Every change to Indian card rewards since 1961

53 cuts against 46 improvements: 81.0 severity points against 50.1, so 81.0 ÷ (81.0 + 50.1) = 61.8%. One cut every 1.7 days across the period's 92.

By kind: devaluation 23 · fee 16 · lounge 14 · partner 17 · improvement 19 · launch 10. Basket: 10,000 points in each of 70 programmes, worth ₹6,10,700 at the open and ₹6,23,500 at the close ( +₹12,800). 95 programmes are left out because we had not started valuing them on 2026-07-01; they are counted, never estimated. Window: 2026-07-01 to 2026-09-30, by the date each change took effect.

Announced in this period, taking effect after it (so counted later): Club Vistara IDFC FIRST stops taking transactions on 1 October 2026 (announced 2026-09-17, from 2026-10-01) ; Amex Platinum loses Lufthansa Lounge access with no replacement (announced 2026-07-14, from 2026-10-01) ; SELECT's ₹3L milestone stops counting the spends it never rewarded (announced 2026-07-28, from 2026-10-01) ; Platinum Travel's domestic lounge access gets a ₹1 lakh prior-quarter spend gate (announced 2026-08-10, from 2026-10-01) ; Privilege's movie buy-one-get-one becomes a spend gate (announced 2026-07-28, from 2026-11-01) ; IndiGo Axis Premium's movie BOGO grows a ₹25,000 spend gate (announced 2026-08-15, from 2026-11-01) ; The movie-BOGO spend gate reaches My Zone too — ₹25,000 in the previous 3 months (announced 2026-08-15, from 2026-11-01) ; NEO's BookMyShow discount grows a ₹25,000 spend gate (announced 2026-08-24, from 2026-11-01) ; The Reserve's memberships stop renewing unless you say yes (announced 2026-07-28, from 2026-11-15) ; Amex India ends Membership Rewards pooling — each card keeps its own balance (announced 2026-09-11, from 2026-11-15) ; Amex MRCC loses its 50% renewal waiver between ₹90,000 and ₹1,49,999 (announced 2026-09-11, from 2026-11-15) .

The 99 changes themselves are in the year-by-year list: 2026 .

30 cuts against 10 improvements: 49.5 severity points against 15.3, so 49.5 ÷ (49.5 + 15.3) = 76.4%. One cut every 3.0 days across the period's 91.

By kind: devaluation 18 · fee 7 · lounge 3 · partner 6 · improvement 5 · launch 1. Basket: 10,000 points in each of 69 programmes, worth ₹6,08,200 at the open and ₹6,08,200 at the close ( ₹0). 96 programmes are left out because we had not started valuing them on 2026-04-01; they are counted, never estimated. Window: 2026-04-01 to 2026-06-30, by the date each change took effect.

No programme in the basket had a priced ₹ move in FY2027 Q1. That is not a claim that nothing was cut: 30 cuts landed, and they count in full in the share above.

Announced in this period, taking effect after it (so counted later): PhonePe SBI SELECT BLACK stops paying on UPI outside PhonePe (announced 2026-05-29, from 2026-07-01) ; PhonePe SBI PURPLE stops paying on UPI outside PhonePe (announced 2026-05-29, from 2026-07-01) ; BOBCARD nearly doubles the spend gate on unlimited domestic lounges (announced 2026-06-12, from 2026-07-15) .

The 40 changes themselves are in the year-by-year list: 2026 .

48 cuts against 12 improvements: 85.8 severity points against 14.4, so 85.8 ÷ (85.8 + 14.4) = 85.6%. One cut every 7.6 days across the period's 365. Among the 4 financial years with at least 20 changes, it ranks first.

By kind: devaluation 35 · fee 4 · lounge 6 · partner 8 · improvement 2 · launch 5. Basket: 10,000 points in each of 69 programmes, worth ₹6,08,800 at the open and ₹6,08,200 at the close ( −₹600). 96 programmes are left out because we had not started valuing them on 2025-04-01; they are counted, never estimated. Window: 2025-04-01 to 2026-03-31, by the date each change took effect.

