Jaws Turn Positive and Loan Growth Hits 19.6%: The ADS Closed Unchanged Anyway

Results vs. Consensus

ICICI Bank released unaudited standalone and consolidated results for the quarter ended June 30, 2026 on Saturday, July 18, and held the analyst call the same evening. All figures below are the standalone accounts as filed unless marked consolidated. The Bank reports in Indian rupees under Indian GAAP; one crore is ten million rupees, and where a dollar convenience translation appears it uses the rate of ₹94.66 per US dollar disclosed with the results. The joint statutory auditors issued an unmodified limited review report.

Q1 FY27 Scorecard

The consensus column blends the two named brokerage previews published in the days before the board meeting, which put profit at ₹13,062 crore and ₹13,093 crore, with the street number implied by post-print coverage describing the print as eleven percent above consensus. The provision consensus is the mean of the two houses that published a credit-cost assumption. The dollar rows are the vendor composite carried against the New York listing on a net revenue basis, meaning net interest income plus non-interest income already net of treasury; it is not comparable to the ₹54,247 crore Indian-GAAP total income line and is shown only because it is what a dollar-based screen would have seen.

Year-on-Year Comparison

Two lines carry this table. Interest expended fell 0.1% in absolute terms while the loan book grew 19.6%, which is what a deposit franchise repricing down through a rate cycle looks like when it works. And operating expenses grew 10.4% against core revenue of 13.5%, reversing the relationship that has defined the last two fiscal years. Last quarter the equivalent table showed total income up 1.8% and every line beneath it growing faster because provisions had disappeared. This one grows from the top.

Sequential Comparison

The sequential table is the mirror image of last quarter's. In March, profit before tax jumped 22.3% because provisions had collapsed to ₹96 crore. Here provisions rose thirteenfold off that base and profit before tax still grew 5.6%, because operating profit before provisions grew 12.0%. The four lines to keep are at the bottom: net non-performing assets up two basis points, coverage down 110, common equity tier-one down sixteen, and the loan-to-deposit ratio up 238. Every one of them is what growing a book 5.0% in a quarter costs.

Revenue assessment. The revenue quality this quarter is materially better than last, and the reason is that all three engines fired rather than one. Net interest income grew 12.7% on a book compounding 19.6% with funding cost falling; fee income grew 23.5% after a year of compounding at half the rate of the balance sheet; and non-interest income excluding treasury grew 16.0%. The honest deductions are that the fee comparison is flattered by a weak base management named without being asked, and that the corporate loan growth behind a good part of the interest income is bond-market substitution rather than new investment. Neither deduction changes the shape: a bank that was monetising each rupee of balance sheet less well every quarter through FY26 stopped doing so in Q1.

Margin assessment. Three quarters ago we could not tell whether the margin was holding or being held up. Now we can. The clean series management disclosed puts the core margin at 4.28% against 4.27% in the March quarter and 4.27% a year ago, which means the underlying margin has been flat for a year while cost of deposits fell 44 basis points and 57% of the domestic book sat on external benchmarks through a transmitted rate cut. That is genuine defence, not a refund artifact. The forward risk has changed shape rather than disappeared: management restated the range-bound frame but added a caveat it did not have in April, which is that mobilising foreign-currency non-resident deposits and levering them will expand the international balance sheet and compress the reported margin. A lower-margin, higher-earnings mix is the right trade, and it will look like a miss on the line the market watches.

EPS assessment. Basic EPS of ₹20.65 annualises to ₹82.60 and puts the Mumbai line at 17.7 times a run-rate that includes both assists. On trailing twelve-month earnings of ₹72.95 the multiple is 20.0 times. The composition question we asked in April now has a better answer: in the March quarter the provision line contributed more incremental profit than the entire operating business, whereas here core operating profit grew 15.6% and would have carried the quarter without any help below it. The two assists are worth roughly eight points of the 15.9% growth between them, so the underlying rate is high single digits. That is not a criticism of the print. It is the number to put in the model.

Where the Growth Actually Came From

A 15.9% profit quarter for a bank this size deserves a bridge rather than an adjective. Working from the ₹12,768 crore of year-ago profit: core operating profit added ₹2,730 crore; the treasury line removed ₹1,090 crore; lower provisions added ₹554 crore; and the tax line, including the write-back, took ₹158 crore. Net ₹2,036 crore, which is the reported increase. The point of laying it out that way is that the largest single contributor is the operating line by a factor of five, which was not true in either of the two prior quarters we have examined.

Segment Performance

ICICI reports four standalone segments. Inter-segment revenue is large relative to gross segment revenue because Treasury acts as the internal funding counterparty, so segment profit before tax is the more informative line. The total revenue row below is gross segment revenue before the ₹52,091 crore inter-segment elimination, not the ₹54,247 crore income from operations.

Retail Banking

Retail segment revenue grew 4.2% year-on-year to ₹42,166 crore, an improvement on the flat line the segment printed in March but still well behind the 12.0% growth in the retail loan book. Segment profit before tax grew 31.8% to ₹6,239 crore, and the entire delta is again credit rather than revenue: gross non-performing additions from the retail and rural books fell to ₹4,331 crore from ₹5,193 crore, and net additions to ₹2,121 crore from ₹2,668 crore, a 20.5% reduction. Sequentially, segment profit fell 9.9%, which is the mirror of the March quarter's near-zero provision charge unwinding.

Within retail products the composition improved on the margin. Mortgages grew 14.6% year-on-year against 13.2% in March. Personal loans grew 12.9% against 7.2%, the first real acceleration in the unsecured book since the regulatory tightening of two years ago. Commercial vehicles and equipment grew 12.8%. Auto loans grew 3.6% against 1.7%. Credit cards declined 1.9% year-on-year and 1.7% sequentially, a third consecutive sequential contraction, though the rate of decline has more than halved from the 5.6% year-on-year fall reported in March.

"PL has picked up quite a bit, actually, 12% year-on-year and 4% sequential growth. I think it's after a long time that we are seeing those kinds of numbers. Cards, I think the revolver rates are lower, so that is impacting the book growth." — Anindya Banerjee, Group Chief Financial Officer

Assessment: Retail is still a credit-normalisation story rather than a revenue story, and a segment growing revenue 4.2% on a book growing 12.0% is still repricing down faster than it is adding volume. But two things improved that were deteriorating in March: personal loans turned decisively, and the card contraction slowed. The segment's profit share fell to 32.6% from 35.1% for FY26 as a whole, which continues the migration of earnings weight toward wholesale that we flagged last quarter.

