Skip to content

Lesson 16 of 34 · Module 6 · Shark forward map

Build a driver tree (not a price tree)

If you cannot name drivers, you cannot weight scenarios.

A shark map starts with what moves the shop, not with 'it will go up'. Volumes, input costs, asset quality, pricing, tax, and funding costs are drivers. A driver is a lever you can point at in a filing or a Books line and say: if this changes, the business changes. In this lab you open a desk and write drivers in plain words, and you write unknown whenever you honestly do not know.

About 17 min · desk walk + key terms · Shark methodology

Do now: Practice on HBL Study Mode →

A shark map starts with what moves the shop, not with 'it will go up'. Volumes, input costs, asset quality, pricing, tax, and funding costs are drivers. A driver is a lever you can point at in a filing or a Books line and say: if this changes, the business changes. In this lab you open a desk and write drivers in plain words, and you write unknown whenever you honestly do not know.

After this lesson, you can…

  • ✓Explain build a driver tree (not a price tree) in plain words
  • ✓Apply the idea to one real PSX company
  • ✓Recognise the most common beginner mistake

Visual mental model

A calm learning loop

  1. 1

    Understand

    If you cannot name drivers, you cannot weight scenarios.

  2. 2

    Check

    What did this lesson train?

  3. 3

    Apply

    Use one real desk or calculator

Move from explanation to evidence, then practice.

Key knowledge

Terms you will use on the desk

Driver
A business lever (volumes, costs, asset quality) — not “market will rise”.
Unknown
Write unknown when Books/Filings are empty — do not invent.

driver tree — study, not tip

This lab trains driver tree on a live desk. Write dated evidence, refuse undated tip calendars, and keep scenario weights as business-path opinion — never as % chance the price moves.

In-depth teaching

5 sections

01

What a driver is, and what it is not

A driver is a business lever: something that, when it moves, changes the company's sales, costs, or cash in a way you could eventually see in the numbers. For a bank, the policy rate, the mix of deposits, and the quality of loans are drivers. For a fertilizer maker, gas availability and price, plus farmer demand and volumes, are drivers. For a cement company, local construction demand, coal and freight costs, and pricing are drivers.

A driver is not 'market will rise', 'good news coming', or 'the group says buy'. Those are moods or claims. A driver should be something you can check later against Filings or Books. If you cannot imagine where the proof would appear, it is not yet a driver; it is a hope. Write the hope down if you want, but label it as a hope and keep it out of the tree.

02

Build the tree: unit economics first

Start from the shop's basic arithmetic. What does one unit of the business sell, at what price, and what does it cost to make or serve? For a manufacturer, think of volume times price minus input and energy costs, minus finance costs, minus tax. For a bank, think of interest income minus interest paid, plus fees, minus operating costs and provisions for bad loans. Each term in that sentence is a branch of the tree.

Write at least two drivers per company, each with a one-line explanation and a pointer to where you would check it. Example format: 'Driver: funding cost of deposits; check: interest expense and deposit mix in Books/notes; status: unknown until I open Books.' If Books are empty, write unknown. Do not invent numbers to make a branch look complete. An honest tree with gaps is stronger than a full tree built on guesses.

Pause and predict · free

Getting check ready…

A forward map is…

03

Worked mini-example: HBL driver tree in plain words

Take HBL as a teaching example. Open the desk, note the as_of date, and check Filings and Books. A teaching tree might read: Driver one, net interest margin, meaning the gap between what the bank earns on loans and securities and what it pays on deposits; check interest income and expense lines. Driver two, asset quality, meaning how many loans may not be repaid; check provisions and any non-performing loan notes. Driver three, cost control; check operating expenses against income over several periods.

Beside each driver write a status: 'seen in Books', 'mentioned in Filings', or 'unknown'. Also write what a bad change would look like, for example 'provisions rising faster than income would weaken the base path.' Notice there is no price in the tree. A driver tree explains why the business might earn more or less. It does not say what the market will pay for those earnings.

04

Link each driver to Books and Filings

Every driver needs an evidence address. If you write 'energy cost' as a driver for cement, the address might be the cost of sales lines in Books and any tariff or coal-related notes in Filings. If you write 'provisions' for a bank, the address is the provision line and the auditor and notes sections. Without an address, the driver cannot later be tested, and an untestable driver is just a belief.

