In light of a couple of threads along this line, I've written a brief article looking at some of the studies along these lines.
One of the most frequent unsubstantiated statements of fact to be found in trading forums, books and courses is that 90% (or some similar number) of traders are unprofitable. For such an important statistic it is extremely rare to find any serious reference to the source of this information, and indeed most of these statements of fact are based on ‘common knowledge’ rather than serious study. Trading is plagued with statements of common knowledge which can be a minefield for new traders. Often they are based on opinion or are regurgitated by writers and educators who have little or no record of success in the markets. I will attempt to apply some science to this important fact relating to trading success – after all it is vitally important for traders to understand the odds of success.
It is likely of course that greater knowledge of the likelihood of success will not deter traders. Many people enter the markets in full acceptance of the risks that they are aware of. The reality of losing however often leads to the abandonment of trading except in cases where individuals have a pathological need for the excitement and thrill of trading, despite the financial consequences. A handful are able to master their emotions and apply the work and study required for lasting success. Traders hearing these often quoted statistics of 90% failure logically must conclude (unless they are completely fatalistic) that they will be in the top 10%. This overconfidence is a common facet of humans, especially males. A majority of people believe they are above average drivers for instance, a statistical impossibility. One benefit of a more scrupulous view of the success rates of traders might be that studies into the success of traders often explore the reasons behind their failure. This knowledge if correctly applied could be of practical benefit to speculators.
One of the most recent studies of the success of traders was by Brad Barber of the University of California, Davis. In his paper Do Individual Day Traders Make Money? (2004) he obtained data on all trading transaction data on the Taiwan Stock Exchange between 1995 and 1999. Day traders make around 20% of total volume and Barber focussed on these participants. He found that “heavy day traders appear to trade at favourable prices, but only a select few are sufficiently savvy to consistently earn profits net of their trading costs. More than eight out of ten day traders lose money in a typical semiannual period”. Interestingly the successful traders (less than 20% of day traders) who were profitable in each six month period assessed continued to earn stellar returns. The average annual income was $NT 1 million. This makes clear that the minority of successful traders were not just lucky.
Brad Barber (with Terrance Odean) also conducted another, US based study in 1998. This study (The Courage of Misguided Convictions) used data from 10,000 randomly successful accounts from a US national discount brokerage that were active in 1987. This study was specifically interested in whether traders sold gains more readily than realising their losses. They found that a stock that was up was about 50% more likely to be sold than one that was down.
One study which did find evidence that day traders are profitable was Harris & Schultz’s The Trading Profits of SOES Bandits. This did find evidence that traders were able to beat market makers. It must be said however that these were specific traders using the Nasdaq SOES platform who were able to trade within the Bid/Ask spread using Instinet or SelectNet. I would argue that these traders are not representative of the typical retail trader for this reason.
More commonly cited is Ronald Johnson’s study Day Trading: An Analysis of Public Day Trading at a Retail Day Trading Firm. This randomly chose 30 accounts from a day trading firm. In this case 70 % of the accounts list money. More importantly all of the losing accounts were traded in a manner that had a risk of ruin of 100%. Of the eight successful accounts three were traded with a high risk of ruin and low average trade size. The performance of these accounts was highly dependent on a single trade. Only three accounts showed profitability, sound reward/risk ratios and low risk of ruin. The best trader wasn’t actually a day trader, holding positions for an average of 47 days with no day trades.
One of the best known studies is that of Blair Stewart (An Analysis of Speculative Trading in Grain Futures, 1949) who obtained the complete trading records of 8922 customers of a trading firm which went bankrupt in the 1930s. The study found that 75% of the speculators lost money. Some of the points noted were:
- a clear tendancy to cut profits short and let losses run
- more likely to be long than short
- long speculators tended to buy on days of price declines, and shorts sold on price rises. Most traders thus based decisions on price levels rather than price movement.
Another study was conducted by Thomas Hieronymus (Economics of Futures Trading, 1977). He obtained summary records from a firm for the year 1969. In this case 65% of traders lost money. Regular traders did better, and their nett profits were nearly enough to overcome their losses.
The authors of the book The Futures Game (Richard Teweles and Frank Jones) studied the records of a ‘respected’ brokerage firm over 10 years beginning in 1962. The percentage of traders with profits in any year ranged from a low of 14% to a high of 42%, with an average of 26%.
This certainly isn’t a completely comprehensive look at every single study ever conducted, however I believe it encompasses most of the well known and verified ones. Overall, the commonly held theory that 90% of traders lose money seems high, with around 75% a more accurate number. It should be noted however that many of these studies were carried out before the advent of deep discount brokers with significantly smaller commissions. It would be interesting to obtain more recent data to see if the reduction in costs has had any impact on the profitability of traders. The vastly increased participation and ease of access might have some effect as well, with more ‘casual’ participants (rather than a more select sector of the population with higher net worth as in the past) perhaps being more or less successful than traders in the past.
