Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • How to run cross-sectional rgression to find residuals

    This is my equation:
    𝐼𝑛𝑣𝑖,𝑑 = 𝛽0 + 𝛽1π‘†π‘Žπ‘™π‘’π‘ πΊπ‘Ÿπ‘œπ‘€π‘‘β„Žπ‘–,𝑑-1 + πœ€π‘–,𝑑

    where: 𝐼𝑛𝑣𝑖,𝑑 denotes total investment; π‘†π‘Žπ‘™π‘’π‘ πΊπ‘Ÿπ‘œπ‘€π‘‘β„Žπ‘–,𝑑-1 denotes the percentage of change in sales; and i and t are subscripts of firms and years, respectively. Both variables are deflated by lagged total assets (TA) to control for heteroskedasticity.

    The authors describe their calculation process like this:

    "The study performs cross-sectional regression of equation (3.1) for each year and industry.

    The residuals of equation represent the deviation of the actual investment from the expected investment level and are considered proxies for investment inefficiency. If the residual is positive, it means that a firm is making more investments than the expected level based on sales growth. In other words, the firm overinvests. If the residual is negative, it implies that a firm is making less investment than the expected level, reflecting an underinvestment problem. To obtain the dependent variable (denoted as IE), the absolute value of the residuals of equation (3.1) is multiplied by minus one (IE=-|πœ€π‘–,𝑑|). Therefore, a higher value of IE means higher efficiency of investment."

    May I ask what I need to do to extract the result of investment efficiency (=residual)? Thank you.

  • #2
    Code:
    help regress postestimation##predict

    Comment

    Working...
    X