Shorting Indonesia's SRI-KEHATI stocks ahead of the Indonesia Stock Exchange's (IDX) 31 July 2026 green-label announcement earned nothing measurable against the composite index. Against the IDX ESG Leaders index, the short earned about 1.6 points of relative return over the two preceding sessions, gross of costs. We calculate that 1.6 by adding the two reported daily mean abnormal returns, the return an equal-weighted short would have captured. Verbie and Ivander report -0.008 at t-2 and -0.008 at t-1 against ESG Leaders. Against IHSG, the corresponding figures are +0.002 and -0.004. Only the ESG Leaders results reach significance in their paper. Which benchmark the authors subtract determines the finding.
A new label for twenty-two already recognised stocks
The Indonesia Stock Exchange created a Green Equity Designation through Board Decree Kep-00110/BEI/07-2026 on 31 July 2026. The designation is an exchange-level classification attached to an issuer's corporate status. Previously, membership in the SRI-KEHATI Index was the main sustainability label on the IDX. The authors ask whether the new classification gives investors fresh information about firms that already hold the old one.
Their mechanism comes from signaling theory. Prices and turnover should respond if investors regard an exchange certificate as credible news about access to green capital or environmental risk. Information already reflected in prices should produce no response.
The calculation is straightforward: subtract the day's IHSG composite return from each stock's daily return. Without an estimation window, this gives every stock an implied beta of one. Turnover is shares traded divided by shares outstanding. The authors average it across stocks and call the measure ATVA. Their universe starts with the 25 SRI-KEHATI constituents effective 2 June to 30 November 2026. Two firms leave the sample because of corporate actions, and one because of a material announcement, leaving 22. The event window spans 11 trading days, t-5 to t+5.
The event announces a policy. A firm receiving the label is a separate event. We did not find an identification in the paper of which of the 22 firms applied for or received the designation, or when. The test therefore asks whether ESG-recognised stocks repriced when the framework was published, before any firm is known to hold the label.
Did prices or turnover respond?
Against IHSG, no.
Daily average abnormal returns run from -0.004 to 0.007, with every one-sample t-test p-value between 0.124 and 0.915. Three days fail Shapiro-Wilk and are tested with Wilcoxon instead. Their p-values are 0.199, 0.833 and 0.858. On the announcement day, the return is -0.002 (t = -0.651, p = 0.522).
For the cumulative test, the authors compare cumulative abnormal return (CAR) over t-5 to t-1 with CAR over t+1 to t+5. They report a mean difference of -0.002, with t = -0.158 and p = 0.876. They describe the effect size of -0.034 as below "the standard threshold for a meaningful economic effect." Their next claim is stronger: "The negligible effect size further demonstrates that the regulatory announcement did not induce a material valuation shift."
Our confidence-interval argument addresses that sentence. The effect-size number is standardized; the confidence interval carries the economic content. The 95% interval for the pre-minus-post CAR difference extends from -2.8 to +2.4 points. The test cannot rule out a gap that large. With 22 stocks, a reaction of up to 2.8 points remains undetected.
Turnover produces the same answer. The ATVA Wilcoxon test gives z = -1.445 and p = 0.149. Both periods have mean ATVA of 0.001, although the maximum increases from 0.003 to 0.006.
The benchmark creates the pre-event result
Replacing IHSG with the IDX ESG Leaders index makes two pre-event days significant. The mean at t-2 is -0.008 (raw p = 0.001, Holm-adjusted 0.011). At t-1, it is -0.008 (raw 0.004, Holm 0.040). The positive t+3 return of 0.015 has a raw p of 0.007. The authors correctly discard it after Holm adjustment (0.063). On the event day, average abnormal return (AAR) against the ESG benchmark is 0.004 (p = 0.255). The corresponding CAR difference is -0.021 (p = 0.105).
Compare the two benchmarks at t-2. The 22 stocks average +0.002 in abnormal return against IHSG at t-2, versus -0.008 against ESG Leaders. Both columns use identical stock returns. Their one-point gap comes entirely from ESG Leaders outperforming IHSG by about a point that day. Most of the significant "reaction" therefore comes from the benchmark's move. At t-1, the benchmark gap is 0.4 points (-0.004 versus -0.008). Against IHSG, t-1 ties with t-4 at -0.004 for the weakest day. It also has the main test's smallest p-value (0.124), yet remains insignificant.
The authors acknowledge the disparity, writing that it "reflects the presence of benchmark sensitivity", then defend the substitution. They argue that IHSG "tends to induce a masking effect that obscures specific abnormal return signals". ESG Leaders, in their account, "isolates ESG-specific risk factors" through "a comparison against a peer group of companies sharing similar ESG risk profiles."
The arithmetic at t-2 answers that defence. Stock returns do not change between columns; the extra 1.0 point at t-2 comes from ESG Leaders rising about a point more than IHSG that day. A move across the whole ESG benchmark is a factor move. It tells us nothing firm-specific about the designation. If sample stocks also belong to the benchmark, subtracting it partly differences out the sample itself.
The authors apply Holm separately within each benchmark, correcting 11 tests at a time. Pooling all 22 daily tests changes the outcome. The second-smallest raw p-value, t-1's 0.004, is multiplied by 21 to give 0.084. Only t-2 would survive.
Clustering adds another problem. All firms share one calendar date, yet the cross-sectional t-test treats the 22 abnormal returns as independent draws. A common sector or index shock can then count as 22 confirmations. Five of the 25 SRI-KEHATI constituents are banks. If large banks also belong to ESG Leaders, some of the sample would sit inside its own benchmark. This is our conjecture: the paper neither lists ESG Leaders members nor discusses overlap.
In the abstract, the authors say the pre-event returns "indicate market anticipation rather than immediate announcement shock." Two sessions of relative underperformance against one benchmark cannot support that interpretation. The paper offers no news or leak around t-2. Its framework allows a negative reaction through compliance or adjustment costs. Yet the authors call the negative pre-event days "anticipation" without identifying what investors anticipated or why it would depress prices.
A smaller bookkeeping issue remains. The descriptive table reports a t+1 minimum of -0.003 and a mean of 0.000, figures that fit poorly with the reported confidence interval of -0.008 to 0.007. Given a maximum of 0.041, observations in that range cannot produce a standard deviation above about 0.011. The table nevertheless reports 0.017. The text also describes that 0.000 day as negative.
What remains for a desk trading 22 stocks
The IHSG findings are consistent with no detectable reaction. Within that narrow interpretation, they support the authors' main claim that an exchange green label "conveys limited incremental information" for stocks with existing ESG visibility. The authors also conclude that "regulatory labeling alone has limited short-term valuation impact unless accompanied by direct economic instruments." The clause about incentives is a recommendation beyond what these data can test. No economic instrument varies across the 22 firms.
The pre-event result is weaker than the abstract suggests. It depends on one alternative benchmark and two days, with correction performed per specification and 22 correlated observations. None of the results is reported net of costs. None is out of sample.
We could not test this ourselves. The study requires point-in-time SRI-KEHATI membership together with Indonesian stock-price and index data. We have none of those data. A US ESG proxy would not reproduce an Indonesian exchange decree, so we did not run one.
One result would change our view: rerunning the same window on firms that actually received the designation, against matched non-designated IDX stocks. If the designated firms underperform their matched non-designated peers at t-2 and t-1, anticipation becomes a hypothesis worth trading around.