MICADO will likely be the first-light near-infrared instrument at the ELT, offering high-angular and high-spectral resolution (R=20,000) capabilities. As part of the MICADO consortium and the MORFEO science working group, we have developed an end-to-end simulation framework tailored to MICADO's unique observing modes. In practice, we simulate exoplanet observations in pupil-tracking mode using the single-conjugate adaptive optics (SCAO) mode and the long-slit with the H+K band filter. As input, we use realistic PSFs generated with the MISTHIC code (Huby et al. 2024), including effects such as atmospheric turbulence, dispersion, and static aberrations. We have named this MICADO data product a "slit-temporal scan" (STS). To reduce and extract a spectrum from an STS, we optimized a spectral deconvolution algorithm that effectively suppresses stellar speckles. This algorithm will be released as open-source software on GitHub to support the community in preparing for first light. In this talk, I will present the simulation framework, our latest results, and how these simulations will guide target selection and observing strategies—focusing on questions such as: How close to the host star, and at what contrast can we extract planetary spectra? What physical processes in the atmospheres of young giant planets can we expect to probe with MICADO?