A stochastic approach to analysing mismatch losses and therefore the potential energy increase of module-level optimisation, as a result of PV module degradation, is presented in this paper. By considering the PV module as a random vector and making use of flash tested degradation data to estimate a continuous multivariate probability density function, the mismatch loss can be estimated as a function of time by computing its average or expected value. This is shown to gradually increase over time as the modules degrade further away from one another and is estimated to be around 5% of the total available power after 30 years of PV plant operation. Whether this loss of energy justifies the investment in optimisation depends on the exact cost in comparison to acquiring new modules. Given the current performance of degraded commercial PV modules, and if acquiring new modules is prohibitively expensive for a specific PV plant, re-using old modules with module-level optimisation may be a viable option if it can be done at a lower cost. Re-using old modules in different applications, instead of disposing or recycling, will open a second-hand market where homeowners and small businesses can use old utility scale PV modules, together with module-level optimisation to ensure maximum efficiency.