Quote:
Originally Posted by scottmci
Academic research on day trading.
Mathematical - how inefficient markets must be to overcome transactions costs.
Psychology - how humans react to randomness, lie to themselves, other biases.
General understanding of markets and exploiting them.
Anecdotal evidence.
Given this I believe that it is extremely hard to almost impossible to make money (in a positive EV/utility sense) day trading as a discretionary trader. There is some evidence it is not as difficult to make money in poker, although many of the above points apply there as well.
Anecdotal evidence is going to depend on who you hang out with. Of course most traders fail, we accept this, and strive to be among the very small group of those that do not.
Most of the academic studies of traders simply aggregate the results of all traders. Of course this will produce a poor result, just as taking the results of all poker players lumped together will show an aggregate loss. This does not mean poker (or day trading) is necessarily unprofitable for those who put in a large amount of time and effort developing their skills.
But I found one study that looked more closely at individual results:
http://faculty.haas.berkeley.edu/ode...e%20040330.pdf
The conclusion:
"Our analysis of performance indicates day trading is treacherous, but not entirely a fool’s game. Heavy day traders, as a group, earn gross profits (before transaction costs). Thus, heavy day traders do appear to have a trading advantage over other investors. The stocks bought by the most active day traders outperform those sold by 31 basis points
per day. Unfortunately, the gross profits of heavy day traders are not sufficiently large to cover reasonable estimates of transaction costs. Thus, as a group, they lose money. In contrast, occasional day traders experience both gross and net losses. The stocks bought by occasional day traders actually underperform those sold, even before considering transaction costs.
There is considerable cross-sectional variation in the performance of day traders. Over the typical six month horizon, using lower range assumptions regarding transaction costs, less than 20 percent of day traders earn profits net of transaction costs.
These results paint a rather dim portrait of day traders. However, we do document a select few are able to consistently earn profits sufficient to cover transaction costs. We identify day traders who earn substantial profits over a six-month period and analyze the performance of their
subsequent trades. These profitable day traders continue to earn stellar returns. The average day trader in this group earns a semi-annual income of over $NT 1 million from his day trading activity, though the group’s median income is a more modest $NT 126,000. The stocks they buy outperform those that they sell by 62 basis points
per day. These profits survive transaction costs. In other words, there is strong evidence of persistence in the ability of day traders.
Our analysis makes clear the need for comprehensive risk disclosure. Prospective day traders should be apprised of their likelihood of success: only two out of ten make money; fewer do so consistently."
The good news is transaction costs have been falling, and if you can profit before transaction costs, there are ways to reduce them further.
Academic studies of price behavior also suffer from this aggregation problem. They usually look at
all price changes over a long period of time and conclude there is not enough serial auto-correlation of prices to overcome transaction costs. That is, past prices do not correlate to future prices enough to profit after costs. But the successful trader doesn't care about the aggregate behavior of prices, he cares about finding the temporary anomalies that do allow profits after costs. The difficulty of being selective with one's trades is similar to the difficulty of "playing tight" in poker. Not everyone can maintain that level of self-control.