This link is also what makes the next labs possible. Catalysts are dated events that may move a driver. Scenarios are paths where drivers move differently. Falsifiers are the filing or book changes that would show a driver going the wrong way. If your tree has no evidence addresses, those later lessons have nothing to attach to, and the whole forward map becomes decoration.

05

Tip-culture trap: index mood as a company driver, and the fork

The scar is listing market-wide mood as a company driver. Tip rooms say 'index will break out so this will fly', turning a general feeling into a company thesis. Index mood can influence all prices, but it tells you nothing about a particular company's sales or costs, and it cannot be falsified with a filing. Refuse it as a driver and treat it as weather.

Wait means you cannot name two real drivers or you cannot say where to check them. Need evidence means you have drivers but Books are empty or thin and many statuses are unknown. Broker talk is not decided at this lab; a tree is only the first block of the forward map. A tree with honest unknowns and clear evidence addresses is a strong output even when the decision remains Wait.

Check your understanding · free

Getting check ready…

Scenario weights mean…

Scar · refuse this

Trap: Listing ‘market will rise’ as a driver.

Refuse: Refuse market-mood drivers.

Tip rooms sell index mood as a company thesis.

Required lab

Do this on the desk now

  • Open HBL. Write as_of.

    Done when: as_of written

    Open →
  • Open HBL Books or note empty.

    Done when: Books checked

    Open →

Lab · Forward map

Build your map: drivers, catalysts, BBB, falsifiers, exit, decision.

Practice on HBL Study Mode →

Loading forward map lab…

Or practice on live desk with Study ModeOpen HBL Study Mode

Worked example

LUCK Driver Tree: Volumes + Energy

Step-by-step walkthrough on LUCK. Open the desk alongside to follow.

Steps (4)▼

1. Identify the business model

LUCK = cement. Revenue = volumes × price. Cost = coal/freight + power + fixed.

💡 Driver tree starts with revenue levers, then cost levers.

2. Write ≥2 revenue drivers

Driver 1: 'Local cement demand (dispatches)'. Driver 2: 'Retention price per ton'.

💡 Drivers are business levers you can check in filings/books.

3. Write ≥2 cost drivers

Driver 3: 'Coal/freight cost per ton'. Driver 4: 'Power/fuel cost'.

💡 Cost drivers often matter more than revenue drivers in cement.

4. Link to evidence

Demand → APCMA monthly dispatches. Price → quarterly results. Coal → import data + quarterly notes.

💡 Each driver must point to a checkable source.

Show completed forward map

Drivers: Local cement demand (dispatches); Retention price per ton; Coal/freight cost per ton; Power/fuel cost

Catalysts: APCMA monthly dispatch data (monthly); Quarterly results (price/cost) (quarterly); Coal price trend (ongoing)

Bull: Demand holds; energy stable; price > cost

Base: Normal cement cycle; cost/price balanced

Bear: Demand drop or coal/freight shock

Weights: Bull 20% / Base 55% / Bear 25%

Rationale: Base until cost/demand evidence flips. Coal is the swing variable.

Falsifiers: Coal shock without price pass-through; Dispatches decline 2+ months

Exit: LUCK liquid among cement — lot 500, avg vol ~500k/day. Exit realistic.

Decision: need_evidence

Failed map examples — learn from scars

Real mistakes that fail the validator. See the error, understand the fix.

These are real beginner mistakes that the MapValidator catches. Study them so you don't repeat them.

PROPHECY

Prophecy Language in Rationale

HBL · 5 error(s)

▼

❌ Failed map — 5 validation errors

drivers: [ "Net interest margin", "Asset quality" ]
catalysts: [ { "event": "Q3 results", "whenOrUnknown": "2024-10-24" } ]
scenarios: { "bull": "Stock will go up 20% after clean results", "base": "Price will stay range-bound", "bear": "Will drop 15% if provisions spike" }
scenarioWeights: { "bull": 60, "base": 30, "bear": 10 }
weightsRationale: HBL will go up because provisions are normalizing and NIM is strong. Target 200.
falsifiers: [ "Market crashes", "Bad news" ]
liquidityExit: HBL is liquid
decision: broker_talk
  • Rationale contains prophecy: 'will go up', 'target 200'
  • Scenarios describe price targets, not business paths
  • Weights reflect price odds (60/30/10), not business path opinions
  • Falsifiers are vague ('market crashes')
  • Decision 'broker_talk' without completing evidence steps
💡 Teaching note

Prophecy veto is automatic. Every field must describe business paths, not price predictions. Weights are your opinion on business outcomes — not % chance price moves.

