# plan.md — A Beginner's Plan for Speculation

**Educational guidance, not personalised financial advice.** Nothing here tells you what to buy or
sell. It is a framework for thinking, researching and documenting. Speculation can and does result
in total loss of capital.

Last reviewed: 28 July 2026

---

## 1. What speculation actually is

Speculation is buying an asset primarily because you expect **other people's beliefs about it to
change**, not because you expect to be paid by the asset itself.

An investment has a claim on something: earnings, rent, coupons, a share of a productive business.
You can be right about an investment even if nobody else notices, because the cash flow eventually
arrives. A speculation has no such backstop. If the crowd never arrives, you are left holding the
thing.

That is not a moral distinction. It is a mechanical one, and it has three consequences:

1. **Time works against you.** An investment compounds while you wait. A speculation burns cash,
   dilutes shareholders, or bleeds carry while you wait.
2. **You cannot value your way to safety.** There is no "cheap" in a pre-revenue explorer or a
   memecoin. Price is set entirely by narrative and flow.
3. **Exit is part of the thesis.** For an investment, "hold forever" is a valid answer. For a
   speculation, it is a missing plan.

### The honest definition

If you cannot answer *"what does this asset pay me if the story never spreads?"* with a number
greater than zero, you are speculating. Say so out loud. Most bad outcomes start with someone
telling themselves they are investing.

---

## 2. Speculation vs investing — the practical differences

| | Investing | Speculation |
|---|---|---|
| Source of return | Cash flows the asset produces | Change in what others will pay |
| Main risk | You mispriced the cash flows | The story dies, or was never true |
| Time horizon | Years to decades; time is an ally | Weeks to quarters; time is a cost |
| Position sizing | Can be large, diversified | Must be small, assume zero |
| Research edge | Financial analysis, industry depth | Narrative timing, flow, catalyst mapping |
| Being early | Usually fine | Often indistinguishable from being wrong |
| Exit plan | Optional | Mandatory, written before entry |
| Correct base rate | Most positions work out | **Most positions go to zero or near it** |

The last row is the one people skip. A well-run speculative book expects the majority of positions
to lose money. It survives because the sizing is small and the rare winners are held long enough to
matter. If your speculative process assumes a high hit rate, it is not a process, it is optimism.

---

## 3. Risk psychology — the part that actually decides your outcome

Your research quality matters less than your behaviour under pressure. The failure modes are
predictable, which is good news: predictable things can be pre-committed against.

**The seven traps, and the pre-commitment that defuses each:**

1. **Position size creep.** After two wins you size up, and the third loss erases both.
   *Defuse:* fix a dollar cap per position and per theme, in writing, before you feel confident.
2. **Averaging down into a broken story.** The thesis was narrative-driven; the narrative broke;
   you add anyway because the price is "better."
   *Defuse:* only average down when a *falsifiable* thesis point improved, never when only price fell.
3. **Sunk-cost narrative repair.** You quietly rewrite the reason you own it. The lithium play
   becomes a gold play becomes a "land bank."
   *Defuse:* your written thesis is immutable. If the reason changed, you closed and reopened.
4. **Confirmation nesting.** You end up only reading the ticker's own forum, where every holder is
   long and every seller is a "manipulator."
   *Defuse:* for each position, write the strongest bear case you can find, sourced from someone
   who is actually short.
5. **Anchoring on your entry.** The market does not know or care what you paid.
   *Defuse:* ask daily, "would I open this position today at this price?" If no, that's information.
6. **FOMO from other people's screenshots.** Gains are posted, losses are not. Every feed you read
   is a survivorship-biased highlight reel.
   *Defuse:* assume the visible winner is one of ten silent losers.
7. **Confusing volatility with opportunity.** Something moving 40% a day is not "active," it is
   telling you the price is close to meaningless.
   *Defuse:* volatility raises your risk tier, it does not raise your score.

**The single most useful rule:** decide what you would do at −50% *before* you enter, and write it
down. Not "I'd assess" — a decision. People who write it down mostly follow it. People who don't,
mostly freeze.

---

## 4. Categories of speculative assets

Ranked roughly by how much verifiable information exists (more information ≠ safer, but it does
mean your research can actually create an edge).

**A. Listed microcaps and small caps** — ASX and US. Real disclosure regimes, real filings,
real announcements. On the ASX this is dominated by resource explorers, junior biotech and
early-stage tech. You can genuinely out-read the crowd here because most participants read nothing.

