Draft, not signed off
Draft last pushed: 31 August 2026, 08:02 AEST
Every figure on this page was read from the document named beside it on 31 August 2026. Figures already set out on the Active and passive page are cited here with one line and a link each and are not repeated. The page source carries the classification of each entry. Christoph signs off as licensee.
This page collects what the published account records report about how much trading costs the people who do it. Four documents carry the measurements: two studies of the same United States discount broker's household records in the 1990s, one study of the complete transaction record of a national stock exchange, and two reports in which the Australian Securities and Investments Commission (ASIC) counts what happened to the clients of one leveraged product. Each is a single study or a single collection, and each is written here so it cannot be read as a settled result.
The information is as of 31 August 2026. Every figure below was read from the document named beside it on that day.
Barber and Odean sorted the households at one United States discount broker into five equal groups by how much of their portfolio they turned over each month, and measured what each group earned. The most active group and its comparison against the market index are set out on the Active and passive page, with the paper's own conclusion about what explained the shortfall. Three figures from the same section of the paper are not there, and they are what makes the comparison a comparison. Before costs, the five groups earned almost the same: an investment copying the average household of each group would have earned a gross return, meaning the return before the bid-ask spread and commissions are taken out, of between 18.5% and 18.7% a year across all five. The bid-ask spread is the gap between the price at which a share can be sold and the higher price at which it can be bought at the same moment, and a trade crosses it. After costs the five groups separated: the least active group earned a net annual return of 18.5%, and the paper puts the cost of the most active group's trading, measured against what the least active households earned, at 57 basis points a month, or 6.8% a year. A basis point is one hundredth of a percentage point. The group the paper calls high turnover is the fifth of households turning over more than 8.8% of their portfolio a month.
Source Barber and Odean, "Trading Is Hazardous to Your Wealth: The Common Stock Investment Performance of Individual Investors", Journal of Finance 55(2), 2000, section IV and Figure 1, read from the authors' own posted copy on 31 August 2026. The same paper's other figures, and the sentence in which the authors say what explained the shortfall, are on Active and passive and are not repeated here.
The five groups are not five treatments. Households chose how much to trade, so the group a household is in is its own decision and the groups may differ in ways the records do not show. What the near-identical gross returns establish is narrower and firmer: whatever else separated the groups, it was not the shares they picked. The period is January 1991 to December 1996, six years in which a market index weighted by company size returned 17.9% a year, so the levels belong to that period even though the cost difference is arithmetic. The accounts are self-directed accounts at one discount broker in the United States, they hold common shares only, and mutual funds, options, warrants and depositary receipts were excluded from the analysis.
The same broker's records carry the gender of the person who opened each household's first account, and Barber and Odean used it to compare the two groups over February 1991 to January 1997. Of the 37,664 households where that person could be identified, 29,659 had accounts opened by men and 8,005 by women. Women turned over about 53% of their share portfolios a year and men about 77%, which the paper states as men trading 45% more than women. The paper's own summary of what that cost is that trading reduced men's net returns by 2.65 percentage points a year against 1.72 percentage points for women.
Where the difference arose is measurable in the same tables, because the paper compares each household against the portfolio that household held at the start of the year and reports the shortfall before and after costs. Before costs the two groups fell short of their own starting portfolios by almost the same amount, and the gap between them was 0.34% a year, which the paper reports as significant at the 1% level, so it is a real difference and a small one. After costs the gap was 0.94% a year, tested directly rather than inferred from two separate comparisons against zero, and also significant at the 1% level. So about two thirds of the measured difference appeared only once the cost of trading was taken out. The authors put it in one sentence: "Men lower their returns more than women because they trade more, not because their security selections are worse." The annualised geometric mean returns, which are the steady yearly rates that would have turned each group's starting value into its ending value, tell the same story from the other side: before costs men earned 18.7% and women 18.6%; after costs men earned 16.3% and women 16.9%. Among single households the same differences were larger: single men traded 67% more than single women and reduced their returns by 1.44 percentage points a year more.
Source Barber and Odean, "Boys Will Be Boys: Gender, Overconfidence, and Common Stock Investment", Quarterly Journal of Economics 116(1), 2001, pages 261 to 292, abstract, section III.A and Table II, read from the authors' own posted copy on 31 August 2026.
One study, one broker, one six-year period, and the comparison is between accounts rather than between people: the recorded gender is that of the person who opened the household's first account, which need not be the person who placed the trades. The authors treat that as a known weakness and answer it by splitting the sample on marital status, which is why the single households are reported separately. The two groups are also of very different sizes, 29,659 against 8,005. The measurements are of how much each group traded and of what that trading cost; whether the difference in trading is explained by confidence in one's own judgement is the paper's hypothesis and its own reading of the result, and no measurement of confidence appears in the records.