Wholesale Banking

Wholesale banking produced ₹7,479 crore of profit before tax, up 38.8% year-on-year and 6.3% sequentially, on segment revenue up 18.1%. It is now the largest profit segment by a clear margin at 39.1% of the total, and it out-earned retail for a second consecutive quarter. Domestic corporate advances grew 18.5% year-on-year and 6.9% sequentially, against 9.3% and 3.1% in March, and business banking grew 28.2% and 6.9%. Exposure to non-bank finance and housing finance companies rose to ₹92,052 crore from ₹85,904 crore at March, which the Bank sizes at about 4.5% of advances. The builder portfolio rose to ₹74,721 crore from ₹71,421 crore, about 4.3% of the loan portfolio, with 0.7% of it internally rated BB and below or non-performing, an improvement from 0.9% at March.

"The corporates were always doing what they were doing. They are doing CapEx to the extent that they were doing CapEx. I think certainly bond markets this quarter have not favored the corporate sector. If you see, we have a reasonable increase in our NBFC portfolio as well. Part of it is a shift from bond markets. Part of it is, I would say working capital utilization." — Anindya Banerjee, Group Chief Financial Officer

Assessment: The doubling of corporate growth is the biggest single change in the print and management was careful not to sell it as an investment cycle. Substitution lending is lower-duration, lower-spread and more reversible than project finance, so a corporate book growing 18.5% on that basis should not be capitalised at the same multiple as one growing on capital expenditure. Set against that, the credit metrics inside the segment are moving the wrong way: net non-performing additions from corporate and business banking rose to ₹586 crore from ₹366 crore a year ago and ₹322 crore in March, a third consecutive increase, in the segment now carrying 39.1% of profit.

Treasury

Treasury profit before tax fell 23.0% year-on-year to ₹4,822 crore, though it rose 28.9% sequentially off the depressed March base. Within the profit and loss account the treasury income line swung back to a gain of ₹151 crore from a loss of ₹106 crore in March, ending the two-quarter run of losses that followed the central bank's foreign-exchange position limits. The year-ago comparison is the difficulty: Q1 FY26 carried a ₹1,241 crore treasury gain, so the ₹1,090 crore swing is the single largest drag on the year-on-year operating profit comparison. Management offered no commentary on the treasury line this quarter and no analyst asked.

Assessment: The trading line has stopped bleeding, which resolves the acute question from April. What it has not done is recover: at 25.2% of segment profit before tax against 26.0% for FY26, treasury remains a quarter of this bank's earnings and it is running well below the year-ago level. The silence on both sides of the call is itself informative. A line that generated three consecutive quarters of headlines went unmentioned within twelve weeks of the central bank partially rolling back the measures that caused them.

Other Banking

The smallest segment produced ₹585 crore of profit before tax on revenue of ₹1,633 crore, up 6.8% year-on-year and 47.5% sequentially. At 3.1% of quarterly segment profit it remains immaterial to the consolidated picture, though it has now held above 3% for a quarter after averaging 1.9% across FY26.

Assessment: Still too small to matter and still growing faster than the bank. The threshold for it becoming an analytical item rather than a rounding one is 5% of profit before tax, and it is not there.

Loan Book Composition

The composition question we asked in April has a different answer now. In March, retail grew 9.5% and domestic corporate 9.3%, so essentially the entire gap to the 15.8% headline was rural and business banking. This quarter every book accelerated, and the corporate line accelerated most. Retail contributes 12.0%, corporate 18.5%, business banking 28.2% and rural 35.4%. Growth is still concentrated in the two least-seasoned portfolios, but it is no longer only there.

Key KPIs

Subsidiaries

Consolidated profit after tax of ₹15,440 crore against standalone ₹14,805 crore implies ₹636 crore from the subsidiary stack, 4.1% of the consolidated figure. That is down from ₹789 crore and 5.8% a year ago and from ₹1,053 crore and 7.1% in the March quarter. Consolidated assets grew 12.5% to ₹30,02,407 crore. Consolidated basic EPS was ₹21.54 against ₹19.02.

Assessment: The general insurer is the story and it is a bad one. Profit fell 46.1% and the combined ratio deteriorated 430 basis points to 107.2%, which the Bank attributes to a reserve increase following a judicial pronouncement it does not name or size. That single entity is most of the reason the subsidiary stack's contribution to consolidated profit fell to 4.1% from 5.8%. Against it, the asset manager grew 23.1% and the life insurer 27.8% with new-business margin up two points to 26.7%. The composition matters for anyone running a sum-of-the-parts: the two subsidiaries with public marks are compounding well, and the underwriting business is absorbing a legal shock of undisclosed size.

Key Topics & Management Commentary

Overall Management Tone: Assured on the numbers and, for the first time in three quarters, willing to be specific about them without prompting: the clean margin series, the Kisan Credit Card seasonality, the base effect inside the fee line and the tax write-back all arrived in prepared remarks rather than under questioning. The forward posture is unchanged, which is to say every quantified request was declined, but the refusals were narrower than in March and were paired with mechanism rather than deflection. The one place the answers thinned was the foreign-currency deposit programme, where management is genuinely early rather than evasive, and said so.

1. Positive Jaws, Finally

Operating expenses grew 10.4% year-on-year against core revenue growth of 13.5%, and core cost-to-income improved to 38.3% from 39.4%. For a franchise that spent all of FY26 growing costs 11.5% against core operating profit of 7.7%, this is the single most consequential line in the print. Core operating profit grew 15.6% against 5.1% in the March quarter.

The cost line itself decomposes usefully. Employee expenses grew only 5.5%, and management flagged that Q1 is the quarter that carries annual increments and promotions, so the low growth rate is not a deferral. Non-employee expenses grew 13.8%, which is where branch build and technology sit: 97 branches added in the quarter to 7,608, and technology at 11.4% of operating expenses against roughly 11% for FY26.

"Employee expenses increased by 5.5% year-on-year in this quarter, reflecting mainly the impact of annual increments and promotions that take place during the first quarter of every fiscal year." — Anindya Banerjee, Group Chief Financial Officer

Assessment: In April we named a quarter of positive jaws on the core revenue comparison as one of two conditions that would move our rating. This is that quarter. Two qualifications keep it from being decisive on its own. Employee cost growth of 5.5% is unusually low for the quarter that carries the annual increment cycle, and it laps a year-ago employee line of ₹4,743 crore that was the heaviest of FY26. Positive jaws sustained for two more quarters is a franchise property; one quarter against a soft comparison is a data point.

2. Loan Growth Accelerates to 19.6%, and the Corporate Book Doubles Its Rate

Advances grew 19.6% year-on-year and 5.0% sequentially, up from 15.8% and 6.0% in the March quarter. Domestic advances grew 18.8%. The distribution changed materially: domestic corporate accelerated to 18.5% from 9.3%, business banking to 28.2% from 24.4%, rural including gold loans to 35.4% from 25.6%, and retail to 12.0% from 9.5%. Every portfolio bar credit cards grew faster than it did in March, in what is seasonally the weakest quarter of the year.

