Financial Shenanigans

Forensic Risk Grade: Watch (Score 38/100)

The financial statements appear faithful to economic reality, but FY2024–FY2025 carries enough pre-IPO clean-up choreography — a ₹7,786 mn one-time founder bonus, a ₹13,397 mn US "outbound merger" exceptional tax, a long-term incentive plan accrued then cancelled, and a fresh independent board appointed inside the four months before listing — that the headline FY2025/FY2026 P&L should not be taken as an unmodified trend baseline. The auditor (B S R & Co. LLP, a KPMG India affiliate) issued an unqualified opinion on the restated consolidated financials and no restatements were required. The single largest forensic finding is not an accounting choice — it is the gap between reported net income and reported cash flow from operations, which is structural to a broker / NBFC holding client float and a growing loan book, but means PAT is a poor near-term cash proxy. The one data point that would most change the grade: a clean FY2027 in which reported PAT, pre-working-capital CFO, and FCF after acquisitions all triangulate without a fresh exceptional line.

1. The Forensic Verdict

The headline CFO/NI of 0.03x over FY2024–FY2026 looks like a five-alarm red flag in isolation. The pre-working-capital reconciliation tells a very different story: in FY2025, operating cash flow before working-capital changes was ₹24,868 mn, slightly above reported PAT of ₹18,244 mn (a 1.36x ratio). The full negative reported CFO is explained by (a) ₹19,054 mn of tax paid net of refunds — largely catching up on the US outbound-merger provision, (b) a ₹10,412 mn increase in the NBFC's loan book, and (c) ₹8,190 mn of bank balances earmarked with exchanges. Those are not accounting tricks — they are the cash mechanics of a broker / NBFC scaling. But they do mean that PAT cannot be read as proxy free cash flow until the lending and float build stabilises.

13-shenanigan scorecard

2. Breeding Ground

Founder-promoter dominance over a brand-new public-company structure is the dominant breeding-ground signal — not because of evidence of misuse, but because the checks that normally constrain it were installed in the four months before listing. All four founders are simultaneously promoters, whole-time directors and key managerial personnel; the CFO is a co-founder; the chairman of the Audit Committee was appointed January 29, 2025, and the other two Audit Committee members February 20, 2025. The auditor's CARO note flags that audit trail at the database level was not enabled until 3 February 2025 for the holding company and 11 subsidiaries, and that one subsidiary's third-party-operated revenue/operations system has no service-organisation report confirming audit trail at all.

The two subsidiary-level SEBI settlements deserve unpacking. Per the May 2, 2025 settlement order, SEBI's inspection of Groww Invest Tech Private Limited (the broking subsidiary) identified, among other findings, "discrepancies in financial ledger balances and margin obligations in 38 instances" in client retention statements, failure to update financial/income details in 83 instances, absence of an alternative trading access during major app failures, and offering non-securities services (UPI payments, bill payments, lending) under the secure login of the trading app. The total settlement consideration across both orders was approximately ₹82 lakh. The matters are operational compliance findings, not financial-reporting fraud, but they are material to the breeding-ground read because they came from a SEBI inspection cycle that ran concurrent to the IPO drafting period.

3. Earnings Quality

Underlying earnings quality is reasonable, but the FY24-to-FY25 inflection is partly an accounting-timing pattern that should not be straight-lined into FY27. The reversal of the ₹1,062 mn LTI plan in FY25 alone added roughly 2.6 percentage points to operating margin in the year. Both FY24 below-the-line cleanups — the ₹13,397 mn US outbound-merger exceptional and the ₹7,786 mn one-time founder bonus — hit in the same fiscal year, immediately ahead of listing.

Margin inflection and the FY24 big bath

The 36-percentage-point jump in operating margin from FY24 to FY25 (27% → 62%) decomposes as: roughly 23pp from absence of the ₹7,786 mn one-time bonus; roughly 4pp from absence of the ₹1,062 mn LTI accrual; roughly 2.6pp from the reversal of the prior-year LTI plan; the remaining ~7pp from genuine operating leverage on a 49.5% increase in revenue from operations. The FY26 OPM at 59% is the closer baseline. Net margin volatility is even more pronounced — FY24 NPM of -28.8% is entirely a function of the US outbound-merger tax classified as exceptional.

One-time items in FY24

The two large items are real economic events. The US outbound-merger tax flows from the genuine redomicile of Groww Inc. (Delaware) into Billionbrains Garage Ventures Limited (India), which triggers a 21% US "outbound" tax on the unrealised gain in transferred assets — exactly the structure SEC and IRS rules contemplate. Classifying it as exceptional, not operating, is defensible. The ₹7,786 mn one-time bonus is harder. It is recorded as an operating employee-benefits expense in FY24, which means operating margin already absorbs it; but the company excludes it from Adjusted EBITDA. So management has it both ways: the headline GAAP operating profit shows the hit; the headline non-GAAP "Adj EBITDA" pretends it never happened.

Revenue vs receivables — clean

DSO rose from 9 days to 22 days from FY25 to FY26, but starting from a trivial base — implied receivables of ₹2,800 mn against revenue of ₹46,450 mn (just 6% of annual revenue). Most of the FY26 rise is a function of the new Margin Trading Facility book (₹10,358 mn as of June 30, 2025) and the NBFC personal-loan book, which are not really "trade receivables" in the manipulation sense — they are interest-earning assets with disclosed ECL provisioning. There is no contract-asset build-up, no unbilled-receivables line, no extended-terms quarter-end pattern.

