How to Day Trade a Stock
To day trade a stock, choose one liquid instrument, mark your levels before the open, and write entry, stop and size before any position exists. Trade only the busiest hours, be flat by the close, and record every trade including the ones you skipped.
Day trading is a cost problem before it is a chart problem. Every decision below exists to keep the number of attempts low enough that the cost base does not consume whatever edge is present.
Before you start
Your exact round-trip cost in the units you trade. Commission both ways plus the spread you actually cross, taken from a filled order rather than a fee schedule.
One instrument chosen for participation rather than for a story. Heavily traded, tight spread, and the same one every day until you know how it normally moves.
Levels marked before the session opens. Prior day high and low, the overnight range, the session open. Marked while flat, because a chart is very persuasive once you have a position.
The steps
1. Work out what a round trip costs you
On the shared price series this is 2% of a median bar’s range. Whatever your figure is, it is paid on every attempt, winners and losers alike.
2. Pick one instrument and stay with it
Choose it on volume and spread, not on news. Watching one thing for months is how you learn what an ordinary move looks like on it.
3. Mark your levels before the open
Prior day high and low, overnight range, the busiest price. Six or fewer. Drawn before the session so they cannot be adjusted to fit what is happening.
4. Write entry, stop and size before any position exists
Three prices and one quantity. The size comes from dividing what you will risk by the stop distance, never from what feels affordable.
5. Trade only the hours with people in them
The open and, where relevant, the session overlap. Outside those the spread widens while the range shrinks, which is the worst combination available.
6. Wait, and expect to wait
Price has to reach a level you marked. Most sessions it will do so once or twice, and many sessions it will not do so at all.
7. Be flat by the close
The entire method assumes no overnight exposure. A position carried past the close is a different trade with a different risk, taken by accident.
8. Record every trade and every skip
Level, reason, entry, stop, size, exit, cost, result. Skips get a row with the reason, because they are the evidence the plan governed the session.
How to tell it worked
Review the last 20 days, not the last 20 trades.
Count the days you took zero trades. If that number is 0 out of 20, the levels are too loose or the waiting is not happening. A method that produces a trade every single session is not selective.
Multiply your round trip by the number of trades in those 20 days. Put the result beside the gross figure. If costs are a large share of the gross, the frequency is the problem, not the entries.
Then count the positions still open at the close. The target is 0. Every one of those was an unplanned change of strategy, whatever the outcome.
Why the frequency decides most of this
The cost arithmetic is unforgiving at high frequency. A round trip on the shared series is 2% of a median bar’s range and exceeds 10% on 15 of 576 bars — so on the quietest bars, participating consumes a tenth of everything available.
And the moves are short. Direction runs average 2.01 bars with the longest at 11, so the thing being captured is usually two bars long while the toll is fixed and paid up front.
Which means the largest single lever is trading less. Improving entries is slow and uncertain; halving the number of attempts halves the cost base immediately and is entirely within your control.
The original data
Of the 24,971 unique videos in research/search-study-corpus.jsonl, 316 have an instruction-shaped
title mentioning day trading, at a median of 36,138 views across 185 channels, with a maximum of
6,602,501. That is the largest instructional subject in the corpus by video count. The counts come
from site/rank_howto.py.
316 videos across 185 channels, and a maximum of 6.6 million views. This is the most contested subject on the site, which changes what a page here can usefully do: competing on enthusiasm is hopeless, and the only available ground is the arithmetic that most of those 316 leave out.
The answer to the question on that chart is no, and the reason is that the session is a fixed budget. A trade taken to recover the day is sized and timed by the deficit rather than by a level. The cost of that trade is certain and its edge is not — and it is the trade that turns an ordinary losing day into an unusual one.
When it fails
A quiet session is where the damage happens, and it happens invisibly. Nothing reaches a level, the spread is wide relative to the range, and the pressure to justify the day produces a series of small marginal trades. Each looks defensible on its own; together they are a fee schedule. Nothing dramatic occurs, which is exactly why the pattern survives for months before the record shows it.
The second failure is watching too many instruments. You arrive late to each and learn none.
A third is marking levels during the session. They then describe what already happened.
A fourth is holding past the close. The method’s premise was intraday, and the risk changes entirely.
A fifth is sizing from the account rather than the stop. It produces positions the plan cannot absorb.
And a sixth is judging a week. At this frequency a week is noise, and 20 days is the minimum honest sample.
Related
Day trading covers the style, its demands, and the cost structure underneath it. Trading plan is where the three prices from step four belong. And bid-ask spread is the cost that decides whether any of this can work.
The change that mattered was cutting to one instrument. I had been watching five, which meant I was always late to whichever one moved and I never learned how any of them normally behaved. One instrument for a few months taught me more about what a typical bar looked like than a year of scanning had, and ’typical’ is the thing every other decision is measured against.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.