Reading the result as evidence, and presenting it
◈ 8 cardsConsistent with the EMH or not — never “proves” — with the limits volunteered before they are asked; and five minutes with one chart per claim, every member owning a number, ready for the three questions the instructor will ask.
Three outcomes, one grammar
The term is over and the sheet has a number. Whatever it is, the verdict uses the same grammar: the result is consistent or inconsistent with semi-strong efficiency. Never proves, never disproves — one term of twelve stocks cannot do either.
Outcome A — the chosen 12 beat the random 12 by 4 points. Lesson 14.1's simulation produced gaps of 4 to 9 points from identical distributions, so a 4-point win sits inside what luck does routinely. Consistent with efficiency — not evidence of skill. Check the Sharpe-style ratios: if the chosen basket also carried a higher σ or a higher β in a rising market, the gap shrinks further.
Outcome B — the random 12 beat the chosen 12. Twelve reasoned picks lost to twelve dice rolls. Consistent with efficiency again — and the more instructive case, because the group must resist the temptation to explain each loser after the fact. ("Cobalt Ridge would have risen if not for the strike" is exactly the public information the price already carried.)
Outcome C — both baskets trailed the index. Twelve stocks is an undiversified sample of a 200-plus-stock index (Lesson 13.3): more unsystematic risk, and in a given term that shows up as a miss in either direction. Consistent with efficiency, and a reminder of why the average active fund lags (Lesson 13.7).
What would be inconsistent? A gap far outside the simulated noise — say 25 points — that survives risk adjustment and cannot be traced to one stock's takeover bid. Even then the honest sentence is "inconsistent with semi-strong efficiency over this term", followed by the limits.
The limits, volunteered
Say these before the instructor does:
- Sample size — one term, twelve stocks per basket, twelve weekly observations. The biggest limit by far.
- Luck and noise — the simulation's 4-to-9-point gaps from pure chance.
- Survivorship — the pool was today's index constituents; the stocks that fell out of it are not in the draw.
- No costs — no commissions, no bid-ask spread, no taxes; real baskets earn less.
- Currency — the S&P 500 comparison depends on the CAD/USD move as much as on the index.
Five minutes
The presentation is 20 % of the course and the instructor grades it hard, asks questions, and lets peer evaluations move the mark. Structure it as one claim per slide, one chart per claim, and one member per number:
| Minute | Slide | Claim | Chart | Who |
|---|---|---|---|---|
| 0–1 | Hypothesis | H₀ in EMH terms, dated week 1 | — | A |
| 1–2 | Method | the draw sheet; dividends added back | draw sheet | B |
| 2–3 | The chart | four cumulative lines, in CAD | cumulatives | C |
| 3–4 | Risk-adjusted | Sharpe-style and β per basket | the table | D |
| 4–5 | Verdict | consistent / inconsistent, with limits | — | A |
Every member must be able to defend every number, because the questions go to whoever looks least ready. The three questions that always come:
- "Why these 12?" — for the chosen basket, the documented reasoning; for the random basket, "the sheet drew them", and here is the sheet.
- "Is that gap within noise?" — the simulation and the sample-size limit, in one sentence.
- "What would change your mind?" — a risk-adjusted gap well outside the noise, repeated across terms.
A group that answers the third question calmly has understood the course.
Cold
Consistent / inconsistent, never proves; five limits volunteered; five slides, one chart per claim, every member owning a number; three questions ready.