Core principle
We do not start with a conclusion.
The research process starts with observable data: PSX market files, company filings, NCCPL flow data, SBP releases, PBS macro data, official notices, and other primary sources where available. Interpretation comes after the facts are assembled.
The objective is not to make a market call. The objective is to explain the mechanism: what changed, how that change moved through the system, which sectors or companies were exposed, and what evidence should be watched next.
01Collect facts
02Verify source
03Identify mechanism
04Test risk
05Publish clearly
Research chain
The mechanism chain used in every analysis.
A market move is not enough. The explanation must show how the move travelled from trigger to impact.
Cause
The starting event: policy decision, macro print, earnings release, institutional flow, sector data, commodity move, or company filing.
Transmission
The path through which the event enters the market: rates, liquidity, currency, commodity costs, margins, sentiment, flows, or valuation multiples.
Market impact
The broad PSX effect: index level, volume, participation, advance-decline balance, leadership, concentration, and liquidity.
Sector impact
The sector-level effect: demand, input costs, regulation, pricing power, utilisation, working capital, and earnings sensitivity.
Company impact
The company-level effect: business model exposure, balance-sheet sensitivity, cash conversion, receivables, cost base, and valuation context.
Risk and watch item
The falsifier: the data point, policy reversal, flow shift, commodity move, or earnings result that could weaken the explanation.
How claims are handled
Facts and inferences are kept separate.
Verified fact
A verified fact is supported by a primary or official source. Examples include a PSX closing level, a company filing number, an NCCPL flow figure, an SBP policy rate, or a PBS inflation print.
Named inference
A named inference is an interpretation drawn from the evidence. It is not presented as certainty. The assumption behind the inference must be visible to the reader.
Example of the distinction
“KSE-100 closed at a stated level” is a fact if it comes from PSX data. “The move likely reflected positioning ahead of an event” is an inference and must be framed as one.
Research lenses
What we check before publishing.
Each lens answers a different question. Missing one can make the explanation incomplete.
Was the move broad, narrow, volume-backed, flow-led, or concentrated in a few index names?
Who bought, who sold, and whether the move reflects organic flow or a possible one-off transaction?
Which sectors gain or lose through revenue, cost, regulation, demand, or balance-sheet sensitivity?
Which company fundamentals are actually exposed: earnings, cash flow, receivables, debt, or valuation?
What evidence would prove the mechanism incomplete, overstated, or wrong?
Pre-publication checks
Nothing is final until it clears these checks.
Source check
Important figures must trace to primary or approved sources. If a figure cannot be verified, it is removed or clearly labelled as unavailable.
Mechanism check
The analysis must explain the chain, not just repeat the headline. Cause and transmission must be visible.
Figure consistency check
Numbers must match across the article, social post, infographic, metadata, and source trail.
Compliance check
The final text must not include buy or sell calls, target prices, entry or exit levels, guaranteed returns, or personalised advice.