"I think not just loan growth, but a range of other high-frequency indicators are showing positive momentum. We are sort of participating in that, and we will keep looking at opportunities as they come." — Anindya Banerjee, Group Chief Financial Officer

Asked for the average rather than period-end growth rate, which is the figure that actually drives net interest income, management confirmed it is lower and would not give it.

"It will be somewhat lower, as I mentioned earlier, this higher loan growth this quarter is reflective of the increase in loan growth in the system, both the banks and the systems is off the base of Q1 last year." — Anindya Banerjee, Group Chief Financial Officer

Assessment: Nineteen-point-six percent is a system-plus number achieved in a seasonally weak quarter, and management's own framing is that a meaningful part of it is system recovery lapping a soft base rather than share gain. The unqualified positive is that growth is no longer confined to the two least-seasoned books. The qualification is that the corporate acceleration behind it is substitution, addressed next.

3. Corporate Growth Is Bond-Market Substitution, Not Capital Expenditure

The domestic corporate book grew 18.5% year-on-year and 6.9% sequentially, having grown 9.3% and 3.1% only one quarter earlier. Exposure to non-bank finance and housing finance companies rose ₹6,148 crore in the quarter to ₹92,052 crore. Asked directly whether this reflects fresh capital expenditure, management gave the most useful answer of the call, and it was not the flattering one.

"Part of it is a shift from bond markets. Part of it is, I would say working capital utilization. Part of it may be corporates taking on some borrowings and just maintaining some liquidity buffers as they also want to have greater resilience in the uncertainties of the environment." — Anindya Banerjee, Group Chief Financial Officer

Pressed a second time on whether the loan pipeline shows more fresh capital expenditure than before, management declined to generalise and pointed instead to pricing: "this quarter in particular, we saw a healthy pipeline at reasonable rates."

Assessment: This matters for how the growth should be capitalised. Bond-market substitution is shorter-duration, thinner-spread and reverses the moment credit spreads normalise; working-capital utilisation and precautionary liquidity buffers reverse faster still. None of it is bad lending, and the Bank is being paid to intermediate. But an investor extrapolating 18.5% corporate growth into FY28 is extrapolating a funding-market dislocation. Management declined to sell it as anything else, which is to their credit and should be taken at face value.

4. Margin: the First Clean Reading

Reported net interest margin was 4.36%, four basis points above March and two above the year-ago quarter. What matters is that management disclosed the adjusted series in prepared remarks rather than waiting to be asked, as happened in the two prior quarters.

"Excluding the benefit of interest on tax refund, the net interest margin would have been 4.28% in this quarter, 4.27% in the previous quarter, and 4.27% in Q1 of last year." — Anindya Banerjee, Group Chief Financial Officer

Cost of deposits fell to 4.41% from 4.43% sequentially and 4.85% a year ago, a 44 basis point annual improvement. The domestic loan mix moved marginally toward external benchmarks, at 57% repo-linked and other external benchmarks, 13% on the marginal cost of funds and older benchmarks, and 30% fixed rate, against 56, 13 and 31 at March. Asked how the yield held despite agricultural interest reversals, management credited funding and the government securities book.

"Overall the NIM reflects the healthy funding franchise, our disciplined approach consistently on both deposit and loan pricing, and as well as our management of the government securities book." — Anindya Banerjee, Group Chief Financial Officer

Assessment: Our April read was that the reported margin was being held flat by a liability tailwind with a defined end, and that core margin would drift down. The core margin did not drift down; it went up one basis point, and it has now been flat at roughly 4.27% for four quarters through a transmitted policy cut with 57% of the book on external benchmarks. That is better than we modelled and it is the reason this pillar moves from at-risk to on-track. The new caveat is self-inflicted and addressed next.

5. FCNR (B): a New Funding Programme With a Two-Sided Effect

The largest genuinely new item on the call is the foreign-currency non-resident deposit programme, which drew questions from two directions and has no precedent in our coverage of this name. Management would not size it.

"We think it is a good scheme, and we will look to mobilize as much as we can. It is very early days yet." — Anindya Banerjee, Group Chief Financial Officer

On economics, the disclosure was specific: a roughly 6% coupon, and an all-in cost after hedging of about 630 to 640 basis points, which management characterised as competitive against wholesale lending rates. The programme is expected to play out over eight to ten weeks. Separately, the Board approved raising the limit for overseas bond, note and offshore certificate-of-deposit issuance to USD 2.50 billion, and the overseas loan book has already grown to 3.1% of total advances from 2.7%.

"From a cost-effectiveness perspective, 6%, and then if you look at kind of the all-in cost after hedging the coupon, et cetera, it will be somewhere maybe 630, 640, which is of course lower than the wholesale lending rates." — Anindya Banerjee, Group Chief Financial Officer

The consequence for the margin was volunteered rather than extracted, and it is the first forward qualification management has attached to the range-bound frame since we picked up coverage.

"Based on current conditions and assuming no real policy rate movements, I would say it should be range bound. We'll also have to, as we go along, factor in sort of the impact of the FCNR deposit mobilization and the related leverage, where there could be some impact on margins." — Anindya Banerjee, Group Chief Financial Officer

Assessment: A hedged 6.35% funding cost deployed at wholesale lending rates is accretive to profit and dilutive to the reported margin, because the spread is thinner than the 4.28% domestic net interest margin even though the return on the incremental equity is fine. That is exactly the trade a bank with a 16.19% common equity ratio should be making. The risk is presentational: the market watches the margin line, and a bank that has built its premium on the highest margin in the Indian large-bank complex is about to dilute it deliberately. Expect the explanation to be required more than once. Size is the unknown, and management has given no number and no target, which makes the second-quarter print the first place the programme becomes measurable.

6. The Provision Line Normalises, and the Recovery Nobody Sized

Provisions other than tax rose to ₹1,260 crore from ₹96 crore in March, or 32 basis points of average advances annualised against three basis points. That is the normalisation the Street expected and the reason the sequential profit comparison is a modest 8.0% rather than a headline number. Underneath it, credit is genuinely better: gross non-performing additions fell 11.1% year-on-year to ₹5,552 crore, including ₹706 crore from the Kisan Credit Card portfolio that management pre-emptively flagged as a first-and-third-quarter seasonal item, and retail and rural net additions fell 20.5%.

Asked whether a one-off sat inside the number, management named the mechanism, again without sizing it.

"I think you would all be aware that in one particular case, there was an NCLT judgment where the company was taken over. That recovery did come through this quarter. This is an asset that had been previously sold to NARCL, and that is there." — Anindya Banerjee, Group Chief Financial Officer

"From a credit cost perspective, the reported of 32 basis points, I guess as we always say, in the more normalized level, adjusting for chunky recoveries would be around 50 basis points, and that is where it stays." — Anindya Banerjee, Group Chief Financial Officer

Assessment: This is a materially better disclosure than March. Management volunteered both the existence of the one-off and the normalised rate implied after adjusting for it, which brackets the undisclosed amount at roughly ₹730 crore pre-tax for anyone willing to do the arithmetic. Two consecutive quarters in which the reported credit cost is below the Bank's own normalised frame means the frame, not the print, is the modelling input. At 50 basis points on the current book, the quarterly charge is close to ₹2,000 crore.