Barber, Lee, Liu and Odean analysed the Taiwan Stock Exchange's own transaction record for 1992 to 2006: 3.7 billion two-sided transactions with the identity of each trader attached, which is the whole market rather than the customers of one firm. A day trader is defined in the paper as an investor who buys and sells the same share on the same day. In the average year of that period about 450,000 Taiwanese individuals day traded, and day trading was 17% of all volume on the exchange, a share the paper describes as stable across the fifteen years. The share of day traders who lost money is set out on the Active and passive page beside the equivalent figures for two other markets, and is not repeated here. Two further results are what this record adds. Of the 277,000 individuals in the average year who day traded more than 600,000 New Taiwan dollars, about 20,000 United States dollars, roughly 20% earned positive abnormal returns net of commissions and transaction taxes in that year, a share the authors report as varying between 17% and 20% depending on how the population is defined. And when the authors ranked traders on one year's returns and then watched the next year, the paper's own sentence is that "less than 1% of the day trader population is able to predictably and reliably earn positive abnormal returns net of fees". The top 500 of the previous year's ranking went on to earn 37.9 basis points a day after fees, and the bottom-ranked group minus 28.9 basis points a day.
Source Barber, Lee, Liu and Odean, "The cross-section of speculator skill: Evidence from day trading", Journal of Financial Markets 18, 2014, pages 1 to 24, abstract and sections 1 and 3, read from the authors' own posted copy on 31 August 2026.
One market, one activity and one fifteen-year period. Taiwan charges a tax of 30 basis points on sales, so the cost a Taiwanese day trader had to clear is not the cost an Australian investor faces. The returns are computed on the assumption that the value of a trade is the trader's capital at risk, which the authors say gives returns of the right sign but understated in size. The paper measures day trading, which is buying and selling the same share within one day; it does not cover investors who hold for longer periods, and the after-fee figure for the top-ranked group depends on the commission rate assumed, which the paper states as a range.
Two statements are usually quoted alongside these measurements: one about a large fund manager's review of accounts that nobody was trading, and one giving a round-number share of short-term traders who lose. Both are set out on the Active and passive page, with what a search of the record finds behind each and what the published loss rates for defined groups of traders actually are. Neither is repeated here.
A contract for difference (CFD) is an agreement to exchange the change in the price of an asset without owning the asset, and it is normally traded with leverage, meaning the client puts up a fraction of the position's value and carries the gain or loss on the whole of it. ASIC counts the outcomes of the clients of the firms that issue these contracts, and it has now published two counts on either side of a rule change. In August 2019, from data collected in its 2017 review of the retail over-the-counter derivatives sector, it reported that 72% of clients who traded contracts for difference lost money, together with 63% of clients who traded margin foreign exchange and 80% of clients who traded binary options. ASIC then made a product intervention order, and its report of January 2026 counts the 2023 to 2024 financial year. The headline figures from that count, and the equivalent published figures for the United Kingdom and for Europe, are set out on the Active and passive page. Three figures in the same report are not there. Alongside the clients who made a net loss, 57,183 clients, or 29.27%, made a net profit, totalling $172.33 million after $26.49 million of fees. The outcome differed by what was traded and by how the client arrived: ASIC's own summary reports that "in FY 24, 85% of retail clients made a net loss trading options CFDs", and that of new retail clients acquired through paid online advertising, 74% lost money.
Sources ASIC, Report 626, "Consumer harm from OTC binary options and CFDs", August 2019, whose three percentages carry ASIC's own note that the data was collected in its 2017 review of the sector; and ASIC, Report 828, "Risky business: Driving change in CFD issuers' distribution practices", 20 January 2026, executive summary, Table 1 and Appendix 2, Table 2. Both were read from ASIC's own download addresses on 31 August 2026.
The two counts are not a controlled before-and-after comparison. They cover different periods, different sets of licensees and, in Report 626's case, three products rather than one, and the rule change was not the only thing that differed between 2017 and 2024. Each count is also of one financial year rather than of a group of clients followed over time, and Report 828's own table of new-client activity shows the number of active new retail clients falling from 8,193 in the first quarter of that year to 2,694 in the fourth, so a one-year count is not a count of what happened to any client over the life of an account. Contracts for difference are a leveraged derivative sold to retail clients, and nothing in either report describes ordinary share investing.
No Australian study was found that sorts retail investors by how much they trade and reports what each group earned, on brokerage records, share registry records or any other Australian source. Nor was one found that compares Australian accounts the way the 1991 to 1997 study above compares them. What exists for Australia in this area is the regulator's count of one leveraged product, which is the entry above. The questions the field could answer with records that already exist are set out in the questions section of the State of Research page.
Source that no such Australian study was found is this practice's own search, of 31 August 2026.
Built by In Your Interest Financial Planning. In Your Interest Financial Planning Pty Ltd, ABN 28 094 300 464 is Authorised Rep. No 308161 of Fiduciary Duty Advisers Pty Ltd AFSL No 527434. Nothing on this page is advice to you or to your clients, and nothing on it is a recommendation of any product, any strategy or any adviser. Corrections are welcome and wanted – contact us.