The pension system's internal rate of return (IRR) synthetically compares benefits received over the entire retirement period to all contributions paid during a person's working life. Within generations, it makes it possible to see if the system truly performs the vertical and horizontal redistribution expected of it, to the benefit of underprivileged groups. At the intergenerational level, its development is limited by demographic and economic variations, yet it makes it possible to see how the burden of adjustments is distributed between successive generations. The "Destinie 2" microsimulation model is used to illustrate both of these different uses. We first compare the rate of return of the generations from 1950 to 1980. Without reforms, the return would have remained at a much higher level than its equilibrium level. The reforms brought it closer to this equilibrium level, but to a degree that remains dependent on economic growth assumptions. One of the factors contributing to this decrease in return is the move towards the indexation to prices, rather than wages, for a certain number of parameters: this caused growth in pension rights to slow down, but to a lesser extent considering the slow economic growth, which impacts the system's financial balance. Within a single generation, the differences in IRR are a reflection of the presence of non-contributory mechanisms and of the differences in life expectancy. The latter have an anti-redistributive effect when comparing returns according to the age at the completion of studies; however, for the 1960-1970 generations studied here, this does go so far as to nullify the redistributive impact of rules which benefit lower incomes and short careers. In addition, reversion improves the return for couples, when the joint return for both members of these couples is considered.