Abstrak Pergerakan reaksi pasar dan rasio keuangan beserta Economic Value Added menjadi topik hangat terutama dengan berkembangnya pasar modal di tanah air. Melalui SEM, model yang dihasilkan mampu mengkontruks indikator-indikator rasio keuangan terhadap pergerakan saham. SEM merupakan pemodelan kuantitatif faktor-faktor yang menunjukkan hubungan sebab akibat antara beberapa faktor dependen dan independen melalui indikator-indikatornya. Analisis SEM merupakan kombinasi dari analisis faktor (Confirmatory Factor Analysis), analisis jalur (Path Analysis) dan analisis regresi. Untuk mendapatkan model yang lebih baik, maka analisis ini dipadukan dengan mengkorelasikan error berdasarkan Modification indices. Modification indices akan mengakibatkan terjadinya penurunan Chi-square serta terjadi perubahan nilai CMINDF dan RMSEA menjadi semakin baik. Begitupula pada p-value, GFI, dan TLI. Sehingga dapat disimpulkan bahwa pengaruh korelasi antar measurement error dalam variabel rasio keuangan dalam variabel reaksi pasar dan antar variabel EVA mengakibatkan perubahan yang signifikan pada kebaikan model. Kata Kunci: Reaksi pasar, modification indices, SEM Abstract Market reaction movements and financial ratios and also the Economic Value Added are becoming hot topics, especially with the development of capital markets in the our country. Through SEM, the resulting model is able to construct indicators of financial ratios to stock movements. SEM is a quantitative modeling of factors that show a causal relationship between several dependent and independent factors through its indicators. SEM analysis is a combination of factor analysis (Confirmatory Factor Analysis), path analysis (Path Analysis) and regression analysis. To get a better model, this analysis is combined by correlating errors based on Modification indices. Modification indices will cause a decrease in Chi-square and changes in CMINDF and RMSEA values will be better. Likewise on p-value, GFI, and TLI. So that it can be concluded that the effect of the correlation between measurement errors in financial ratio variables in the market reaction variable and between EVA variables resulted in a significant change in the goodness of the model.