**B. Pre-revenue thematic equities** — quantum, space, nuclear/SMR, defence tech, neuromorphic
chips. Listed and liquid, but valued entirely on a story about the 2030s. Volatility of 30–50% is
routine and not a signal.

**C. Crypto narratives** — token sectors that rotate on a roughly 6–14 week cycle. Highest
information asymmetry, most reflexive, thinnest protections. Includes the memecoin end, where the
asset is explicitly nothing but attention.

**D. Prediction markets** — Kalshi, Polymarket and the regulated venues. Structurally different:
positions expire and resolve, so you get scored objectively and quickly. Excellent training ground
for calibration, poor vehicle for compounding.

**E. Collectibles** — graded cards, watches, memorabilia. Slow, illiquid, high friction (grading,
authentication, 10–20% round-trip costs), and increasingly regulated. Long holding periods and
genuine physical risk.

**F. Emerging and frontier markets** — country and sector bets where currency, custody and
governance risk sit on top of the market risk.

**G. Pre-IPO and private secondaries** — access is the whole game; retail entry points usually
carry fee layers that eat the edge.

---

## 5. How a beginner should actually start

**Step 1 — Fund it from a separate account.** Speculative capital is money whose complete loss
changes nothing about your life. Not "would hurt but survivable." Nothing. If you cannot answer
what that number is, it is zero for now.

**Step 2 — Paper trade for one full cycle first.** A cycle, not a fortnight. You need to see a
narrative be born, get crowded, and die. Roughly one quarter minimum. Track it in a real journal
with real timestamps, because the retrospective memory of your own conviction is unreliable.

**Step 3 — Set the arithmetic before the ideas.** A common structure: speculation is a small,
capped slice of total net worth; no single position exceeds a fixed fraction of that slice; no
single *theme* exceeds a larger fixed fraction. The specific numbers are yours to choose. The
principle is that they exist and are written down before you have an opinion about any ticker.

**Step 4 — Take your first position deliberately small.** Small enough to be boring. The purpose of
the first position is to observe your own behaviour, not to make money.

**Step 5 — Journal every entry and exit.** Non-negotiable. Section 8 covers the format.

**Step 6 — Review monthly, adjust quarterly.** Judge the *process*, not the P&L. A profitable
position taken for a bad reason is a bad position that got lucky, and it will teach you a lesson
that costs you later.

---

## 6. Building a research routine

The point of a routine is to make research a habit rather than a reaction to price movement. If you
only research after something has already moved, you will systematically buy late.

**Daily (15 minutes)**

- Company announcements for your watchlist (ASX announcements platform, SEC filings)
- One sentiment sweep: what changed in tone, not in price
- Log anything that moved >15% without news — that is either information leaking or a liquidity event

**Weekly (60–90 minutes)**

- Narrative review: what is rising, what has gone quiet, what has become consensus
- Update the catalyst calendar — dated, known events for the next 90 days
- Read one bear case in depth on something you own
- Cull the watchlist. Watchlists rot; a 60-name list is a list you don't read

**Monthly (2–3 hours)**

- Score every open position again from scratch, ignoring your entry price
- Compare each position's current score to its score at entry — a falling score is your earliest exit signal
- Review the journal: which *reasons* worked, not which tickers
- Check theme concentration. Six positions in five different critical-minerals stories is one position

**Quarterly**

- Full hit-rate and average-win/average-loss review
- Re-examine your sizing rules against actual behaviour
- Ask honestly whether the edge is real or the market just went up

---

## 7. How to track narratives

A narrative is a compressed story that lets people buy without doing work. Your job is to identify
it early, judge how much runway it has, and know when it has become consensus.

**Narrative lifecycle — the five stages**

1. **Latent** — a real change occurs (policy, technology, supply shock). Almost nobody connects it
   to specific tickers. Best risk/reward, hardest to distinguish from noise.
2. **Emergent** — specialist forums, niche newsletters and a few analysts name the theme. Ticker
   lists begin circulating. This is where research effort pays most.
3. **Acceleration** — mainstream financial media picks it up. New listings, capital raisings and
   name-changes appear to serve the demand. Prices go vertical.
4. **Saturation** — the story appears in general news, an ETF launches for it, and the marginal
   buyer knows nothing about it. Good news stops moving prices — *this is the key tell*.
5. **Decay** — the narrative doesn't get disproven so much as it gets boring. Volume falls first,
   price follows.

