Which thing actually repeats
A saved layout is a tool, and a tool is worth keeping when the job comes back. So the only question is which part of your reading comes back: the companies, or the things you ask them.
The companies change constantly. Some you look at once and never again. Some you own for a decade and rarely open. The questions do not change at all. Am I still comfortable owning this? Where has this price reacted before? What actually happened in today's session? You will ask those of a hundred different companies.
So the layout belongs to the question. Each question needs its own timeframe, its own chart type and its own scale — for reasons chapters 03.01 and 03.02 already settled — and those settings are the same whichever company you point them at.
Three questions, three layouts, and one mistake
The settings under each question follow from rules earlier in this Part, not from anybody's strategy. The fourth panel is what happens when you name a layout after a company instead.
These are illustrations of how a question fixes settings. They are not recommended layouts, and Granthara recommends no strategy.
Why one per company goes wrong
The failure is quiet, and it is not laziness. You open a company, the chart does not show what you were hoping to see, and you adjust — a longer lookback, a different timeframe, the axis dragged a little. Then you save it. You have not cheated; you have tuned the instrument until it produced the answer, and then preserved the instrument.
Do that across two hundred companies and you own two hundred different instruments. Nothing measured by one is comparable with anything measured by another, which is the same objection 03.01 raised against two raw rupee lines: the comparison the screen invites is not one the data supports.
A layout per question inverts it. The instrument is fixed before the company arrives, so the company has to meet the chart. When it looks different from the last one, that difference is information rather than a setting.
- Layouts, if saved per question
- three or four
- Layouts, if saved per company
- one for every company you have opened
- Two companies compared through a question layout
- same instrument, so the difference is real
- Two companies compared through their own layouts
- two instruments, so the difference means nothing
- What either approach tells you about the next price
- nothing
You open two companies through the same saved layout and one looks far more volatile than the other. What can you conclude?
What travels with a layout, and what must not
Travels: the six settings from 03.04.01 — data and symbol source, timeframe, chart type, scale and axis range, panels, colour rule. Those are statements about how you look, and they apply to any company.
Does not travel: anything drawn. A trendline belongs to one series on one scale — 03.02.02 showed it becomes a different line the moment the scale changes, and it means nothing at all on a different company. The same goes for a level, a note or a marked date. Those are facts about one stock and belong with that stock, not in the instrument.
Does not travel: anything tuned. A lookback stretched until one company's history looked right is no longer a decision about how you read charts. It is a decision about that chart, made while you were looking at it.
Name each layout after its question — long-horizon review, level check, session read. If you cannot name a layout after a question you repeatedly ask, you have not saved a method. You have saved a mood.
Question or company — which of these should a saved layout be named after?
Fix the instrument before you meet the evidence. Chapter 03.01 chose what a chart keeps, 03.02 chose how it is drawn, 03.03 asked whether the numbers are the right numbers, and this chapter saved all of it so the decisions are made when you are calm rather than when you already have an opinion. None of it tells you what a price will do. All of it tells you whether what you are looking at is worth being surprised by.
The workspace is settled, and every setting in it was chosen by a question. What remains is the thing the workspace was built to show: a single mark on a chart, four numbers wide, and everything people claim to read in it.