VAGUE-FALSIFIER

Vague Falsifiers

FFC · 4 error(s)

▼

❌ Failed map — 4 validation errors

drivers: [ "Gas cost", "Volumes" ]
catalysts: [ { "event": "Gas price notification", "whenOrUnknown": "unknown" } ]
scenarios: { "bull": "Gas manageable, volumes hold", "base": "Normal agri cycle", "bear": "Things go wrong" }
scenarioWeights: { "bull": 30, "base": 50, "bear": 20 }
weightsRationale: Base case until something changes
falsifiers: [ "Market crashes", "Economy gets bad", "Bad news" ]
liquidityExit: FFC is liquid
decision: need_evidence
  • Falsifiers are generic ('market crashes', 'economy gets bad') — not checkable
  • Bear scenario 'things go wrong' is not a business path
  • Weights rationale 'until something changes' is not a working case explanation
  • Catalyst 'unknown' without falsifier for 'no gas notification in 90 days'
💡 Teaching note

Falsifiers must be specific, dated, and checkable: filing events, book lines, governance signals. 'Market crashes' is lazy — name the filing or book line that kills your base case.

WEIGHT-ERROR

Weights Don't Sum to 100

LUCK · 3 error(s)

▼

❌ Failed map — 3 validation errors

drivers: [ "Cement demand", "Coal cost" ]
catalysts: [ { "event": "APCMA dispatch data", "whenOrUnknown": "monthly" } ]
scenarios: { "bull": "Demand up, coal stable", "base": "Normal cycle", "bear": "Cost shock" }
scenarioWeights: { "bull": 40, "base": 40, "bear": 30 }
weightsRationale: Bullish on demand
falsifiers: [ "Coal spike", "Demand drop" ]
liquidityExit: LUCK liquid
decision: broker_talk
  • Weights sum to 110 (40+40+30), not 100
  • Decision 'broker_talk' with invalid weights
  • Rationale 'bullish on demand' doesn't explain base/bear weights
💡 Teaching note

Weights must sum to exactly 100. The validator catches this automatically. Your rationale must explain why base is the working case, not just 'bullish on X'.

MISSING-DRIVER

Missing Key Driver (Circular Debt)

OGDC · 4 error(s)

▼

❌ Failed map — 4 validation errors

drivers: [ "Oil price", "Production volumes" ]
catalysts: [ { "event": "Monthly production data", "whenOrUnknown": "monthly" } ]
scenarios: { "bull": "Oil up, production up", "base": "Stable", "bear": "Oil down" }
scenarioWeights: { "bull": 30, "base": 50, "bear": 20 }
weightsRationale: Oil price drives everything
falsifiers: [ "Oil crash", "Production drop" ]
liquidityExit: OGDC liquid
decision: need_evidence
  • Missing critical driver: circular debt receivables / cash collection
  • OGDC is an E&P — cash collection from refineries is the real driver, not just oil price
  • Falsifiers don't mention receivables aging or government payment delays
  • Liquidity exit ignores that OGDC shares can be thin in panic
💡 Teaching note

Every sector has hidden drivers. For E&P: cash collection = production × price × collection rate. Missing receivables driver = incomplete map. Always ask: 'What kills cash flow even if price is good?'

