Avoid increasing trading during market booms and crashes
Shanghai Stock Exchange data covering roughly 40 million accounts during 2014–2015 found that the smallest 85% of household accounts earned RMB 250 billion less through active trading than by holding unchanged—around 30% of their starting stock value. Comparable gaps in calmer periods were only 1%–3%.
What it takes
Free. Resisting extra exposure while others appear to be making money is difficult.
What you may gain
Daily data covered all approximately 40 million exchange accounts from July 2014 to December 2015, 18 months. The Shanghai Composite first rose over 150%, peaked at 5,166.35 on June 12, 2015, then fell 40% by year-end. Household accounts were divided into four starting-value groups. The smallest group contained 85% of accounts and the largest 0.5%; their initial total stock holdings were similar, RMB 880 billion and RMB 808 billion. Relative to holding unchanged, the smallest group lost RMB 250 billion through trading and the largest gained RMB 254 billion: 28% and 31% of initial holdings. In any 18-month window during the calmer two-and-a-half years before June 2014, comparable gaps were only 1%–3%. During the boom/crash, households turned over holdings about every three weeks, nearly 18 times annually. The main period is Shanghai 2014–2015, with 2012–2014 comparison data.
Context & considerations
The study does not say never buy stocks. It compares active trading with unchanged holdings during the same market, finding transfers from smaller to larger individual accounts. The beneficiaries were not institutions; retail accounts generated about 90% of exchange turnover then. The common claim that 90% of retail investors lose money lacks a checkable official statistic and is not used. Compare nearly 18 annual turnovers here with 75% in the US trading study. This is not investment advice.
Research & references
An L, Lou D, Shi D (2022). Wealth redistribution in bubbles and crashes. Journal of Monetary Economics 126:134-153. https://doi.org/10.1016/j.jmoneco.2022.01.001 ; Authors' October 2021 public working paper, used to check the detailed numbers. https://personal.lse.ac.uk/loud/AnLouShi.pdf