Investor Memory

Summary: The paper studies how memory shapes investor beliefs and behavior. We provide experimental evidence of a systematic memory bias. One week after observing investment outcomes, individuals over-remember gains and under-remember losses from their investment choices. The memory bias affects beliefs and subsequent investment decisions. Individuals form overly optimistic beliefs about their investment compared to a normative benchmark and re-invest even when doing so leads to a lower expected return. We further implement several treatments to provide evidence on the underlying mechanism driving the memory bias. Our findings are in line with the notion of a motivational mechanism involving self-image concerns. Importantly, individuals only exhibit the memory bias if they actively chose the investment. In contrast, individuals who did not choose their investments or did not invest but merely observed outcomes do not have a memory bias. Our findings contribute to the understanding of how people learn from experience in financial markets. Further, the documented memory bias offers a consistent explanation for stylized facts about investor behavior, such as stock repurchase mistakes and persistent investing of unsuccessful investors, as well as dynamic risk taking in many economic domains.