NO-EXIT

No Exit Plan for Thin Name

THIN · 4 error(s)

▼

❌ Failed map — 4 validation errors

drivers: [ "Unknown — need filings" ]
catalysts: [ { "event": "None known", "whenOrUnknown": "unknown" } ]
scenarios: { "bull": "Could be anything", "base": "Wait and see", "bear": "Risks everywhere" }
scenarioWeights: { "bull": 10, "base": 20, "bear": 70 }
weightsRationale: Too many unknowns
falsifiers: [ "Any bad news" ]
liquidityExit: Will figure out later
decision: broker_talk
  • Liquidity exit 'will figure out later' is not a plan
  • Decision 'broker_talk' on thin data with no exit = dangerous
  • Drivers 'unknown' but decision not 'wait'
  • No lot size, volume, or free float analysis
💡 Teaching note

Thin data → Wait is the only valid decision. If you don't know lot size, volume, and free float, you cannot size a position. 'Broker talk' on unknown liquidity = gambling.

THIN-DATA-INVENTED

Inventing Numbers on Thin Coverage

THIN · 4 error(s)

▼

❌ Failed map — 4 validation errors

drivers: [ "Revenue growth 15%", "Margins expanding" ]
catalysts: [ { "event": "Next quarter results", "whenOrUnknown": "2024-12-15" } ]
scenarios: { "bull": "Growth continues", "base": "Stable", "bear": "Slowdown" }
scenarioWeights: { "bull": 40, "base": 40, "bear": 20 }
weightsRationale: Management guided growth
falsifiers: [ "Results miss" ]
liquidityExit: Assume liquid
decision: need_evidence
  • Drivers invented ('Revenue growth 15%') — Books are empty, no data to support
  • Catalyst date invented — no filing confirms Dec 15
  • Weights reflect optimism, not evidence
  • Liquidity exit 'assume liquid' on unknown name = dangerous
💡 Teaching note

Thin coverage → Wait. Honesty beats invented drivers. If Books are empty, write 'unknown — need filings/books'. Never invent numbers to make a map look complete.

What you can say now

Evaluation language — not a price call.

  • →I can name at least two business drivers — or write unknown.
  • →‘Market will rise’ is not a company driver.

Worth keeping

  • Veterans keep scars next to every step.
  • A complete map still is not a price call.
  • A driver tree is closer to a mechanic's diagnostic list than to a prediction; it tells you where to look when something changes.
  • Writing 'unknown' honestly is often the most valuable line on a beginner's tree.

Write Wait when evidence is thin — that is skill, not weakness.

Try it on a live desk

Practice · support fades as you go

Practice with support

Write Wait when evidence is thin — that is skill, not weakness.

Need a hint?

A scar-backed habit on the live desk — not a tip.

Done when: You can explain what you checked and what remains uncertain.

Now without hints

Try it independently

What did this lesson train?

A strong answer: A scar-backed habit on the live desk — not a tip.

Lesson FAQ

What did this lesson train?

A scar-backed habit on the live desk — not a tip.

Does a strong map mean buy?

No. It means homework. Decision may still be Wait.

Are weights price odds?

No. They are opinion on business paths only.

How many drivers should a beginner write?

Two to four real ones is plenty. Quality matters more than count: each driver should have a plain explanation and a place to check it.

What if I do not know what drives a sector?

Use the sector literacy page and the later sector lens lessons for starter questions, and write unknown where the desk cannot answer.

Mark this lesson complete

What did this lesson train?

Show calm answer

A scar-backed habit on the live desk — not a tip.

A scar-backed habit on the live desk — not a tip.

Tick every required lab step above before you mark complete.

Sign in before completing lessons if you want them counted toward the certificate.

Optional: graded test (sign-in)

Member lesson tools

Full lesson test & member tools

Checking sign-in…

Ask coach

Locked

Ask: review my map gaps

Share drivers and falsifiers — coach checks process.

Sign in to unlock →

Stay on the course path

Follow the Shark Methodology path. How-tos and metrics are optional reference.

  • Shark Methodology course

    Main beginner path: evidence, forward maps, and evaluating claims — not competing starts.

  • Learn PSX hub

    Start Lesson 1, then optional guides and tools when you need them.

  • PSX today

    Pakistan stock market today in plain words - breadth, gainers, losers.

  • PSX companies list

    Browse all Pakistan Stock Exchange listed companies by name or sector.

  • KSE 100 index today

    Educational constituent board - not the official index level on psx.com.pk.

  • PSX gainers today

    Biggest % rises among ordinary listed shares - homework, not a buy list.