Chapter 7

Why now

Published 29 September 2026

Every issuer will tell you this is about "sustainability". They are not entirely wrong. They are just leaving out who is being sustained.

The economics are real, and ET's March 2026 review lays them out: reward programs were priced for an era of cheap customer acquisition, when banks burned interchange income and investor patience to buy market share. As portfolios matured, the bill came due. Points liabilities sit on the balance sheet. Lounge access costs hard cash per visit. Co-brand vouchers are negotiated line items. When a bank's board asks the cards division to show profit, the reward program is the largest discretionary cost on the page - and the only one that can be cut with a PDF.

There is also the behavioral math the industry does not say aloud. Reward points have breakage built in: a meaningful share of points is never redeemed. Devaluations increase breakage twice over - they make points worth less, and they make customers disengage, so more points expire unclaimed. A liability that melts on its own is, from a certain desk, not a scandal but a strategy. Every 30-day expiry, every 50% redemption cap, every excluded category is a small transfer from customers who did not read the notice to shareholders who did not have to.

The one time it failed

Which is what makes the Infinia episode so instructive. When HDFC Bank moved to gut the reward multipliers on its flagship card, the backlash was immediate and organized - affluent customers, personal-finance communities, and social media united in outrage, and the bank backed down. No regulator intervened; no rule was broken. The devaluation was stopped by the only force that has ever stopped one: customers who noticed, together, out loud.

Everyone else's cuts went through. The PNB cardholder whose point halved from ₹0.50 to ₹0.25 did not trend on social media. The SBI SimplyCLICK user whose Swiggy earn fell from 10x to 5x got an email. The Axis Magnus holder who watched the voucher, the milestone, the miles and the fee all move in one notice got a PDF. The quietness is the point: each cut is small enough to absorb alone, and only the sum of them - the grid - tells the real story.

That is why we built the measurement before the narrative. The Credit Card Maximizer dataset exists because nobody else was keeping score: 232 cards, their earn rates, caps, exclusions, redemption fees and point values, extracted from issuer documents and kept current as the notices land. This file is the story the dataset tells. The next notice is already being drafted somewhere; the dataset will catch it, and we will name it.