The Cost That Does Not Scale

What a fixed cost is, in R

Every cost this course has priced so far arrives with a trade. Lesson 12’s spread is paid on the way in and again on the way out. Lesson 59’s impact is paid in proportion to what you send. Lesson 78’s rate is charged on a gain that has to happen first. A fixed cost is the other kind: the data feed, the second connection lesson 77 priced, the accountant, the entity filing. It arrives in a year you traded four hundred times and it arrives in a year you traded none.

Three divisions turn it into a number that belongs beside your edge, and almost nobody does the third. Divide the annual cost by the capital and you have the return the account owes before it has earned anything for you. Divide that by the round trips a year and you have the cost of one round trip as a fraction of the account. Divide that by the risk per trade and you have it in R, which is the unit the edge is already in, and now the two can be subtracted.

Do the substitution and something exact falls out. Lesson 79 sized the capital so that the expected annual gain equals the wage, which means the wage is the gross edge itself, expressed in money. So a cost base measured as a fraction of that wage is the same fraction of the edge, at any pace, any account size and any currency. A quarter of the wage is a quarter of the edge. There is no approximation in that sentence.

The second column is the part people plan for and the last column is the part they do not. A cost base of a quarter adds just under three years of wage to the capital, which sounds like a purchase you could make. It also adds eight years to the wait, and those eight years are years in which the wage and the cost base are both being drawn. The reason the last column moves so much faster than the second is that the fraction enters through a square: lesson 67’s test length goes as one over the edge squared, so halving the edge does not double the trades, it quadruples them, and leaving a quarter of it multiplies them by sixteen. This is the same shape lesson 78 found in the relief fraction, arriving through a different door.

Almost nobody has divided their own annual fixed costs by the wage they intend to draw. It is one division on a number you already pay, and the answer is the share of your edge that the structure is taking before a single trade is placed.

Now hold the money still and move the pace, because the first table did the opposite. Take lesson 79’s capital of 11.80 years of wage and a cost base of a quarter of a wage, which is 2.12 per cent of that capital a year, and run it at each of lesson 76’s four paces.

Read the last two columns against each other, because the gap between them is the whole of what this page has to say. The same bill, on the same account, costs 8.11 years at the top of the table and 0.08 of a year at the bottom, which is a month. A fixed cost is divided among the trades it is spread over, so the fewer trades there are the larger each one’s share, and lesson 76 counted how many there actually are.

That is the reversal against the page before it. Lesson 79’s finding was that the pace cancels and the risk per trade is the only lever. It cancels because both terms in that ratio were proportional to the pace. A fixed cost is not, by definition, so the moment one exists the pace stops cancelling and becomes the term that matters most on the page.

Module 10’s list of what the day contains that is not the decision gains its fifth item: the overhead, and the number of trades it has to be spread across.

A quarter of a wage

Take the book module 9 licensed and lesson 79 priced: two rules at one and a half per cent a trade, 56.5 round trips a year, a hypothetical tenth of an R. Now give it the smallest cost base that looks like a business rather than a hobby — a data subscription, the second connection lesson 77 priced, an accountant once a year — and suppose it comes to a quarter of the wage you intend to draw.

Capital, if you fund only the wage: 11.80 years of it, unchanged from lesson 79.

Capital, if you fund the wage and the cost base: 14.75 years, an extra 2.95.

Edge after the overhead: 0.075 R rather than 0.10.

Verdict: 1,047 trades at 56.5 a year, which is 18.53 years rather than 10.42.

Now put the last two lines together, which is where the arithmetic stops behaving. Over 18.53 years you draw the wage and pay the cost base, so 23.16 years of wage leave the account. Against the 14.75 that funded it, that is 157.0 per cent, and lesson 67’s median worst drawdown of 8.82R adds 13.23 per cent of the account on top. Lesson 79 reached 101.6 per cent and called it more than the account holds. Adding a cost base of one quarter takes it to 170.3.

A fixed cost is not a bill the account settles out of its profits. It is a piece of the edge, and the pace decides how big a piece.

Which is why the order this subject is usually taught in is backwards. It says form the entity, open the accounts, engage the professionals, and then trade. On these numbers the structure is the last purchase rather than the first, and the thing that decides when you can afford it is not the size of the account but the number of trades a year you have to divide the bill among. Two accounts of identical size, one taking 56.5 round trips a year and one taking 480, are not the same business, and the second can carry a cost base the first cannot.

So add up what you pay in a year whether you trade or not, divide it by the wage you mean to draw, and subtract that fraction of your edge before you run any test on it.

What this does not settle

That a quarter of a wage is the right cost base. It is a parameter, every row of the first table is printed, and the number is yours to supply. This page names no price for anything for the same reason lesson 78 named no jurisdiction: a data feed costs what it costs where you are, in a currency this page does not know, and the map from the fraction to the years is the part that travels.

That lesson 78’s deduction rescues this. It helps, and the help is smaller than the deduction looks. At a rate of thirty per cent with the costs fully relieved, a cost base of a quarter of a wage costs 17.5 per cent of it after relief, the edge is 0.0825, and the verdict arrives in 15.31 years rather than 18.53. That is three of the eight years back, on the most favourable relief assumption lesson 78 allows, and it is a reduction in a cost rather than a removal of one.

That the fixed costs are waste. Lesson 77 priced exactly what some of them buy: a second connection takes the hours a year in which you hold an unprotected position from 96.3 to 27.0, and that reduction is real and is nowhere in this arithmetic. This page prices what the base costs and not what it prevents, which makes it half of an argument, and the missing half is a risk that a return series cannot see by construction.

That the pace is a lever you may pull. The second table reads like an instruction to trade more often, and lesson 76 is the reason it is not one. The top row is what the course’s own grid actually produced on sixty closes; the bottom row is what lesson 65 stated and nothing ever measured. Getting from one to the other means running more rules or faster ones, and every rule has to clear lesson 64’s bar, lesson 67’s test, lesson 68’s capacity and lesson 70’s filter condition first. You do not reach the bottom row by deciding to.

And the concession that costs most: this is the fifth subtraction in a row. Module 10 has now priced the day, the machine, the rate, the wage and the overhead, every one of them comes out of the same tenth of an R, and not one page in the module has priced what any of it buys. That is not an oversight and it is not defensible either. It is what happens when the only evidence a course will accept is a return series, because a return series contains every cost and none of the reasons the arrangement exists. Lesson 81 is the capstone and it has no more room to defer: it puts all eleven thresholds this course has produced against the one book module 9 licensed, at once, and reports how many of them that book actually clears.