Announced in this period, taking effect after it (so counted later): SBI Card starts charging late fees on dues of ₹101–₹500 (announced 2026-03-29, from 2026-05-01) ; Aeroplan's expiry pause ends on 30 November — 18 months idle and the balance lapses (announced 2025-11-04, from 2026-11-30) .

The 60 changes themselves are in the year-by-year list: 2025 , 2026 .

21 cuts against 3 improvements: 30.3 severity points against 4.5, so 30.3 ÷ (30.3 + 4.5) = 87.1%. One cut every 4.3 days across the period's 90.

By kind: devaluation 15 · fee 1 · lounge 2 · partner 3 · launch 3. Basket: 10,000 points in each of 69 programmes, worth ₹6,08,200 at the open and ₹6,08,200 at the close ( ₹0). 96 programmes are left out because we had not started valuing them on 2026-01-01; they are counted, never estimated. Window: 2026-01-01 to 2026-03-31, by the date each change took effect.

No programme in the basket had a priced ₹ move in FY2026 Q4. That is not a claim that nothing was cut: 21 cuts landed, and they count in full in the share above.

Announced in this period, taking effect after it (so counted later): SBI Card starts charging late fees on dues of ₹101–₹500 (announced 2026-03-29, from 2026-05-01) .

The 24 changes themselves are in the year-by-year list: 2026 .

6 cuts against 4 improvements: 8.7 severity points against 4.2, so 8.7 ÷ (8.7 + 4.2) = 67.4%. One cut every 15.3 days across the period's 92.

By kind: devaluation 3 · fee 1 · lounge 2 · partner 1 · improvement 2 · launch 1. Basket: 10,000 points in each of 69 programmes, worth ₹6,08,800 at the open and ₹6,08,800 at the close ( ₹0). 96 programmes are left out because we had not started valuing them on 2025-07-01; they are counted, never estimated. Window: 2025-07-01 to 2025-09-30, by the date each change took effect.

No programme in the basket had a priced ₹ move in FY2026 Q2. That is not a claim that nothing was cut: 6 cuts landed, and they count in full in the share above.

Announced in this period, taking effect after it (so counted later): HDFC kills the 6E Rewards IndiGo co-brands — both tiers, and the currency with them (announced 2025-08-26, from 2025-11-01) .

The 10 changes themselves are in the year-by-year list: 2025 .

15 cuts against 2 improvements: 32.7 severity points against 3.0, so 32.7 ÷ (32.7 + 3.0) = 91.6%. One cut every 6.1 days across the period's 91.

By kind: devaluation 13 · fee 1 · lounge 1 · partner 2. Basket: 10,000 points in each of 69 programmes, worth ₹6,08,800 at the open and ₹6,08,800 at the close ( ₹0). 96 programmes are left out because we had not started valuing them on 2025-04-01; they are counted, never estimated. Window: 2025-04-01 to 2025-06-30, by the date each change took effect.

No programme in the basket had a priced ₹ move in FY2026 Q1. That is not a claim that nothing was cut: 15 cuts landed, and they count in full in the share above.

The 17 changes themselves are in the year-by-year list: 2025 .

17 cuts against 7 improvements: 30.6 severity points against 9.3, so 30.6 ÷ (30.6 + 9.3) = 76.7%. One cut every 21.5 days across the period's 365. Among the 4 financial years with at least 20 changes, it ranks #2.

By kind: devaluation 11 · fee 3 · lounge 2 · partner 4 · improvement 2 · launch 2. Basket: 10,000 points in each of 62 programmes, worth ₹5,54,800 at the open and ₹5,54,800 at the close ( ₹0). 103 programmes are left out because we had not started valuing them on 2024-04-01; they are counted, never estimated. Window: 2024-04-01 to 2025-03-31, by the date each change took effect.

No programme in the basket had a priced ₹ move in FY2025. That is not a claim that nothing was cut: 17 cuts landed, and they count in full in the share above.

The 24 changes themselves are in the year-by-year list: 2024 , 2025 .

15 cuts against 12 improvements: 28.2 severity points against 14.7, so 28.2 ÷ (28.2 + 14.7) = 65.7%. One cut every 24.4 days across the period's 366. Among the 4 financial years with at least 20 changes, it ranks #3.

