Temporal aggregation is seen as an intuitively appealing forecasting strategy that reduces uncertainty by transforming the series into lower frequencies allowing one to identify the time series characteristics better. The effect of temporal aggregation on the stock control, via inventory costs and service levels, has been analyzed empirically and its benefits are discussed in the literature. However, there is little theoretical support identifying when temporal aggregation can improve supply chain utility. Herein, we analytically investigate the impact of non-overlapping temporal aggregation on supply chain inventory and production costs. For an ARMA(1,1) demand process, forecasted with the MMSE approach, we model a supply chain consisting of a retailer and a manufacturer. We calculate the inventory and production costs at both stages using non-aggregate and non-overlapping aggregated demand. They reveal how temporal aggregation affects supply chain costs and when each approach is superior. We provide suggestions about when to use temporal aggregation for minimizing supply chain costs.