Brad Barber and Terrance Odean examined the accounts of 66,465 households at a large US discount broker over 1991 to 1996.
The average household turned over about 75 percent of its portfolio each year and earned 16.4 percent annually against a market return of 17.9 percent, underperforming by about 1.5 percentage points.
Households in the most active group earned 11.4 percent, underperforming by about 6.5 percentage points.
The authors traced the shortfall primarily to transaction costs and attributed the elevated trade counts to overconfidence.
Now the finding, and it is the most useful thing anybody can be told at the start.
The average household picked well enough to finish close to the market and gave the difference back through activity.
Nothing about their selection was the problem.
The largest measured cost in retail participation is not being wrong about what to hold. It is how much is done.
And per O100-04 the driver the authors identified is the thing you just measured about yourself.
Ten minutes, no exercise, and it is the last thing before your path begins. Per O100-01 through O100-04 you have four numbers about yourself. These are the four numbers about everybody else.
The four base rates
Most published findings in this area do not replicate. Per O100-01, 82.1 percent of 452 asset pricing anomalies failed a consistent replication at the multiple-testing hurdle, leaving about 81.
Most people who pursue trading as an income do not achieve one. Per O100-02, of 19,646 people followed, 1,551 persisted beyond 300 trading days and 97 percent of those lost money net of fees, with 1.1 percent clearing a minimum wage.
Most retail accounts in leveraged products lose money. Per O100-03, between 74 and 89 percent, with average losses per client from about EUR 1,600 to about EUR 29,000.
And most measured underperformance is activity rather than selection. Per the autopsy, 1.5 percentage points a year for the average household and 6.5 for the most active, from portfolios that were otherwise close to the market.
None of these is a prediction about you. Per S203-04 they are priors, and per O100-04 the evidence that would move them is a record you have not yet produced.
What this curriculum can and cannot change
Per P10, stated plainly because everything downstream depends on it.
It cannot change the base rates. Nothing in it makes a market easier or a distribution kinder.
It can change what you know, per O100-01, which is what most of the material is.
It can change what you can compute. A liquidation price, a position size, an expectancy, a cost stack, a coverage figure. Per the autopsy the terms that decided those households' outcomes were arithmetic.
It can change what you do not do, which per the autopsy is the largest single term available.
And it can change how much you commit, per O100-03, which is the only control that works regardless of whether anything else in it is right.
Part one: the four outputs.
A competency map across six domains, per O100-01, which produced your path.
A written intent, one sentence, per O100-02.
A risk ceiling with its three subtractions and the date, per O100-03.
And a calibration curve, per O100-04, which is re-measured at every gate.
Part two: the arithmetic to remember from this lesson.
Average household: 17.9 - 16.4 = 1.5 percentage points a year.
Most active group: 17.9 - 11.4 = 6.5 percentage points a year.
6.5 / 1.5 = about 4.3 times the shortfall, from activity rather than from selection.
Part three: the single sentence.
Doing less is the cheapest improvement available to almost everybody who arrives here, and per the autopsy it required no skill, no information and no timing from the households that would have benefited.
Part four: where to go next.
Per O100-01 your path is per domain. Most people start at Freshman, which is custody, threat models and what you actually own. Per O100-02 if your intent is served entirely by that, the honest advice is to stop there.
And per part three, stopping is a legitimate outcome of this curriculum rather than a failure to complete it.
Ten minutes of reading and one filing job.
Put the four outputs from part one in one place, dated. Per O100-03 the ceiling drifts if it is not recorded, and per O100-02 an intent recalled later has been shaped by what happened.
Then read the four base rates once more before your first real decision. Per S203-04 they are your priors until your own record says otherwise, and per O100-04 your record does not exist yet.
And treat the least glamorous finding as the most useful one. Per part two the most active households gave up about 4.3 times what the average ones did, and per part three the correction required nothing except doing less.
Per P10 nothing here promises anybody a return. Per part four it offers a competency map, four numbers about yourself, and a path that is allowed to be short.
If the base rates are this bad, why would anybody do this at all?
That is a reasonable question and the honest answer is that the base rates describe an average, not a prohibition.
What the numbers say. Per the four base rates, most participants lose, most published findings do not replicate, and activity is the largest measured cost. Anybody presenting this field without those numbers is presenting an advertisement.
What they do not say. Per O100-02 the same study found a small group who did clear a living, and per R408-06 a much larger study found persistent skill differences at the top of a distribution. The distribution has a right tail and it is not empty.
And most reasons to be here are not that tail. Per O100-02 the intents this curriculum serves directly include holding assets without losing them, understanding what you already own, not being defrauded again, and being able to read a claim. Per the four base rates none of those requires beating anybody.
Nor is the alternative neutral. Per O100-03 many people already hold exposure they did not choose deliberately, and per the four base rates the cost of understanding it is far smaller than the cost of not.
So per P6 the accurate framing: the base rates argue against one specific intent and support most of the others. Per part four the honest path for many people is short, and finishing it is the point rather than continuing past it.
Four base rates, given before anything else. Most published findings in this area do not replicate, at 82.1 percent of 452 anomalies failing at the multiple-testing hurdle. Most people who pursue trading as an income do not achieve one, at 97 percent of the persistent group losing money net of fees. Most retail accounts in leveraged products lose money, at 74 to 89 percent with average losses to about EUR 29,000. And most measured underperformance is activity rather than selection: 66,465 households picked well enough to finish 1.5 points behind the market on average, while the most active group finished 6.5 points behind, about 4.3 times the shortfall, from portfolios that were otherwise fine. This curriculum cannot change any of those. It can change what you know, what you can compute, what you do not do, and how much you commit. You leave orientation with a competency map, a written intent, a risk ceiling and a calibration curve, and with a path that is allowed to be short.