By kind: devaluation 15 · improvement 2 · launch 10. Basket: 10,000 points in each of 56 programmes, worth ₹5,26,300 at the open and ₹5,22,300 at the close ( −₹4,000). 109 programmes are left out because we had not started valuing them on 2023-04-01; they are counted, never estimated. Window: 2023-04-01 to 2024-03-31, by the date each change took effect.

The 27 changes themselves are in the year-by-year list: 2023 , 2024 .

6 cuts against 8 improvements: 15.0 severity points against 13.8, so 15.0 ÷ (15.0 + 13.8) = 52.1%. One cut every 60.8 days across the period's 365.

By kind: devaluation 6 · partner 1 · improvement 1 · launch 6. Basket: 10,000 points in each of 50 programmes, worth ₹4,84,500 at the open and ₹4,83,300 at the close ( −₹1,200). 115 programmes are left out because we had not started valuing them on 2022-04-01; they are counted, never estimated. Window: 2022-04-01 to 2023-03-31, by the date each change took effect.

The 14 changes themselves are in the year-by-year list: 2022 , 2023 .

9 cuts against 12 improvements: 16.2 severity points against 25.8, so 16.2 ÷ (16.2 + 25.8) = 38.6%. One cut every 40.6 days across the period's 365. Among the 4 financial years with at least 20 changes, it ranks #4.

By kind: devaluation 6 · fee 3 · partner 4 · improvement 4 · launch 4. Basket: 10,000 points in each of 44 programmes, worth ₹4,27,500 at the open and ₹4,27,500 at the close ( ₹0). 121 programmes are left out because we had not started valuing them on 2021-04-01; they are counted, never estimated. Window: 2021-04-01 to 2022-03-31, by the date each change took effect.

No programme in the basket had a priced ₹ move in FY2022. That is not a claim that nothing was cut: 9 cuts landed, and they count in full in the share above.

The 21 changes themselves are in the year-by-year list: 2021 , 2022 .

9 cuts against 8 improvements: 21.0 severity points against 12.3, so 21.0 ÷ (21.0 + 12.3) = 63.1%. One cut every 40.7 days across the period's 366.

By kind: devaluation 7 · fee 1 · partner 1 · improvement 2 · launch 6. Basket: 10,000 points in each of 41 programmes, worth ₹4,06,000 at the open and ₹4,06,000 at the close ( ₹0). 124 programmes are left out because we had not started valuing them on 2019-04-01; they are counted, never estimated. Window: 2019-04-01 to 2020-03-31, by the date each change took effect.

No programme in the basket had a priced ₹ move in FY2020. That is not a claim that nothing was cut: 9 cuts landed, and they count in full in the share above.

The 17 changes themselves are in the year-by-year list: 2019 , 2020 .

6 cuts against 7 improvements: 15.0 severity points against 13.2, so 15.0 ÷ (15.0 + 13.2) = 53.2%. One cut every 60.8 days across the period's 365.

By kind: devaluation 5 · fee 1 · partner 2 · improvement 2 · launch 3. Basket: 10,000 points in each of 40 programmes, worth ₹4,04,500 at the open and ₹4,03,500 at the close ( −₹1,000). 125 programmes are left out because we had not started valuing them on 2018-04-01; they are counted, never estimated. Window: 2018-04-01 to 2019-03-31, by the date each change took effect.

The 13 changes themselves are in the year-by-year list: 2018 , 2019 .

6 cuts against 4 improvements: 10.2 severity points against 7.2, so 10.2 ÷ (10.2 + 7.2) = 58.6%. One cut every 60.8 days across the period's 365.

By kind: devaluation 6 · partner 2 · improvement 2. Basket: 10,000 points in each of 37 programmes, worth ₹3,76,700 at the open and ₹3,76,700 at the close ( ₹0). 128 programmes are left out because we had not started valuing them on 2016-04-01; they are counted, never estimated. Window: 2016-04-01 to 2017-03-31, by the date each change took effect.

No programme in the basket had a priced ₹ move in FY2017. That is not a claim that nothing was cut: 6 cuts landed, and they count in full in the share above.

