiry to an emerging market, introduces new concentration measures, and investigates whether institutional investors moderate these relations. The main research questions are: (H1) Is there a meaningful relationship between corporate customer concentration and stock price crash risk? (H2) Is there a significant negative relationship between government customer concentration and stock price crash risk? (H3) Does institutional investor presence affect the relation between corporate customer concentration and crash risk? (H4) Does institutional investor presence affect the relation between government customer concentration and crash risk?

The theoretical framing relies on agency theory, focusing on conflicts between principals (shareholders) and agents (managers) and how customer concentration may influence disclosure incentives and risk. Two competing views about corporate (non-government) customer concentration exist: (1) it heightens risk through dependence on a few customers, relationship-specific investments, increased buyer bargaining power, and higher operating leverage and uncertainty; (2) it reduces risk through operating efficiencies, lower costs (e.g., inventory and administrative), and improved performance due to stable relationships. For government customers, prior work generally suggests stabilizing effects for suppliers (lower bankruptcy risk, long-term contracts, cost-plus pricing). Empirical studies show mixed evidence: positive links between corporate customer concentration and risk/costs (e.g., cost of equity, tighter loan terms), but also improved performance and conservative accounting under higher concentration. Research on government concentration indicates more stable cash flows and lower cost of equity. Institutional investors’ role is also debated: they may enhance monitoring and conservatism (reducing crash risk) or align with management, weakening oversight and potentially increasing risk. This study addresses these mixed results by testing corporate and government customer concentration with multiple metrics and examining the moderating role of institutional investors in Iran.

Data and sample: 82 non-financial companies listed on the Tehran Stock Exchange (TSE) that had at least three years of activity between 2013 and 2020 and disclosed major customers (customers accounting for ≥10% of sales), yielding 656 firm-year observations. Financial institutions (banks, holdings, insurance, etc.) were excluded due to different operations and disclosure practices. Data were sourced from audited financial statements and board reports of TSE-listed firms. Empirical strategy: Panel data multiple regression using R. The study estimates baseline and moderation models with industry and year effects and standard controls. Model selection used Chow and Hausman tests; pooled OLS was used for Sales-based Index and HHI specifications, and fixed effects for the Ranking Index specification when indicated. Dependent variable (stock price crash risk): NCSKEW, the negative conditional skewness of firm-specific monthly returns within a fiscal year (following Chen et al., 2001; Kim et al., 2011). Monthly abnormal returns are obtained from a market model with leads/lags of market returns; demeaned log residuals form the basis of NCSKEW. Independent variables (customer concentration): Measured separately for corporate (CC) and government (GC) customers using three metrics: (1) Sales-based Index: sum of sales to significant customers (≥10% revenue) divided by total sales; (2) HHI: sum of squared revenue shares of significant customers; (3) Ranking Index: quintile ranks of concentration, where higher quintiles represent higher concentration. Government customer identification: A customer is considered government if the government or a state-owned enterprise (SOE) holds at least 50% of the customer’s shares. Moderator (institutional investors, IST): Measured as institutional ownership; operationalized as a dummy equal to 1 if the percentage of institutional shareholders is above the sample median and 0 otherwise; interacted with CC and GC to test moderation. Controls: SIZE (log assets), ROA, LEV, DTURNOVER (change in monthly turnover), SIGMA (SD of firm-specific abnormal returns), RET (average monthly firm-specific returns), MB, PCF (free float), CIM (dummy for crash risk above industry average). Models: Baseline (Model 1) regresses crash risk on lagged CC and GC with controls, industry and year effects. Moderation (Model 2) augments Model 1 with IST and interactions CC×IST and GC×IST. Statistical significance is assessed via t-tests and F-statistics; multicollinearity checked by VIF; autocorrelation by Durbin–Watson.

