Controlling endogeneity using instrumental variables approach
| Variables | First stage | Second stage | ||||
|---|---|---|---|---|---|---|
| LP_QUALITY | RP_QUALITY | (1) | (2) | (3) | (4) | |
| NCSKEW | DUVOL | COUNT | CRASH | |||
| AVE_ LP_QUALITY | 1.312*** (0.437) | |||||
| AVE_ RP_QUALITY | 0.127*** (0.041) | |||||
| PRED_LP_QUALITY | −0.886*** (0.306) | −0.783** (0.356) | −1.892*** (0.610) | −4.389*** (1.463) | ||
| PRED_RP_QUALITY | −0.678** (0.295) | −0.599** (0.285) | −1.447** (0.629) | −3.356*** (1.119) | ||
| C and control variables | Yes | Yes | Yes | Yes | Yes | Yes |
| Year FE | Yes | Yes | Yes | Yes | Yes | Yes |
| Industry FE | No | No | No | No | No | No |
| N | 2,898 | 2,898 | 2,898 | 2,898 | 2,898 | 2,898 |
| R2 | 0.234 | 0.241 | 0.218 | 0.035 | 0.080 | |
| Adj_R2 | 0.219 | 0.226 | 0.202 | 0.015 | 0.062 | |
| Pseudo R2 | 0.084 | |||||
| F | 15.460*** | 16.110*** | 14.130*** | 1.829*** | 4.423*** | |
| LR Chi2 | 274.260*** | |||||
| Variables | First stage | Second stage | ||||
|---|---|---|---|---|---|---|
| (1) | (2) | (3) | (4) | |||
| 1.312 | ||||||
| 0.127 | ||||||
| Yes | Yes | Yes | Yes | Yes | Yes | |
| Year | Yes | Yes | Yes | Yes | Yes | Yes |
| Industry | No | No | No | No | No | No |
| 2,898 | 2,898 | 2,898 | 2,898 | 2,898 | 2,898 | |
| 0.234 | 0.241 | 0.218 | 0.035 | 0.080 | ||
| Adj_ | 0.219 | 0.226 | 0.202 | 0.015 | 0.062 | |
| Pseudo | 0.084 | |||||
| F | 15.460 | 16.110 | 14.130 | 1.829 | 4.423 | |
| 274.260 | ||||||
This table presents the results of the two-stage least squares (2SLS) analysis using industry-average partner quality (AVE_LP_QUALITY and AVE_RP_QUALITY) as instruments. The first-stage estimates confirm strong instrument relevance, while second-stage coefficients for PRED_LP_QUALITY and PRED_RP_QUALITY remain negative and significant, indicating robustness to endogeneity. All models include control variables and year fixed effects. Robust standard errors are in parentheses. Statistically significant coefficients are shown in italics for ease of reference. *, ** and *** denote statistical significance at the 10, 5 and 1% levels, respectively
Sharing content requires targeting cookies to be enabled. Please update your cookie preferences to use this feature.