7. Fee Income Snaps Back 23.5%, With an Asterisk Management Supplied

Fee income grew 23.5% year-on-year to ₹7,286 crore, against 7.5% in the March quarter and 7.8% for FY26 as a whole. This is the line that explains most of the improvement in core revenue growth, and management named the caveat before anyone asked.

"As I briefly mentioned in our opening remarks, there is some amount of a base effect. Last year, the fee income growth in Q1 was in fact sequentially lower over the preceding quarter." — Anindya Banerjee, Group Chief Financial Officer

Pressed on which fee pools drove it, management declined the disbursement-linked framing an analyst offered and listed the components instead: cards, loan processing fees, transaction banking and trade, foreign exchange and derivatives, and deposit-linked fees. The sequential increase was put at about ₹500 crore, which reconciles with the filed ₹7,286 crore against ₹6,779 crore. Retail, rural and business banking customers accounted for about 72% of fees, down from 78% in March, which is the corporate mix shift showing up in the fee line as well as the loan book.

"As you said, maybe cards is one area where fee growth is less than what we would want it to be. Although, at a PBT level, the business has done very well because of the reduction in credit costs." — Anindya Banerjee, Group Chief Financial Officer

Assessment: A 23.5% print against a base management itself describes as weak is worth perhaps 15% as a run rate, which is still the first quarter in two years in which fee income grew faster than the loan book. The cards problem we identified in April is unresolved on the fee line and improving on the profit line, which is a distinction worth holding: lower revolver share compresses both interest and fee income while the credit cost of the same book falls. That trade is net positive for profit and net negative for revenue quality.

8. The Loan-to-Deposit Ratio Nobody Mentioned

Advances grew 5.0% sequentially and deposits 2.2%, taking the loan-to-deposit ratio to 89.0% from 86.6% at March and 84.8% a year ago. That is 416 basis points of annual increase, on top of 327 basis points the year before. Average liquidity coverage was about 124% against about 126% in March. Average deposits grew 14.0% year-on-year and 6.1% sequentially, so the period-end and average growth rates have converged, which removes the ambiguity that made the March comparison hard to read.

Neither management nor any questioner raised the ratio on this call, in contrast to March when the deposit-versus-loan gap drew the most persistent questioning of the session.

Assessment: Deposit growth of 14.0% against loan growth of 19.6% is a five-and-a-half point gap, wider than the four-point gap that dominated the March call, and it went unremarked because deposit growth accelerated in absolute terms. Both things are true. The funding position remains comfortable on every disclosed measure, and the trend line has now added 700 basis points to the ratio in two years. The FCNR programme is the answer management is building, and the second quarter is where we find out whether it closes the gap or funds a wider one.

9. Asset Quality: Better Everywhere Except the Segment That Now Earns Most

Gross non-performing assets fell to 1.38% of gross customer assets from 1.67% a year ago, and the net ratio to 0.35% from 0.41%. Beyond specific provisions the Bank holds ₹22,963 crore, 1.4% of loans, including the untouched ₹13,100 crore contingency buffer now worth 0.8% of advances rather than 0.9% because the book grew. Performing corporate exposure rated BB and below fell slightly to ₹3,485 crore. Standard borrowers under resolution fell to ₹1,363 crore. Provisioning coverage fell to 74.7% from 75.8%, and the net non-performing ratio rose two basis points sequentially.

One line continues in the wrong direction. Net additions to gross non-performing assets from the corporate and business banking books were ₹586 crore against ₹366 crore a year ago and ₹322 crore in March, because gross additions rose to ₹1,221 crore from ₹1,052 crore while recoveries fell to ₹635 crore from ₹686 crore. Neither management nor any questioner addressed it.

Assessment: Flagging this line in April as a watch item off a near-zero base was the right call and it has now risen for a third consecutive quarter, in the segment that has just become the Bank's largest profit contributor at 39.1% and whose fastest-growing book compounds at 28.2%. The absolute numbers remain trivial against a ₹16.3 trillion book and a ₹13,100 crore buffer. The direction is the point, and the reason it belongs in the thesis rather than the footnotes is that it is the only credit series in this print moving against the trend.

10. The Agricultural Priority-Sector Provision, One Quarter Later

In the December quarter the Reserve Bank directed the Bank to take a ₹1,283 crore standard-asset provision against agricultural priority-sector facilities whose terms were found not fully compliant with the classification requirements. In April management said an update would come "a quarter-or-so from now." The provision is still held in full at June 30, and the update, when requested, was procedural.

"I wouldn't really be able to comment on the timing, but yes, of course, we have been working on the required remediation of the portfolio, and we just want to make sure that whatever we have done is correct and signed off and validated before we discuss a writeback of the provision." — Anindya Banerjee, Group Chief Financial Officer

Pressed on whether the validation process was under way, management confirmed it is, described the portfolio as granular, and again refused a timeline. The associated interest reversals on the Kisan Credit Card book also flowed through the yield line this quarter without a separate disclosure.

Assessment: This is the one explicit forward commitment we carried into the quarter and it was met in the weakest available sense: an update was given, no release occurred, and the timeline was withdrawn rather than extended. The provision is worth about 1.9% of FY26 profit before tax if fully released. It is now better treated as unquantified optionality with no date than as a second-half catalyst, and we do not carry it in our estimates.

11. The ₹446 Crore Tax Write-Back

The effective tax rate fell to 22.59% from 24.59% a year ago, which is worth about ₹446 crore of profit, or 3.0% of the reported figure and roughly three and a half points of the 15.9% growth rate. Management disclosed it in a single sentence in prepared remarks.

"The bank has written back tax provision of INR 4.46 billion pursuant to favorable tax orders." — Anindya Banerjee, Group Chief Financial Officer

No questioner returned to it. Excluding the write-back, profit after tax would have been roughly ₹14,358 crore, up 12.5% rather than 15.9%.

Assessment: The disclosure is clean and the item is plainly non-recurring, so the criticism is not of the Bank. It is that a quarter carrying two identified below-the-operating-line assists, worth roughly eight points of growth between them, was received by the market and by the sell-side as a straightforward sixteen percent print. The operating result is strong enough that it did not need either. Model the effective tax rate back at 24% and the credit cost at 50 basis points, and the underlying growth rate is high single digits.

12. Capital Starts Being Consumed, and the Transition Nobody Can Size

The common equity tier-one ratio fell to 16.19% from 16.35% and total capital adequacy to 16.84% from 17.18%, against minimums of 8.20% and 11.70%. That is the arithmetic of growing risk-weighted assets 5.0% in a quarter, and it is the first sequential decline in either ratio in our coverage. The FY26 dividend of ₹12 per share, roughly ₹8,610 crore, has not yet been paid and awaits the August 21 annual meeting, so the ratios take a further step down in the second quarter.

