questions about Leuz 2003 earnings management measure

[Zihao Zhou]

Based on Leuz 2003, there are 4 measures of earnings management. I have done the first three, but the last one confused me.
Leuz 2003 said: "Following Burgstahler and Dichev (1997), the ratio of ‘‘small profits’’ to ‘‘small losses’’ is computed, for each country, using after-tax earnings scaled by total assets.Small losses are defined to be in the range [-0.01, 0.00) and small profits are defined to be in the range [0.00, 0.01]. In order to reliably compute this ratio, we require at least five observations of small losses for a country to be included in the sample"

It is easy to compute small loss and small profits as followings:
gen ni_at=ni/at
gen small_loss=ni_at if ni_at>-0.01 & ni_at<0
replace small_loss=0 if small_loss==.
gen small_gain=ni_at if ni_at>0 and ni_at<0.01

However, since then, I am not sure how to compute the ratio of small gains to small losses for each country, because if the industry year report a small gain then it could not be a small loss. Moreover, I am not sure how to compute the ratios with keeping at least 5 observations of small losses for each country to be included in the sample. Should I sum all the small gains or loss by loc?

Please help me with this thanks!

[Nick Cox]

You need to count over countries, perhaps

egen total1 = total(ni_at >=  -0.01 & ni_at < 0), by(country) 

egen total2 = total(ni_at >=  0 & ni_at  <= 0.01), by(country)

and from there can calculate ratios, checking for small values, especially zero denominators.

I've stuck to the usual definitions of closed and open intervals, etc. -- compare your code. That said, exact equality with 0.01 or its negation is elusive as neither value can be held exactly in binary.

Perhaps there needs to be a time variable in there too. I am not familiar with this literature. Please note also our comments on bare name (date) references that assume we're all in the same sub-sub-field. https://www.statalist.org/forums/help#references Here the references don't seem crucial as you define what you want but that's not always true.