Good morning traders
On the desk today
š The study. 66,465 real brokerage accounts, six years, and a seven point gap between the busiest and the quietest.
š The bit everyone gets wrong. Before costs, both groups performed about the same.
š§¾ The Tape. Where the week actually finished, which isn't where Friday finished.
They weren't worse at picking. They were worse at leaving it alone.
In 2000, two finance professors, Brad Barber and Terrance Odean, published a paper in the Journal of Finance built on something nobody had been able to see properly before. A large discount broker gave them the real account records of 66,465 households, covering 1991 to 1996. Real trades, real outcomes, not a survey.
They sorted those households by how often each one traded.
The most active fifth earned 11.4 percent a year, net. The least active earned 18.5 percent. The market over the same stretch did 17.9.
Seven points a year, between people using the same broker in the same market at the same time.
The part that gets left out
This almost never makes it into the retelling, and it's the only part that matters.
Before costs, the two groups performed about the same.
The busy traders weren't picking worse. Their gross returns were close to the index and close to each other. Barber and Odean put the gap down to the cost of trading and the frequency of trading, not the portfolio selections.
So the seven points didn't come from bad ideas. It came from acting on them too often.
The busiest households were turning over more than 250 percent of their portfolio a year. The average household was doing more than 75 percent, which is roughly the whole book once. The busy ones were doing it two and a half times, and paying a spread on every one of those turns.
You have probably done this, and it didn't feel like churn
You didn't trade 250 percent turnover last year. Almost nobody reading this did. The mechanism doesn't need that scale to work on you.
š The add. You sized the position correctly, then added because it was working.
š The round trip. You exited at a level you hadn't planned, then got back in forty minutes later at a worse price, because you realised you'd been right the first time.
š The tightened stop. You moved it up to lock something in, and it took you out flat on a move you had called correctly a week earlier.
None of those feel like overtrading. Every one of them is a turn, and every turn is paid for at the same window.
And the costs didn't go away, they moved
The obvious objection is that this was the 1990s and commissions are gone, so the paper has expired.
I don't buy it. What happened is that the toll stopped being a line on a statement and moved into the fill you get, the slippage on a fast move, and the tax on a short holding period. It stopped sending you an invoice.
Something else replaced it those households never had. Their friction was a phone call. Yours is a thumb. The mechanism the paper found gets stronger the easier trading becomes, not weaker.
š§¾ The Tape
Where the week actually finished. Friday was green and the week was not.
Friday 11 Sep | On the week | |
|---|---|---|
S&P 500 | 7,656.97, +0.86% | down ~0.7% |
Dow Jones | 52,573.29, +0.98% | down ~1.5% |
Russell 2000 | 2,902.4, +0.45% | down 2.2% |
VIX | 15.85, ā9.58% | |
WTI crude | 99.98, high 104.46 | up |
Four straight down sessions Monday to Thursday, the longest S&P losing streak since March, then one relief day after the inflation print. One green day doesn't undo four red ones.
šÆ The Desk Take
Most people respond to a flat month by trying to get better at analysis. The study says that's the wrong end of the problem. Those households weren't bad at choosing. They were bad at leaving things alone, and the market charged them seven points a year for it whether they were right or not.
If you had a flat month recently, count your turns before you blame the market. Not your trades, your turns. A scale-in and a scale-out are two turns on one idea, and that's what your account experienced.
Most people have never counted. It's usually not the number they'd have guessed.
Counting turns by hand is miserable, which is the point
Working out your own turn count off a broker statement takes an evening and most people never do it twice. That's what the journal is for. Every logged position carries its R and its percentage, so you can sort by month and see the thing the study is actually about in about a minute.
It's 20% off with RBT20 at rbtrading.site.
What the paid edition does that this one doesn't
This free list is roughly weekly and it teaches the mechanics, like today.
The paid edition of Inside the Trade goes out three times a week and does a different job: the individual names and the exact levels behind them, written down before the open, plus the live board that updates as things trigger. Every trade that closes goes on a public record either way, win or lose, and that record is open to everyone:
If you want the levels as well as the method, that's the door: rbtrading.substack.com
Next issue: the second half of this one. Three professors, one auction, and why the second-highest bidder always pays the most.
Educational analysis, not financial advice. Study figures are quoted from Barber and Odean, "Trading Is Hazardous to Your Wealth", The Journal of Finance, April 2000, and describe US discount brokerage accounts between 1991 and 1996. Market levels are Friday 11 September 2026 closes from my own charts. Past performance is not indicative of future results.