Pakistan Fashion Price Index

A naive “average price this month” is worthless here. A brand adding five thousand cheap items would read as deflation even if nothing on the shelf changed price. So the index only ever compares a product to itself.

1. Take each product's price in every calendar month, from our own recorded price history. We write a price down when it changes, so a month with no new reading means the price did not move and the previous one carries forward. A product leaves the calculation after the month we last saw it in the catalog.
2. Keep only matched pairs — products present in both this month and last. New arrivals and delisted items are excluded from the month they appear or vanish, so a brand dumping 5,000 cheap items into the catalog cannot read as deflation.
3. Compute the 2%-trimmed geometric mean of those price ratios — a Jevons index. Geometric so a doubling and a halving cancel out; trimmed so one mis-scraped price cannot move the reading.
4. Chain the result onto the previous month. April 2026 is fixed at 100.

Months with fewer than 300 matched pairs are carried flat and marked estimated rather than dropped — dropping a month would break the chain and silently shift every reading after it.

Category series use the same method over the same data, restricted to products in that category. A product counts toward every category it belongs to, and toward the overall index.

Prices come from brand stores directly. Related: the Sale Honesty Report measures whether advertised discounts are real; Brand Insights ranks brands on price and stock today.