Reserves and provisions — sector-normal

The NBFC ECL allowance reconciliation shows opening allowance ₹477 mn → closing ₹536 mn (June 2025) with ₹190 mn written off and ₹249 mn new originations — sector-normal coverage as the loan book scales. No evidence of under-provisioning. Trade-payables include client float; trade-receivables are largely exchange-mediated (no historical loss). Inventory does not exist — this is a services business.

4. Cash Flow Quality

The reported CFO/NI ratio is misleading without the broker-specific structural overlay. Three lines drive every quirk: client float (booked as trade payables), the NBFC loan book (booked as "Loans"), and bank balances earmarked with stock exchanges (booked as "Other bank balances"). Movements in all three flow through CFO, often in opposite directions in the same year.

The CFO/NI gap

In FY2025 the headline CFO/NI ratio was -0.53x. Strip out the working-capital movements and CFO before working-capital changes was ₹24,868 mn — above reported net income of ₹18,244 mn (1.36x). The gap is fully explained by the components below.

What broke the FY25 reported CFO

This walk is the single most important picture in the forensic file. The negative reported CFO is overwhelmingly a function of (a) paying ₹19,054 mn of tax, largely catching up on the FY24 outbound-merger provision and on a profitable FY25, and (b) deploying ₹10,412 mn into the lending books. Neither is recurring at FY25 magnitude — the merger tax will not repeat, and the lending book either stabilises or becomes part of investing-style growth that should be evaluated against FCF after lending growth, not against PAT.

Client-float dependence

Per the prospectus, "Our outstanding dues of creditors other than micro enterprises and small enterprises increased to ₹60,932.20 million as of June 30, 2025 from ₹13,732.44 million as of March 31, 2023 primarily due to an increase in client funds which are not yet invested." That growth — ₹47,200 mn (~US$530 mn) of client-pending-invest float in roughly 27 months — has the same forensic implication on both sides: it inflates reported CFO when growing and will deflate reported CFO if customer behaviour reverses. It is honest disclosure, not a shenanigan, but it means reported CFO is a participation-rate reading rather than an earnings-quality reading.

FCF after acquisitions

Capex remained tiny throughout (₹70–170 mn/year on an ₹40 bn revenue base). The cash-flow distortion from acquisitions has been modest so far — Groww AMC and Groww Creditserv added ₹436 mn of intangibles to the balance sheet. The Finwizard (Fisdom) acquisition completed October 2025 for ₹9,611 mn (~US$108 mn) is the first material acquisition outflow and will show up in FY26 investing-activity disclosure. Track FCF after acquisitions in FY27 once Fisdom is bedded down.

5. Metric Hygiene

Management discloses a long list of non-GAAP measures — Adjusted EBITDA, Cost to Serve, Cost to Operate, Adjusted Cost to Operate, Contribution Margin — and reconciles every one explicitly to the Ind AS line items. The hygiene is structurally good. The substance question is whether the exclusions from Adjusted EBITDA are appropriate.

The Adjusted EBITDA gap

In FY24, reported PAT was a loss of ₹(8,055) mn while Adjusted EBITDA was a gain of ₹14,709 mn — a gap of ₹22,764 mn. The arithmetic is exact and disclosed: ₹13,397 mn US merger tax + ₹7,786 mn one-time bonus + ₹1,062 mn LTI accrual + ₹207 mn SBC + ₹67 mn associate share + ₹1,870 mn tax + ₹42 mn finance + ₹201 mn D&A − ₹1,867 mn other income = ₹22,765 mn. None of the add-backs are inventive. But the "Adjusted" series effectively asks the reader to imagine FY24 without the costs that actually happened. An investor underwriting Groww should look at the Adj EBITDA series as an operating-leverage diagnostic and at the reported series as the economic outcome — not the other way around.

Stopped or renamed metrics

The pre-IPO prospectus refers to several KPIs (Active Users, MAUs, Broking Transacting Users, Orders per User, AARPU, Annual Average Revenue per User, Cost to Grow per New Transacting User). The first two post-IPO shareholder letters continue to disclose Active Users, Net Active Customers Added, and AARPU on a consistent basis. There is no evidence yet of a KPI definition change or stopped disclosure.

6. What to Underwrite Next

The forensic risk grade is Watch (38/100). Translation: this is not a footnote. It is a position-sizing limiter and a "trust the next two annual reports before extrapolating margins" instruction. It is not a thesis breaker, and there is no evidence of accounting fraud or undisclosed misconduct.

Top diligence items to watch

Grade-change triggers

Position-sizing implication

This forensic risk profile should affect how much you size, not whether you own. Reported FY26 PAT of ₹20,830 mn (~US$222 mn at year-end FX) is supported by clean accounting; reported FY26 CFO of ₹(210) mn is not a quality signal because it absorbs growth in client float, exchange-earmarked balances and a growing NBFC loan book. An investor should haircut headline operating margin by ~3 percentage points to remove the LTI reversal benefit, refuse to capitalise the FY24 big-bath pattern as a permanent trend, and demand a discount-rate premium of perhaps 100–200 bps over an Indian broker peer with a longer-listed track record and a non-founder CFO. None of this is a thesis breaker — it is the price of being underwriting a 6-month-listed company whose pre-IPO P&L was reshaped by two large below-the-line cleanups in the same fiscal year.