What Survived

This is the last video of the course. Almost everything in it was a claim traders repeat, counted on real prices.

Many of those claims fell apart. This video is about the ones that did not: ten things this course can say with numbers behind them.

Every number here is collected from the video it came from, by a script that stops if any of them has changed. So the finale cannot disagree with the lessons.

One. The size of a day carries over to the next. The shape of a day does not.

After a large day, more than one and a half times the usual range, the next day moved zero point six nine of a range. After a quiet day, zero point one nine.

More than three times as much. Meanwhile, wherever a bar closed, top, middle or bottom, the next day was up fifty two to fifty three per cent of the time, the same as any day.

So read a bar for its size against the instrument's usual range, not for its shape.

Two. Almost everything the market moves, it moves back.

Over a year, ninety three to ninety nine point six per cent of the daily moves on the stocks cancelled out. Ninety six to ninety seven on crypto.

The net move is a thin remainder of a lot of noise, which is why most single days tell you nothing.

Three. A third of a stock's movement happens while you cannot trade.

Thirty three to forty five per cent of the stocks' movement came between the close and the next open. The worst overnight gap was eight point eight per cent of price.

Crypto trades around the clock and has no such gap. So rules travel between the two markets. Position size does not.

Four. When price closes beyond a level, it reaches the next one about two times in three.

On one thousand nine hundred and eighty seven crossings, sixty nine per cent went on to the next level. The claim said eighty.

Not eighty, but not made up either. The rule is real; the number that usually comes with it is not.

Five. What a trade costs is arithmetic, and on crypto it decides the result.

A round trip took two to five per cent of the stop on the stocks. On crypto at an ordinary fee, thirty five to fifty nine per cent.

On the lowest-volume fee tier, one hundred and thirty six to two hundred and twenty seven per cent: the trade has lost before price moves.

Six. Leverage does not change what a trade makes per unit of risk. It adds a second stop that you did not place.

At ten times, the exchange closes you out after a nine and a half per cent move against you. At fifty times, after one and a half per cent.

On Tesla, the move from the open reached that fifty-times line on forty six per cent of days.

Seven. A tighter stop makes every trade more expensive in R, by exactly as much as it is tighter.

On the index fund, a round trip cost zero point one two of an R at the daily stop. Four times tighter, zero point four nine. Eight times tighter, zero point nine eight.

That is why moving the same method to a faster chart is not free. The pattern may scale. The bill does not.

Eight. A line can be drawn almost anywhere.

A candidate level turned up forty eight to fifty six times in every hundred bars. About every other day.

So a line existing proves nothing. What matters is the rule that picked it, written down before price got there.

Nine. The best result found in a test does not survive new data.

Out of twenty one combinations of entry, stop and exit, the best one made plus zero point three eight R a trade on the first half of the data.

The same combination on the second half: minus zero point one. And seventeen of the twenty one did worse on the second half.

Whatever you find by searching, check it on data you did not search.

Ten. How you count decides the result.

The first count of the level setup said plus zero point nine two of an R a trade, on one thousand one hundred and thirty four trades.

Counted honestly, with the entry day handled the way four prices allow, it was eight hundred and eighteen trades at minus zero point one seven.

And it lost on twenty four of twenty four combinations of instrument and buffer.

Put together, that is what this course leaves you with. Not a setup that wins. A way to tell whether one does.

Size before shape. Costs before edge. The second half of the data before the first. And two error bars before any conclusion.

Two limits on this video. One: everything here was measured on daily bars of six instruments, over the years the course had. A different market needs its own count.

Two: surviving here means a claim was tested and held. It does not mean it holds forever. Run the count again when the data changes.

So, what to go and do. Take the ten, and for each one write down the one thing it changes in how you trade.

Then take the next claim you hear, and before you use it, ask the question this course asked every time: has anyone counted it, and how?

No trades yet? None of the ten needs one. They are about the market, the arithmetic and the counting, and they hold before your first trade.

Measure first. Thank you for watching the course. Educational content only. Nothing here is financial advice.