<![CDATA[Penelitian terdahulu tentang carbon emission disclosure (CED) dan kinerja keuangan masih didominasi pengujian linier pada periode yang sama; efek yang berubah menurut tingkat disclosure dan respons keuangan tertunda jarang diuji bersama, khususnya pada sektor energi Indonesia. Penelitian ini menguji pengaruh CED terhadap return on equity (ROE) pada periode berjalan, hubungan nonlinier berbentuk U, dan efek lag satu tahun. Kontribusinya membedakan CED sebagai lapisan transparansi dari environmental performance serta menguji dimensi level dan waktu dalam two-way fixed effects. Panel seimbang mencakup 17 perusahaan sektor energi BEI periode 2019–2024 (102 observasi); model lag menggunakan 85 observasi. CED diukur dengan indeks 18 item. Estimasi menggunakan fixed effects perusahaan dan tahun dengan firm-clustered standard errors. CED periode berjalan tidak signifikan. Koefisien kuadratik membentuk arah U secara matematis, tetapi komponen linear dan kuadratik tidak signifikan secara individual maupun bersama, sehingga pola U tidak terbukti. Efek lag satu tahun juga tidak signifikan. Dalam batas sampel, periode, ROE, dan ukuran CED yang digunakan, keluasan disclosure belum berkaitan secara statistik dengan profitabilitas. Prior studies of carbon emission disclosure (CED) and financial performance remain dominated by contemporaneous linear tests; level-dependent effects and delayed financial responses are rarely examined jointly, particularly in Indonesia's energy sector. This study tests the effect of CED on return on equity (ROE) in the current period, a U-shaped nonlinear relationship, and a one-year lagged effect. Its contribution is to distinguish CED as a transparency layer from environmental performance and test level and timing within two-way fixed effects. The balanced panel covers 17 IDX energy-sector firms during 2019–2024 (102 observations); lagged models use 85 observations. CED is measured with an 18-item index. Estimation employs firm and year fixed effects with firm-clustered standard errors. Current CED is insignificant. The quadratic coefficients form a U-shaped direction algebraically, but the linear and quadratic terms are insignificant individually and jointly, so the U-shape is not supported. The one-year lagged effect is also insignificant. Within the sample, period, ROE proxy, and CED measure used, disclosure breadth is not statistically associated with profitability. ]]>
