The costs of overtrading
When investors switch between asset classes (marketing timing), they do not do this very well. Equally, when they choose to sell one company and replace it with another in their portfolio, they also tend not to do very well. A strategy of choosing an asset allocation, and sticking with it in a well-diversified way seems to be a powerful and effective strategy over long periods of time. Too much trading is the opposite of a passive style strategy, and it is interesting to review some information around the way we trade.
How much trading is there in the Australian market?
It is interesting to consider the level of trading on the ASX. According to trading economics, in 2020 the portfolio turnover of the Australian market was 71.3%. Given that there is much spoken about the value of ‘buy and hold’ investing, it is fascinating to see how much of the market is traded every year.
Some evidence from individual investors
Barber and Odean are two researchers who look at the trading behaviour of individual investors, using a database of more than 60,000 accounts held with a discount broker. In analysing returns from investors from 1991 to 1996, in an article titled Trading is Hazardous to your Wealth published in the Journal of Finance they spit the accounts into five groups, quintiles, based on how often the portfolios were traded.
Returns were discussed as ‘gross’ returns and ‘net’ returns. Gross returns were the returns before taking into account fees, while net returns were returns after fees (the bid-ask spread and ‘commissions’, or brokerage).
It is worthwhile considering the difference in portfolio turnover between the 20% of portfolios that traded the least, trading around 0.19% of the portfolio per month, a rate that would take around 44 years to see the whole portfolio turnover over once, and the 20% of portfolios that traded the most, trading at least 8.7% of the portfolio per month, turning over the whole portfolio about every year.
The distinguishing factor of the research? All 5 quintiles based on trading had relatively similar gross returns (returns before looking at costs), between 1.47% and 1.55% per month using the raw data. However, the net returns (after costs) are where there becomes a clear difference.
Looking at the raw returns data, the monthly returns from the highest trading to lowest trading portfolios were:
Highest trading quintile = 1.01% per month
2nd highest trading quintile = 1.27% per month
3rd highest trading quintile = 1.36% per month
4th highest trading quintile = 1.41% per month
Lowest trading quintile = 1.47% per month
Various methods were used to calculate the net (after cost) returns of the portfolios, and they found that the high turnover quintile of portfolios underperformed the low turnover quintile by 5.5% pa to 9.6% pa.
The authors discussion of their results included the idea that overconfidence led to excessive trading, and refer to a famous Carhart (1997) study of managed funds that found managed funds also saw a negative relationship between higher trading and net returns – it is not just a phenomenon of induvial investors.
The study can be found here.
One of the costs of trading that we don’t immediately see is around tax. If you have a shareholding that was purchased at $5,000 and has increased to $10,000 you have the benefit of $10,000 working for you – dividends on the $10,000 holding and any future capital growth from the $10,000 value. However, as soon as you sell that holding you have to pay capital gains tax. The $10,000 holding will have $5,000 in capital gains. Assuming that the asset has been held for more than 12 months the 50% discount will mean a taxable gain of $2,500. For an investor paying tax at the 32.5% tax rate that is $812.50 of tax paid. By selling that $10,000 holding you have to pay $812.50 in tax, leaving a reduced $9,187.50 in the portfolio.
Unrealised capital gains in portfolios are a wonderful gift – unlike income tax, GST, superannuation contributions tax you have a pathway to deferring the payment by not realising the capital gain. They support the ‘less is more’ approach to investing.
We would expect that higher turnover portfolios in a taxable environment would be even further behind in terms of portfolio returns once tax costs are taken into account.
Market impact costs
For investors who use managed investments, for example managed funds, there may be a more insidious cost that comes with trading. Institutional investors like managed funds can be selling or buying in such large quantities that they create a ‘market impact cost’. For example, if they are selling a portion of a holding it might be such a large amount that the selling pushes the price of the shares down, reducing the sales price. This is sometimes seen in institutional sales where blocks of shares might be sold through brokers at a discount to the market price.
It is a shame that managed investments are not required to report their portfolio turnover in Australia, as this would be useful and easily calculated information for investors and potential investors to have to consider their managed investment.
Amongst the many portfolio metrics, we have at our fingertips, perhaps we don’t think about ‘portfolio turnover’ enough. Part of this might be a lack of access to this information, particularly if managed investments are used. Given the relationship between higher portfolio turnover and poorer after-cost returns, and the capital gains tax regime in Australia that only taxes gains once assets are sold, there are two key reasons to keep an eye on your overall portfolio, and thinking carefully about the benefits of lower turnover, whether through being patient with direct holdings, or favouring low turnover ETFs and index style funds.