**Practical tracking method**

- Keep a one-line definition of each narrative and the date you first logged it
- List the specific, dated catalysts that would advance it, and the events that would kill it
- Track *mentions* and *tone* separately from price. Rising mentions with falling price is a
  distribution warning; rising price with flat mentions can indicate genuine accumulation
- Note when the *quality* of discussion drops — when the questions get less technical, you are late

**The saturation test:** when a company in the theme releases genuinely good news and the stock
closes flat or down, the narrative has run out of new buyers. That signal is worth more than any
valuation work you will do.

---

## 8. Documenting decisions

Use one file per position. Written before entry, appended never rewritten.

```
TICKER / ASSET:
Date opened:
Category (A–G) and risk tier:
Position size ($ and % of speculative capital):

THE STORY IN ONE SENTENCE:

WHY NOW (the catalyst, with a date if it has one):

WHAT MUST BE TRUE for this to work (3 falsifiable points):
  1.
  2.
  3.

WHAT WOULD PROVE ME WRONG (specific, observable, not "if it goes down"):

THE STRONGEST BEAR CASE (sourced from someone who disagrees):

MY PLAN AT −50%:            (decided in advance, not "assess")
MY PLAN IF IT DOUBLES:      (decided in advance)
TIME STOP:                  (a date by which the thesis must show progress)

ENTRY SCORE (see spec.md):   /100

--- APPEND ONLY BELOW ---
[date] Observation / score change / action taken and why
```

The time stop is the underrated field. Most speculative losses are not dramatic collapses; they are
slow bleeds in positions whose story quietly stopped progressing while the holder was distracted.

---

## 9. Traps to avoid

**Structural traps**

- **Illiquidity** — check average daily traded value, not market cap. If your intended position is a
  meaningful share of a day's volume, you cannot exit at the price you see on screen.
- **Serial dilution** — a pre-revenue company's real product is shares. Check the raise history. Two
  placements a year at successive discounts means your upside is being issued to someone else.
- **Perpetual near-term milestones** — "results expected next quarter," every quarter, for four years.
  Read old announcements, not just new ones.
- **Shell recycling** — a company that has been a mining explorer, then a cannabis play, then an AI
  play, is selling narratives, not building anything.

**Social traps**

- **Ramping** — coordinated promotion, often disguised as research. Anonymous accounts with strong
  price targets and no falsifiable claims.
- **Paid research and "sponsored" coverage** — check the disclosure at the bottom. It is always there.
- **The influencer's entry price** — they are, by definition, in earlier than you and their exit is
  not a public event.
- **Forums as due diligence** — a ticker's own forum is a room of holders. Useful for sentiment,
  worthless for evidence.

**Cognitive traps**

- Mistaking a *sector* being real for a *company* being real. Rare earths matter; that says nothing
  about whether a specific explorer has an economic deposit.
- Mistaking a big market for a big company. "The TAM is $400bn" is a sentence that has preceded
  more losses than almost any other.
- Treating price action as evidence. In a thin market, price is the opinion of whoever traded last.

**The meta-trap:** believing that reading about traps immunises you against them. It does not. Only
written pre-commitments and small sizing do.

---

## 10. The six-week learning pathway

Detail and exercises for each week are in `spec.md` and rendered on the page in `index.html`.

- **Week 1 — Understanding speculation.** Definitions, base rates, why most positions fail, sizing arithmetic.
- **Week 2 — Understanding narratives.** Lifecycle stages, saturation tests, three live narratives tracked daily.
- **Week 3 — Understanding catalysts.** Dated vs undated events, building a 90-day catalyst calendar, "sell the news."
- **Week 4 — Building a research routine.** Sources, filters, the daily/weekly/monthly cadence, watchlist hygiene.
- **Week 5 — Tracking sentiment.** Measuring tone vs price, divergence, crowding, sentiment heatmaps.
- **Week 6 — Evaluating opportunities.** Full scoring framework applied to five live ideas, with written journals.

At the end of six weeks you should be able to look at any speculative idea and, within ten minutes,
state its narrative stage, its next dated catalyst, its liquidity, its risk tier and its score —
and then explain what would make you wrong.

---

## Related files

- `skills.md` — how this research system is built and what it runs each day
- `spec.md` — deep dive on categories, risk tiers, sentiment and the scoring framework
- `history.md` — append-only log of every opportunity this system has flagged
- `index.html` — the live page, updated daily

---

*Educational guidance, not personalised financial advice. Speculative assets can lose all of their
value. Verify everything independently before acting on it.*