The 10 changes themselves are in the year-by-year list: 2016 , 2017 .

5 cuts against 6 improvements: 13.5 severity points against 10.2, so 13.5 ÷ (13.5 + 10.2) = 57.0%. One cut every 73.2 days across the period's 366.

By kind: devaluation 5 · partner 2 · improvement 1 · launch 3. Basket: 10,000 points in each of 36 programmes, worth ₹3,66,700 at the open and ₹3,66,700 at the close ( ₹0). 129 programmes are left out because we had not started valuing them on 2015-04-01; they are counted, never estimated. Window: 2015-04-01 to 2016-03-31, by the date each change took effect.

No programme in the basket had a priced ₹ move in FY2016. That is not a claim that nothing was cut: 5 cuts landed, and they count in full in the share above.

The 11 changes themselves are in the year-by-year list: 2015 , 2016 .

cut share   = cut points ÷ (cut points + improvement points) × 100
cut points  = sum of severity weights of changes that went AGAINST the reader
severity    = major 3 · moderate 1.5 · minor 0.6
against     = devaluation, fee rise, lounge cut, hostile partner change
for         = improvement, launch, a partner added
basket ₹    = sum over programmes valued at the open of
              (value at close − value at open) × 10000

• Period: the Indian financial year, and its four quarters.
• Window: by the date a change takes effect; start exclusive, end inclusive.
• Weighting: every programme counts once; changes are weighted by severity.
• Basket: 10,000 points per programme, fixed on the period's first day.
• An edition needs 10 changes and a closed period; a superlative needs 20.

Worked on FY2026: 60 changes, 48 cuts worth 85.8 points and 12 improvements worth 14.4. 85.8 ÷ (85.8 + 14.4) = 85.6%.

1. A share, not a count. "N devaluations this year" measures how hard we looked as much as what happened: the archive holds one change for 1961 and over a hundred for the current year. A ratio does not rise just because we logged more. The severity weights are the same ones the programme pages' risk grades use (methodology).

2. The Indian financial year. 1 April to 31 March, the frame RBI and every listed bank report in, with every label spelling the months out. A change counts in the period it takes effect; one announced earlier is listed under "announced in this period". Only a closed period gets an edition.

3. Every programme counts once. Weighting by the cards that earn a programme was refused: the two biggest ₹ cuts in the archive hit Marriott Bonvoy (1 card) and Avios (BA / Qatar / Finnair) (4 cards), against 67 for RBL Reward Points, so that weight would shrink the biggest events to nothing. Weighting by cardholders needs numbers nobody publishes. Each edition prints the cards and routes behind each move instead, so you can re-weight by hand.

4. What counts. Cuts and improvements are both counted and shown. Fee rises and lounge cuts count in the share but never in the rupees. The ₹ figure is a floor. A move caused by the rupee shifting against a programme pegged to a foreign currency is real and is labelled "re-mark", not "event".

5. Two counts, never mixed. The share counts changes: one change is one, however many programmes it touches. The ₹ figure counts programme moves against a basket fixed at the open; a programme first valued mid-period is left out and counted.

• It is not weighted by what readers hold. Nobody publishes Indian membership by programme, so every programme counts once.
• It is not a forecast. A programme's forward risk grade is a different measure, on the methodology page.
• It is not the size of the damage: the ₹ figure is, and it is conservative. The worst single move on record is Marriott Bonvoy, 2022-03-29, −47.6%: −₹5,900 on 10,000 points.
• The direction of each change comes from the site's one classifier, edge cases included: a regulator freezing a bank's card sales reads as a partner change with no hostile word in it, so it counts as an improvement. We publish the classifier rather than patch it for one page.
• Editions are recomputed on every build. A programme we start valuing today with history back to 2015 can restate a past edition, as a real index restates.

• The change log above: 436 dated, sourced changes, 1961-01-01 to 2026-10-26 in force (2026-12-17 including announced ones).
• The dated point values behind points values: 165 programmes, 246 dated ₹-per-point marks, 23 of them moves.
• Both are open files in the site's public repository, and one module computes every figure here from them.

The whole series as JSON: /d/devaluation-index.json.