Looking further out, the transition to expected credit loss provisioning arrives in FY28. Management gave the shape and refused the magnitude, including when offered a range as wide as five to ten basis points.

"On the net worth, we will not really have any impact in our assessment based on whatever pro forma estimates we've done based on the existing position of the balance sheet of the portfolio, is that whatever impact is there will be well absorbed by our provisioning buffers." — Anindya Banerjee, Group Chief Financial Officer

"On an ongoing basis, the fact that we will be providing for Stage 2, obviously that is a provision that is currently not made by any bank. That will result in higher provisions, which will be partly offset by lower provisions on the Stage 3" — Anindya Banerjee, Group Chief Financial Officer

Assessment: In April we argued that a common equity ratio 815 basis points above the minimum, on a book earning 16.5%, was a lot of unemployed capital and that the return could only fall by arithmetic unless growth or payout rose. Growth rose, capital started being consumed, and annualised return on equity went to 17.7%. That is the constructive resolution of the question we asked. The expected credit loss transition is the offsetting item and it is genuinely unknowable from outside: a permanent step-up in the ongoing charge from Stage 2 provisioning, sized by nobody in the sector, arriving in the same year the currently suppressed credit cost has to normalise anyway.

Guidance & Outlook

ICICI does not publish numeric guidance and did not start here. What exists is a set of qualitative frames, most of them restated and two of them new.

The strategic frame from the top of the call has not moved in several years and was restated verbatim in substance for a second consecutive quarter.

"At ICICI Bank, our strategic focus continues to be on growing profit before tax, excluding treasury, through the 360-degree customer-centric approach and by serving opportunities across ecosystems and micro markets." — Sandeep Bakhshi, Managing Director and Chief Executive Officer

The closing frame dropped one word that was present in March. In April the Bank said it saw "many profit opportunities"; this quarter the same sentence reads "many opportunities." That is a single word in a boilerplate paragraph and it would be over-reading to treat it as guidance, but the sentence is otherwise identical and the deletion is on the profitability qualifier in the quarter the Bank started building a lower-margin funding book.

"Looking ahead, we see many opportunities to drive risk-calibrated profitable growth and grow market share across key segments." — Sandeep Bakhshi, Managing Director and Chief Executive Officer

On competition, management declined the invitation to say pricing has improved as public-sector banks exhaust their liquidity buffers, and reframed the question around its own franchise.

"I think it's a large market and there is enough for us to do. We are, for example, some of the rates that get quoted, we are not doing that kind of business, but there is enough other business where we are comfortable with risk-adjusted returns on a customer basis, which we do." — Anindya Banerjee, Group Chief Financial Officer

Implied FY27 arithmetic: hold the core margin at 4.28% and grow average advances 17%, and net interest income lands near ₹1,02,000 crore, up about 16%. Grow operating expenses at 10%, the rate Q1 actually printed, and the cost base reaches roughly ₹52,000 crore. Normalise credit cost to 45 basis points on an average book near ₹16.9 trillion and provisions become roughly ₹7,600 crore against ₹5,380 crore in FY26. Put the effective tax rate back at 24%. On those inputs profit after tax grows low double digits for the year, which is a step up from the high-single-digit path we modelled in April but well below the 15.9% this quarter printed.

Street at: the sell-side raised FY27 and FY28 earnings estimates by roughly 2% to 5% after the print and models average return on assets near 2.3% and return on equity near 17% across FY27 and FY28. The post-print target cluster on the Mumbai line ran ₹1,710 to ₹1,800, implying 17% to 23% upside from the reaction-day close, against a ₹1,570 to ₹1,785 cluster after the March print.

Guidance style: unchanged and still deliberately unfalsifiable, with one improvement. The Bank continues to publish no number it can be graded against, but it volunteered three adjustments this quarter that make its own results harder to over-read: the clean margin series, the fee base effect, and the normalised credit cost that brackets an undisclosed recovery. A management team that supplies the deductions from its own headline is a different proposition from one that waits to be asked, and it is the second consecutive quarter in which the disclosure has improved.

Analyst Q&A Highlights

Whether High-Teens Loan Growth Can Hold Through the Year

The call opened on the number that defines the quarter, and the framing was pointed: growth of this rate achieved in a seasonally weak quarter invites the question of whether it is a floor. Management neither confirmed nor denied a high-teens run rate, and instead relocated the credit for it from the franchise to the system, describing fiscal and monetary policy measures working through and a broad set of high-frequency indicators turning up. The answer contained no target and one clear signal, which is that the Bank does not regard this quarter's rate as something it produced on its own.

Q: "Given that you already achieved such good loan growth in a seasonally weak quarter, do you find this sustaining? As in, would you be able to do, say, a high teen loan growth through the year? Is there enough visible demand across your segments?"
— Mahrukh Adajania, Tara Capital

A: "I think not just loan growth, but a range of other high-frequency indicators are showing positive momentum. We are sort of participating in that, and we will keep looking at opportunities as they come."
— Anindya Banerjee, Group Chief Financial Officer

Assessment: "Participating" is the operative word and it is not the language of share gain. Set against a later exchange in which management conceded average loan growth is "somewhat lower" than the 19.6% period-end figure and attributed part of the print to a soft prior-year base, the honest read is that the run rate is mid-to-high teens rather than nineteen. That is still the best growth in the large-bank complex.

The Economics of the New Foreign-Currency Deposit Programme

The single most useful new disclosure of the call came from the opening question, which asked three things at once: whether there is a mobilisation target, how much has been raised, and how the cost compares to domestic term deposits. Management answered the third precisely and declined the first two on the grounds of genuine earliness rather than policy. The programme is a matter of weeks old and management put the window over which it will become visible at eight to ten weeks, which places the first measurable data point at the second-quarter print.

Q: "Then on FCNR, if you have a target in mind and how much you have already mobilized and what is the cost effectiveness of FCNR relative to your domestic term deposits."
— Mahrukh Adajania, Tara Capital

A: "From a cost-effectiveness perspective, 6%, and then if you look at kind of the all-in cost after hedging the coupon, et cetera, it will be somewhere maybe 630, 640, which is of course lower than the wholesale lending rates."
— Anindya Banerjee, Group Chief Financial Officer

Assessment: An all-in hedged cost of roughly 635 basis points against a 441 basis point cost of deposits tells you this funding is expensive relative to the domestic book and cheap relative to what it will be lent at. It is accretive to profit and dilutive to the margin, which is a trade a bank with sixteen percent common equity should take and a trade that will be misread on the line the market watches. The absence of any size makes the programme unmodellable for now.

