Cash transfer programmes have spread rapidly as an instrument to raise household consumption and reduce poverty. Questions remain if cash transfers can foster investments in productive assets in addition to raising immediate consumption among the very poor, particularly in rural Africa. Further, can asset investment gains be sustained over the medium-term after transfer programme termination? We use quasi-experimental methods combining difference-in-differences with matching on baseline observables to estimate the impacts of a bundled transfer programme on asset levels of beneficiary households 18 months after project completion. The results suggest that small regular cash transfers combined with enhanced saving mechanisms can generate asset accumulation among the extreme poor. In this case, the bundled intervention induces sustained investments in livestock assets. These gains occur among the poorest households and are not driven by gains among less poor beneficiaries. A lasting increase in participation in tontines (rotating saving groups) is also found among beneficiary households, suggesting bundled saving enhancement activities may contribute to observed asset gains.