Before you begin
- Before this: Chapter 5

Lesson 01 · Section 1 of 10 · Reading
Chapter 5 showed you a business working. This chapter is about the fact that it might not.
Chapter 5 showed you a business working. This chapter is about the fact that it might not.
Every one of the forty modules assumes you can tell a bad outcome you absorb from a bad outcome that ends you, and that you know when to use an average and when averages are the wrong tool entirely. Neither is taught anywhere in them.
risk appears 851 times across the modules. It is the single most-used word in the whole corpus and it is used in at least four different senses, none of them marked. 6.1 sorts them out.
| Subchapters | ||
|---|---|---|
| 1 | What we are even talking about — risk, uncertainty, chances, expected value | 6.1 – 6.4 |
| 2 | Shapes — distributions, tails, and the sort that decides everything | 6.5 – 6.7 |
| 3 | Where numbers mislead — spread, averages, scenarios, sensitivity, stress | 6.8 – 6.12 |
| 4 | What to do about it — the four responses, and paying to be less exposed | 6.13 – 6.14 |
| What it does | ||
|---|---|---|
| 6.1 | Risk is not uncertainty | You know the shape, or you don't. Different responses |
| 6.2 | Probability without maths | Frequencies, not notation. And the one minus none move |
| 6.3 | Expected value | The average of what could happen. Computed three times |
| 6.4 | When expected value is the wrong answer | A +$165,000 bet that should be refused |
| 6.5 | Distributions, drawn not calculated | The whole picture, not its balance point |
| 6.6 | Tails | Rare and large beats common and small |
| 6.7 | Ruinous or survivable? | ★★ The spine of the entire frh course |
| 6.8 | Variance | Why two things with the same average differ |
| 6.9 | The flaw of averages | Plan for the average and you plan for a day that never happens |
| 6.10 | Scenarios | A few whole futures, not one number ± a bit |
| 6.11 | Sensitivity | Which assumption actually matters |
| 6.12 | Stress tests | Not what's likely but what would break us |
| 6.13 | The four things you can do | Ignore · reduce · insure · hedge |
| 6.14 | Insurance and hedging | Negative expected value on purpose, and why that's rational |
6.4 — when expected value is wrong. A coin flip worth +$165,000 that Sam should refuse, because the bad side is $70,000 and Sam can absorb $25,000. Expected value assumes you get to play again. The three conditions under which it is the right tool, and the one under which it is not.
6.7 — ruinous or survivable. The sorting question that ignores probability entirely. Sam's three ruinous outcomes are individually 5%, 10% and 2% — dismissible one at a time, and together a 16.2% chance in any year and 58.7% within five. More likely than not that the shop ends, and nobody had multiplied it out.
6.9 — the flaw of averages. Sam plans on average demand and the plan overstates profit by $17,850 a year — a third of operating profit — with no arithmetic error anywhere in it. The error is having averaged before the arithmetic instead of after.
6.14 — paying to be less exposed. Every honest insurance policy has negative expected value, by construction. Sam pays $4,200 to remove $1,200 of expected loss and is right to. Understanding why is what makes hedging, cash buffers and turning down good projects stop looking like timidity.
Forty-four glossary terms, including six of the dangerous ten and three of the six two-sense words:
risk (851 uses) · hedge/hedging (735/257) · exposure (621) · premium (487) · forward (328) · option (267) · tail (178) · distress (151) · downside (133) · upside (376) · uncertainty (111) · insurance (111) · ruin (112) · base case (109) · loading (96) · worst case (85) · scenario (76) · variance (74) · natural hedge (69) · deductible (63) · probability (54) · frequency (48) · sensitivity (46) · payoff (46) · expected value (38) · volatility (37) · critical ratio (29) · stress test (23) · hedge ratio (23) · expected loss (24) · distribution (22) · self-insure (22) · retention (14) · swap (12) · stockout (10) · basis risk (8) · notional (8) · futures (8) · service level (6) · severity (4) · reorder (4) · bad debt (2) · collar (72)
premium, option and variance each carry two unrelated senses, flagged where they appear.
Sam's coffee shop again, now viewed through what could go wrong. Six risks carry the whole chapter.
| Chance | Loss | Expected loss | |
|---|---|---|---|
| Landlord does not renew | 10% | $95,000 | $9,500 |
| Supplier raises ingredient prices | 33% | $23,040 | $7,680 |
| A wet summer | 25% | $18,000 | $4,500 |
| Street closed for resurfacing | 5% | $37,600 | $1,880 |
| Espresso machine fails | 20% | $9,000 | $1,800 |
| Employment claim | 2% | $60,000 | $1,200 |
| $26,560 |
$26,560 is 49% of Sam's operating profit — the expected annual cost of things going wrong, invisible on all three of Chapter 5's statements.
Sam's survival line is $25,000 — $12,600 of cash plus $12,400 of unused overdraft. Losses above it are ruinous unless they arrive slowly enough to absorb from trading (6.7, 6.14); losses below it are not. And note the ranking: the employment claim is last by expected loss and fatal.
Every figure is checkable on paper.
About two hours for all fourteen, plus three minutes for Gate E.
If you are short of time, the spine is 6.4 → 6.7 → 6.14 — about 30 minutes, and those three carry the entire frh course.
Gate E — The Bad Outcome Board. Six cards, two zones, and a survival line you can drag. Moving it re-sorts everything on the board, which is the point: an outcome is not ruinous because it is large or likely. It is ruinous relative to a buffer, and the buffer is a choice.