The existence of irreversible demand is tested, whereby price increases induce a different absolute magnitude of quantity change than price decreases. Irreversibility is potentially likely in retail food settings for storable products that are consumed regularly and can affect pricing strategy performance. If irreversibility exists, the subsequent research question for storable product demand is whether loss aversion effects dominate stockpiling effects, or vice versa. A two-period theoretical model is developed, which predicts more elastic responses to downward price movements via stockpiling, but empirical tests on secondary data are needed to evaluate offsetting loss aversion effects. A variant of the Rotterdam demand model is developed to allow differential response to price increases and decreases. The model is applied to scanner data of short periodicity (weekly in this case), which are necessary to measure meaningful demand responses to food price changes. The products selected are U.S. cheeses and table spreads that are storable over multiple weeks. The results suggest that stockpiling dominates loss aversion. One potential cause of this behavior may be that marketers asymmetrically provide consumers with more reference price information when lowering prices, but not when raising prices. When stockpiling effects dominate, given the typically price-elastic store-level demand for food products, high-low pricing strategies should produce higher revenue. Regarding measurement of average demand response, reversible demand models applied to weekly data may overestimate own-price elasticities.