One of the most frequent unsubstantiated statements of fact to be found in trading forums, books and courses is that 90% (or some similar number) of traders are unprofitable. For such an important statistic it is extremely rare to find any serious reference to the source of this information, and indeed most of these statements of fact are based on ‘common knowledge’ rather than serious study. Trading is plagued with statements of common knowledge which can be a minefield for new traders. Often they are based on opinion or are regurgitated by writers and educators who have little or no record of success in the markets. I will attempt to apply some science to this important fact relating to trading success – after all it is vitally important for traders to understand the odds of success.
It is likely of course that greater knowledge of the likelihood of success will not deter traders. Many people enter the markets in full acceptance of the risks that they are aware of. The reality of losing however often leads to the abandonment of trading except in cases where individuals have a pathological need for the excitement and thrill of trading, despite the financial consequences. A handful are able to master their emotions and apply the work and study required for lasting success. Traders hearing these often quoted statistics of 90% failure logically must conclude (unless they are completely fatalistic) that they will be in the top 10%. This overconfidence is a common facet of humans, especially males. A majority of people believe they are above average drivers for instance, a statistical impossibility. One benefit of a more scrupulous view of the success rates of traders might be that studies into the success of traders often explore the reasons behind their failure. This knowledge if correctly applied could be of practical benefit to speculators.
One of the most recent studies of the success of traders was by Brad Barber of the University of California, Davis. In his paper Do Individual Day Traders Make Money? (2004) he obtained data on all trading transaction data on the Taiwan Stock Exchange between 1995 and 1999. Day traders make around 20% of total volume and Barber focussed on these participants. He found that “heavy day traders appear to trade at favourable prices, but only a select few are sufficiently savvy to consistently earn profits net of their trading costs. More than eight out of ten day traders lose money in a typical semiannual period”. Interestingly the successful traders (less than 20% of day traders) who were profitable in each six month period assessed continued to earn stellar returns. The average annual income was $NT 1 million. This makes clear that the minority of successful traders were not just lucky.
Brad Barber (with Terrance Odean) also conducted another, US based study in 1998. This study (The Courage of Misguided Convictions) used data from 10,000 randomly successful accounts from a US national discount brokerage that were active in 1987. This study was specifically interested in whether traders sold gains more readily than realising their losses. They found that a stock that was up was about 50% more likely to be sold than one that was down.
One study which did find evidence that day traders are profitable was Harris & Schultz’s The Trading Profits of SOES Bandits. This did find evidence that traders were able to beat market makers. It must be said however that these were specific traders using the Nasdaq SOES platform who were able to trade within the Bid/Ask spread using Instinet or SelectNet. I would argue that these traders are not representative of the typical retail trader for this reason.
More commonly cited is Ronald Johnson’s study Day Trading: An Analysis of Public Day Trading at a Retail Day Trading Firm. This randomly chose 30 accounts from a day trading firm. In this case 70 % of the accounts list money. More importantly all of the losing accounts were traded in a manner that had a risk of ruin of 100%. Of the eight successful accounts three were traded with a high risk of ruin and low average trade size. The performance of these accounts was highly dependent on a single trade. Only three accounts showed profitability, sound reward/risk ratios and low risk of ruin. The best trader wasn’t actually a day trader, holding positions for an average of 47 days with no day trades.
One of the best known studies is that of Blair Stewart (An Analysis of Speculative Trading in Grain Futures, 1949) who obtained the complete trading records of 8922 customers of a trading firm which went bankrupt in the 1930s. The study found that 75% of the speculators lost money. Some of the points noted were:
- a clear tendancy to cut profits short and let losses run
- more likely to be long than short
- long speculators tended to buy on days of price declines, and shorts sold on price rises. Most traders thus based decisions on price levels rather than price movement.
Another study was conducted by Thomas Hieronymus (Economics of Futures Trading, 1977). He obtained summary records from a firm for the year 1969. In this case 65% of traders lost money. Regular traders did better, and their nett profits were nearly enough to overcome their losses.
The authors of the book The Futures Game (Richard Teweles and Frank Jones) studied the records of a ‘respected’ brokerage firm over 10 years beginning in 1962. The percentage of traders with profits in any year ranged from a low of 14% to a high of 42%, with an average of 26%.
This certainly isn’t a completely comprehensive look at every single study ever conducted, however I believe it encompasses most of the well known and verified ones. Overall, the commonly held theory that 90% of traders lose money seems high, with around 75% a more accurate number. It should be noted however that many of these studies were carried out before the advent of deep discount brokers with significantly smaller commissions. It would be interesting to obtain more recent data to see if the reduction in costs has had any impact on the profitability of traders. The vastly increased participation and ease of access might have some effect as well, with more ‘casual’ participants (rather than a more select sector of the population with higher net worth as in the past) perhaps being more or less successful than traders in the past.