When the Agricultural Priority-Sector Provision Gets Released

The one explicit forward commitment carried into this quarter was the promised update on the ₹1,283 crore standard-asset provision the regulator directed in December. The questioner asked precisely the right thing, offering management two timelines to choose between rather than an open-ended request. Management took neither, described the remediation and validation process as under way, and pushed the horizon out to "the next few months" while explicitly withdrawing the ability to give a date. A follow-up confirming the process had actually started drew a firmer yes.

Q: "How is the approach and when do we see the recoveries coming through or maybe the reversals of the provision? Is it expected to come through in the second quarter, or maybe it might take some time, and we should see it more towards the end of the fiscal year?"
— Kunal Shah, Citigroup

A: "I wouldn't really be able to comment on the timing, but yes, of course, we have been working on the required remediation of the portfolio, and we just want to make sure that whatever we have done is correct and signed off and validated before we discuss a writeback of the provision."
— Anindya Banerjee, Group Chief Financial Officer

Assessment: A commitment made with a stated horizon was met with an update that removed the horizon. That is a mild negative on the commitment itself and close to neutral on the economics, because the amount is small and the release, if it comes, is a one-time credit to a line the Bank is already over-providing. Stop carrying it as a catalyst.

Whether the Margin Frame Survives Accelerating Growth

With the loan book compounding nineteen percent and the reported margin up sequentially, a questioner asked whether the range-bound guidance still stands. This produced the only forward qualification management has attached to that frame in three quarters, and it was volunteered rather than extracted: the foreign-currency deposit programme and the leverage built on it will expand the international balance sheet and may compress the reported margin. Everything before that caveat was a restatement.

Q: "You've managed it extremely well in this cycle, now with loan growth accelerating meaningfully for you, would you continue to maintain your guidance on range-bound margins even going ahead?"
— Rikin Shah, IIFL Capital

A: "Based on current conditions and assuming no real policy rate movements, I would say it should be range bound. We'll also have to, as we go along, factor in sort of the impact of the FCNR deposit mobilization and the related leverage, where there could be some impact on margins."
— Anindya Banerjee, Group Chief Financial Officer

Assessment: The frame survives with an escape hatch attached, and the escape hatch is the right kind. A margin that falls because the Bank is deploying cheap hedged offshore funding into wholesale assets is a mix effect, not a pricing failure. The obligation now sits with the reader to distinguish the two when the reported number drops, because the Bank will report one line and the drivers will be two.

Whether a One-Off Recovery Sat Inside the Provision Line

The same question that dominated the March call returned in one sentence and got a materially better answer. Where the prior quarter produced a mechanism without a magnitude, this one produced a named event, its origin, and the normalised credit cost the Bank believes applies after adjusting for it. That last figure is the disclosure that matters: it brackets the size of the undisclosed recovery for anyone willing to do the arithmetic against the reported thirty-two basis points.

Q: "The fourth one is on recoveries. Was there any one-off recovery in this quarter from any accounts?"
— Rikin Shah, IIFL Capital

A: "From a credit cost perspective, the reported of 32 basis points, I guess as we always say, in the more normalized level, adjusting for chunky recoveries would be around 50 basis points, and that is where it stays."
— Anindya Banerjee, Group Chief Financial Officer

Assessment: Two quarters ago the equivalent exchange ended with the questioner no better informed. Here the answer supplies the adjustment factor without supplying the number, which is a real improvement in disclosure quality and functionally equivalent for modelling purposes. The gap between thirty-two and fifty basis points on the current book is roughly ₹730 crore pre-tax per quarter.

What the Expected Credit Loss Transition Costs

The most consequential unanswered question on the call concerns a change that has not happened yet. Indian banks move to expected credit loss provisioning from next year, and the questioner asked for both the one-time net-worth effect and the recurring credit-cost effect. Management gave a confident answer on the first and a structural answer on the second, then declined a follow-up that offered a range as generous as five to ten basis points, on the grounds that the estimate requires running the portfolio through the new guidelines.

Q: "If you could just talk about the impact on transition from next year, both on one-time basis on net worth and also on the recurring credit cost, please."
— Rikin Shah, IIFL Capital

A: "On an ongoing basis, the fact that we will be providing for Stage 2, obviously that is a provision that is currently not made by any bank. That will result in higher provisions, which will be partly offset by lower provisions on the Stage 3, where currently we follow this percentage-based approach, there it will be much more on a predicted loss kind of approach."
— Anindya Banerjee, Group Chief Financial Officer

Assessment: The net-worth answer is credible precisely because of the buffers this bank has been criticised for hoarding: a ₹13,100 crore contingency provision exists to absorb exactly this. The ongoing answer is the open item, and the refusal to bound it is defensible because nobody in the sector has bounded it. What it means practically is that FY28 carries an unquantified permanent step-up in the charge, in the same year the currently suppressed credit cost has to normalise.

Whether Corporate Growth Is Capital Expenditure or Substitution

A domestic corporate book that doubled its growth rate in one quarter invites the most important question about the Indian cycle, and the questioner asked it cleanly: how much of this is investment demand and how much is credit substitution or short-term working capital. Management gave a three-part answer that put substitution first, and then declined a follow-up inviting the conclusion that fresh project proposals had picked up, saying it did not want to make a comment that could be extrapolated to the system.

Q: "If you could just kind of give us some colors to how much of it is a demanded growth that one is seeing because of CapEx versus probably a credit substitution or a short-term working capital demand that has come through."
— M.B. Mahesh, Kotak Securities

A: "Part of it is a shift from bond markets. Part of it is, I would say working capital utilization. Part of it may be corporates taking on some borrowings and just maintaining some liquidity buffers as they also want to have greater resilience in the uncertainties of the environment."
— Anindya Banerjee, Group Chief Financial Officer

Assessment: Three of the three named drivers are cyclical or defensive, and none is capital formation. Management could have let the number speak and did not, which is the same candour it showed on the fee base effect. For the model this is the most important qualification in the print: the fastest-accelerating book in the bank is growing on a funding-market dislocation and on corporate precaution, both of which reverse.

Growing Business Banking Into an Energy Shock

The sharpest challenge of the call did not dispute the growth, it disputed the timing: accelerating the least-seasoned portfolio into an energy shock and an adverse monsoon signal, while coasting on the previous year's macro tailwind. Management's answer was that the shock is being handled inside customer selection and onboarding rather than through portfolio-level restraint, and rested the case on the portfolio being granular, secured and currently performing. It is the same position as March, applied to a book growing four points faster.

Q: "We have seen some very strong loan growth, particularly business banking. I'm just wondering, is this the right time to accelerate, given we are going through the energy shock and El Niño is upon us?"
— Chintan, Autonomous Research

A: "It is a granular portfolio, reasonably secured portfolio, and we are quite comfortable growing it."
— Anindya Banerjee, Group Chief Financial Officer

Assessment: The defence is coherent and it is the same defence as last quarter, offered while the book accelerated from 24.4% to 28.2% and while net non-performing additions in the segment containing it rose for a third consecutive quarter. Granular and secured is a real mitigant and it is not a substitute for a cycle. The Bank is choosing to underwrite through the shock rather than pre-position against it, which is a position, clearly held, and the entire macro risk sits in front of the investor rather than behind a provision.