- Descriptive statistics: Mean crash risk (NCSKEW) = 0.307 (median 0.175). Institutional ownership averages 65% (0–99%). Mean corporate concentration (Sales-based Index) = 20%; mean government concentration = 33%. Mean LEV = 0.674; mean ROA = 0.083. - Model selection: For Sales-based Index and HHI, pooled models were used; for the Ranking Index, panel fixed effects were indicated (e.g., Hausman test significant; Table 6/9). - H1 (Corporate concentration): Using Sales-based Index and HHI, CC is not significant. Using the Ranking Index, the highest concentration tier (5th quintile) shows a significant negative relation with crash risk (e.g., p=0.015, t=−2.435, coef ≈ −1.312), indicating lower crash risk at the highest corporate concentration. - H2 (Government concentration): Using Sales-based Index and HHI, GC is not significant. Using the Ranking Index, there is a significant inverse association between higher government concentration and crash risk (text reports p≈0.011, t≈−2.556, coef ≈ −0.666), indicating lower crash risk at higher GC levels. - H3 (Moderation by institutional investors—corporate): Interactions CC×IST are positive and significant with Sales-based Index and HHI (p=0.003, t=2.999, coef=1.675; p=0.002, t=3.092, coef=1.899), implying institutional ownership strengthens the positive association between corporate concentration and crash risk in those specifications. In additional tests with the aggregate major-customer criterion and Ranking Index, higher-quintile interactions show significance (e.g., CCTA*IST p=0.049; CCT5*IST p=0.0387), supporting a moderating role. - H4 (Moderation by institutional investors—government): Interactions GC×IST are not significant with Sales-based Index and HHI (p=0.846; p=0.464) and are generally insignificant in Ranking Index specifications, indicating no discernible moderating effect for government concentration. - Model diagnostics: F-statistics indicate overall significance; adjusted R² around 0.11 in pooled models; VIFs <5 suggest no multicollinearity; Durbin–Watson near 2 indicates no serious autocorrelation.

The results address the research questions by showing that, in Iran’s developing market, very high corporate customer concentration correlates with lower stock price crash risk when measured via a ranking-based approach, aligning with the view that strong customer relationships can yield operational efficiencies and stability, thereby reducing bad news hoarding. For government customers, higher concentration also associates with lower crash risk, consistent with stabilizing long-term contracts and low bankruptcy risk on the government side. These findings diverge from some evidence in developed markets (e.g., Lee et al., 2020 on corporate concentration) and may reflect Iran’s distinct economic structure, regulatory environment, cultural and geopolitical conditions, and the nature of customer–supplier relationships. Institutional investors exhibit a significant moderating role for corporate concentration in several specifications, suggesting that ownership structure can shape disclosure incentives and monitoring, potentially influencing crash risk dynamics. However, institutional ownership does not appear to moderate the government concentration–crash risk link, possibly due to the unique features of government transactions and limited influence of external owners in that domain.

This study contributes to the literature by examining customer concentration–crash risk dynamics in a developing economy and by introducing three complementary measures of concentration, including a novel Ranking Index. Using a panel of 82 TSE-listed firms (2013–2020), it documents an inverse relation between very high corporate and government customer concentration and stock price crash risk, and finds that institutional investors moderate the corporate concentration–crash risk link but not the government concentration link. The results help reconcile mixed prior findings by highlighting metric choice and contextual differences across markets. Future research could incorporate richer measures of disclosure practices, governance mechanisms, and communication between institutional owners and management; examine political economy and sanction-related factors; and extend the analysis to other developing markets and sectors to test the generalizability of the observed patterns.

Key limitations include data constraints and reliance on publicly available disclosures, potential measurement error in identifying and quantifying significant customers, and the temporal and market-specific nature of the sample (Iran, 2013–2020), which may limit generalizability. The inconsistency across different concentration metrics underscores sensitivity to measurement choices. Future work should add variables on disclosure quality, governance, and owner–management communication, and consider broader contextual and macroeconomic factors.