What They're NOT Saying

1. The size of the insolvency-court recovery: the event was named and its origin disclosed, but the amount inside the provision line was not. The normalised-credit-cost answer brackets it at roughly ₹730 crore pre-tax, which is a better outcome than March produced and still leaves the exact figure undisclosed for a second consecutive quarter.
2. Any size for the foreign-currency deposit programme: no target, no amount mobilised to date, no expected steady-state balance. The largest new item in the print is entirely unmodellable, and management framed the refusal as earliness rather than policy.
3. The expected credit loss impact on the ongoing charge: offered a range as wide as five to ten basis points, management declined to confirm or reject even that, saying the estimate requires running the portfolio through the new guidelines.
4. Average loan growth: confirmed to be "somewhat lower" than the 19.6% period-end figure and never quantified, in a quarter where the gap between the two is the difference between a mid-teens and a high-teens net interest income trajectory.
5. Any FY27 loan growth number: asked directly whether high-teens growth can hold through the year, management answered about system momentum and gave no figure, as in each of the prior two calls.
6. A release date for the agricultural priority-sector provision: the timeline given in April was replaced with "the next few months" and an explicit refusal to commit to one.
7. The corporate and business banking net non-performing addition: ₹586 crore against ₹366 crore a year ago and ₹322 crore in March, disclosed inside a sequence of numbers in prepared remarks, never returned to by management and never raised by a questioner, in the segment that now earns 39.1% of profit.
8. The loan-to-deposit ratio: at 89.0% from 84.8% a year ago, it was neither mentioned by management nor asked about, in contrast to March when the deposit-versus-loan gap was the most persistent line of questioning on the call.
9. What the judicial pronouncement affecting the general insurer was, or what it cost: ICICI Lombard's profit fell 46.1% and its combined ratio deteriorated 430 basis points to 107.2%, attributed to a reserve increase following a court ruling that is neither named nor sized.
10. Any commentary on the treasury line: after three quarters in which the trading book was the most-discussed item on the call, a ₹1,090 crore year-on-year swing that is the largest single drag on the operating profit comparison went unmentioned by management and unasked by every questioner.
11. Period-end CASA: as in March, only the average ratio is given, at 38.1% against 38.6% in the prior quarter. The period-end figure is not disclosed.
12. The cost of the FCNR-related overseas borrowing authorisation: the Board raised the offshore issuance limit to USD 2.50 billion in the same meeting, and neither the current utilisation nor the expected cost of that funding was discussed on the call.

Market Reaction

• Pre-print setup, New York ADS: IBN closed at $29.76 on Friday, July 17, the last session before the Saturday board meeting. That was 0.1% below the December 31 close of $29.80, against an 8.9% year-to-date gain for the S&P 500, and 10.4% below the $33.23 close twelve months earlier. The trailing thirty days had been strong, up 6.7% from $27.89 on June 17. The fifty-two-week closing range entering the print was $25.20 to $34.41.
• Pre-print setup, Mumbai line: ICICIBANK closed at ₹1,444.30 on July 17, up 7.6% year-to-date against a 6.9% decline for the Nifty 50, and up 1.8% over twelve months from ₹1,418.70. Trailing thirty days were up 8.0% from ₹1,336.80. The Friday session itself added 1.8% from a ₹1,418.20 close, so the stock went into the print on a strong bid. The fifty-two-week closing range was ₹1,205.90 to ₹1,488.60.
• Reaction session, Monday July 20, Mumbai: opened ₹1,444.50, flat to the prior close, traded ₹1,439.50 to ₹1,480.00, and closed ₹1,460.20, up 1.1% or ₹15.90. Volume was 22.2 million shares against a thirty-day average of 14.4 million, 1.5 times normal. The Nifty 50 closed down 0.4%.
• Reaction session, Monday July 20, New York: IBN opened $29.81, a 0.2% gap up, traded $29.37 to $30.32, and closed $29.76, unchanged to the cent. Volume was 12.5 million shares against a thirty-day average of 7.7 million, 1.6 times normal. The S&P 500 closed down 0.2%.
• Peer reaction: the largest private-sector peer reported the same weekend and fell 5.1% on the July 20 session on roughly two and a half times its prior-week average volume, against ICICI's 1.1% gain. The banking complex was weak that day and ICICI was among the index's gainers.

A twelve percent profit beat, the fastest loan growth in the large-bank complex, a margin that expanded on both a reported and a clean basis, and the first positive-jaws quarter in two years produced a 1.1% move in Mumbai. The intraday high of ₹1,480 was 2.5% above the prior close and more than half of it was given back by the bell, on a session when the index fell 0.4% and the closest comparable fell 5.1%. Relative to the tape that is a strong reception; relative to the print it is a measured one, and the measurement is happening on the same composition question the March quarter raised. The market can see the two below-the-operating-line assists as clearly as we can.

Volume is the more interesting signal. Both listings traded at roughly 1.5 to 1.6 times their thirty-day averages, against 1.1 and 0.9 times after the March print. Real repositioning happened on this print in a way it did not on the last one, and it happened on a day when the sector was being sold. The peer's 5.1% decline on an upbeat quarter of its own is the context: capital rotated within the Indian private banks rather than into them, and ICICI was the destination.

The divergence between the two listings is the observation that matters most for this report's rating, and it got worse rather than better. On the same session, on the same results, Mumbai rose 1.1% and New York closed unchanged to the cent. Over the trailing twelve months into this print the ADS lost 10.4% while the underlying shares gained 1.8%: 12.2 points of gap on identical earnings, against 9.0 points at the March print. The conversion embedded in the two closes moved from roughly ₹94 on April 20 to roughly ₹98 on July 20, so a rupee holder's 7.7% gain between the two reaction sessions arrived as 3.5% in dollars. The rupee took 4.2 points of a 7.7 point move in a single quarter, which is more than the entire twelve-month excess return we are underwriting.

Street Perspective

Debate: Are the positive jaws a turn or a print?

Bull view: Core revenue grew 13.5% against costs of 10.4% and the gap was produced by both halves: fee income accelerated to 23.5% while employee costs grew only 5.5% in the quarter that sets the annual pay base. Core cost-to-income improved 144 basis points sequentially to 38.3%. The branch build continues at 97 additions in the quarter, so this is operating leverage rather than under-investment.

Bear view: The revenue half rests on a fee comparison management itself flagged as a weak base, and the cost half laps a year-ago quarter that carried the highest employee expense of FY26. Take five points off fee growth and two off the cost comparison, which laps the heaviest employee-expense quarter of FY26, and the jaws are roughly neutral. Management still declines to commit to positive jaws as a target and still says it does not manage to a cost-to-income ratio.

Our take: The bull has the print and the bear has the durability question, and one quarter cannot settle it. What tips the balance is that the improvement showed up in the structural line as well as the cyclical one: cost of deposits fell 44 basis points year-on-year and core margin held, which is not a base effect. We raise conviction on this pillar and will not retire the bear point until a second consecutive quarter of positive jaws lands against a normal comparison.

Debate: Does nineteen percent loan growth deserve a re-rating?

Bull view: Advances grew 19.6% with a 4.36% margin, a 2.49% return on assets and a 17.7% annualised return on equity, in a seasonally weak quarter, while the closest large private peer sold off 5% on its own results. Growth broadened this quarter rather than narrowing: retail accelerated to 12.0%, personal loans to 12.9%, mortgages to 14.6%. This is the franchise operating on every cylinder and the multiple should reflect it.

Bear view: Management put the credit for the acceleration on the system rather than the franchise, conceded average growth is lower than period-end, and identified the corporate half as bond-market substitution and precautionary working capital rather than capital formation. The loan-to-deposit ratio has added 700 basis points in two years to 89.0%, coverage fell to 74.7%, and common equity tier-one declined for the first time. Growth of this rate is not free.

Our take: The bear is describing the price of the growth accurately and the bull is describing its quality accurately, and both are consistent with a modest re-rating rather than a large one. Nineteen percent is not the run rate; mid-to-high teens is, and mid-to-high teens on a stable margin with positive jaws is worth more than the roughly three times book the market is paying. We move our fair value up and stop short of a multiple expansion the substitution-driven corporate growth does not support.

Debate: What does the FCNR programme do to the franchise?

Bull view: Hedged funding at roughly 635 basis points deployed into wholesale assets is accretive on every measure that matters, it uses balance-sheet capacity the Bank has been criticised for leaving idle, it closes the deposit-growth gap that has been the most persistent question on these calls, and management describes the earnings effect as clearly positive. The Board raising the offshore issuance limit to USD 2.50 billion in the same meeting shows a coordinated funding strategy rather than an opportunistic raise.

Bear view: The Bank's premium rests on having the highest margin in the Indian large-bank complex, and it is about to dilute that margin voluntarily, at an unknown size, into an international balance sheet that is the least-scrutinised part of the group. The overseas book has already moved from 2.7% to 3.1% of advances. There is no target, no disclosed mobilisation, and the first measurable data point is a quarter away.

Our take: The economics are right and the disclosure is not yet adequate to underwrite them. Our working assumption is that the programme is meaningful enough to compress the reported margin by a few basis points over FY27 and to add to earnings, which nets to roughly nothing in our estimates and materially changes how the margin line should be read. The genuine risk is neither credit nor spread: it is that a bank with a very good domestic story starts being valued on a group balance sheet a third of the sell-side does not model.

Debate: Is the ADS a good way to own this?

Bull view: Same asset, dollar-denominated dividend, deep New York liquidity, and a currency that has now absorbed two years of depreciation with the central bank actively managing the pace and the emergency measures of the spring already partly withdrawn. The ADS at $29.76 has gone nowhere for twelve months while the underlying earnings compounded at mid-teens, which is a widening discount rather than a deteriorating asset.

Bear view: Twelve points of ADS underperformance against the underlying over twelve months is the argument, and the gap widened this quarter rather than narrowing. The conversion embedded in the two listings moved four percent in a single quarter. A rupee investor who bought at the April print is up 7.7%; a dollar investor is up 3.5%, and neither did anything different.

Our take: This remains the debate the sell-side does not have, because it publishes rupee targets to rupee investors, and it is still the single largest reason our rating differs from a unanimous Buy. A ₹1,710 to ₹1,800 target cluster is a 17% to 23% rupee return. A dollar holder receives that less whatever the rupee does, and over the last four quarters what the rupee did was take more than the whole excess return. Nothing in this print changes that, and this print is the second consecutive one in which the operating result deserved better than the ADS delivered.

Model Update

Our FY27 estimates were set at the initiation in April. This print raises most of them. Aardvark Labs does not currently maintain a published financial model for ICICI Bank; these are the working assumptions behind the rating.

Valuation. At the ₹1,460.20 reaction-day close, the Mumbai line trades at 3.05 times June 2026 net worth per share of ₹478.08 and 20.0 times trailing twelve-month basic earnings of ₹72.95, with a 0.9% dividend yield on the ₹13 we expect for FY27. Rolling book forward from the June base at a 17% return on equity and a 17% payout, and taking the ₹8,610 crore FY26 dividend out in the second quarter when it is paid, gives roughly ₹530 of March 2027 book value per share. Holding the multiple at 3.0 times, essentially where the market is already paying, produces a ₹1,590 twelve-month value on the domestic line, about 8.9% above the reaction-day close and 9.8% including the dividend.

Fair value: $31.50 per ADS, raised from $30.00; ₹1,590 per share, raised from ₹1,500. One ADS represents two ordinary shares, so the domestic value translates into the ADS at whatever the rupee does. At the conversion embedded in the July 20 closes of both listings, roughly ₹98, ₹1,590 is worth about $32.45 per ADS. At ₹102 it is worth about $31.18. We anchor at $31.50, which embeds continued but decelerating depreciation, giving 5.8% price appreciation and roughly 6.7% total return in dollars against 9.8% in rupees. The currency assumption remains a larger swing factor in the dollar return than any operating line in the table above, and it is the whole of the difference between our rating and the Street's.

Thesis Scorecard Post-Earnings

The six pillars below were established at the April initiation and are graded here against this quarter's print and call, with the status tag each carried into the quarter shown alongside where it moved.

Overall: the thesis is strengthened on the operating pillars and weakened on the two that constrain the rating. Three quarters of evidence now say the franchise is the best-run large bank in India and that its cost problem was cyclical rather than structural, which is more than we granted it in April. Against that, the growth is being funded by a loan-to-deposit ratio climbing 400 basis points a year and priced in a currency that has taken more than the entire expected excess return over the last four quarters.

Action: Hold, conviction raised from 6 to 7, fair value raised to $31.50 per ADS and ₹1,590 per share. In April we named two conditions that would move us to Outperform: a quarter of positive jaws on the core revenue comparison, or a rupee that stabilises without emergency support. The first was met. The second moved the other way in the same three months, and because the rating is a dollar-return call, the two net out. We do not move the goalposts, so the trigger is now stated precisely: a second consecutive quarter of positive jaws on a normal cost comparison, with the embedded conversion holding below ₹100, takes this to Outperform. The two things that would move us to Underperform are unchanged from April: credit cost normalising above fifty basis points while business banking is still compounding above 20%, or a policy-driven currency move that resets the dollar earnings base again.