Groww Q2 FY27: Cost-to-Serve rises on AI, compute, and one-offs

Billionbrains Garage Ventures Limited (Groww) — Q2 FY27 Earnings Call (14 Oct 2026)

1. Overall Tone of Management: Neutral

Management is generally confident on growth drivers (“trend should continue”, “feedback is really good”, “we go slow” on US product), but repeatedly qualifies with uncertainty/variability:
Regulatory timing dependence (“waiting for the circular to understand the date”)
Volatility-driven outcomes (“between the quarter, there will always be some days in which it goes down”)
No hard timelines (“I don’t think we have the timeline like this” for US beta→full)

2. Key Themes from Management Commentary

Cost dynamics explained quarter-to-quarter (not a demand pullback):
Cost to Grow down due to absence of IPL branding spend this quarter.
Cost to Serve up due to AI-related costs, compute cost increase, more trading days, and a one-time ~INR10–13 crore operational issue on an expiry day.
Regulatory overhang: CAS (SEBI) expected to normalize volumes
Management expects CAS changes to normalize derivatives behavior faster after applicability date; impact framed as derivatives largely VWAP/weighted-average driven.
MTF / margin lending book: steady scaling with volatility
MTF book volatility attributed to market movements and policy/security eligibility changes (e.g., circuit/group changes affecting MTF).
Still expects ~INR600–700 crore quarterly addition and long-term growth due to underpenetration.
Wealth expansion narrative continues
Wealth AUM disclosure introduced for the first time; Groww Prime is the main contributor to user growth.
Prime/MF Prime described as early; limited disclosure to avoid over-specifying.
Product expansion via AI and new offerings
AI used for faster shipping, better quality, and customer support; also in-house stack to reduce per-token cost.
US stocks testing: “feedback is really good” but no fixed beta→full timeline.
Competitive stance: value-first, “paranoid about markets”
Responds to discount brokers/AI tools by emphasizing customer value and first-principles product building.

3. Q&A Analysis

Theme A: Cost structure—Cost to Grow vs Cost to Serve

Core question(s):
Why did Cost to Grow drop while Cost to Serve rose?
Management response:
Cost to Grow lower because IPL branding event cost didn’t recur this quarter.
Cost to Serve higher due to AI costs, compute costs, more trading days, and one-time INR10–13 crore expiry-day file delay hit.
Assessment (evasive/strong/partial):
Direct and detailed explanation with quantified one-time impact; no clear evasion.

Theme B: SEBI CAS regulatory impact & volume normalization

Core question(s):
What to expect from SEBI CAS circular; when will volumes normalize?
Management response:
Derivatives impact expected limited because closing depends mostly on VWAP/weighted average.
Normalization expected faster after applicability date, but timing depends on circular effective date.
Assessment:
Partially evasive on timing (“waiting for the circular to understand the date”), but provides a clear mechanism.

Theme C: MTF / margin lending book trajectory & risk controls

Core question(s):
Why did MTF book moderate then scale up again?
Any one-off costs during September similar to March?
How are stock-level risk measures/haircuts managed?
Management response:
Volatility driven by market direction, position building/squaring, and security eligibility changes (e.g., MTF restrictions).
Policy/rules implemented after prior quarter; no one-off cost this time from MTF; MTF volatility explained as execution/policy strength.
Risk management: proactive exclude from MTF-approved list or apply higher haircuts (example: “80% as haircut” → funding only 20%).
Assessment:
Strong on risk mechanics; “no one-off cost” is asserted, but details remain high-level.

Theme D: Competition & moats (discount brokers, AI tools)

Core question(s):
How competitive landscape is evolving; how to protect moats/customers?
Management response:
“Be paranoid about the markets and focus on the customer”; don’t follow competitors.
If value delivered > price charged, customers stay.
AI features framed as customer decision support.
Assessment:
Value-based but not quantified; typical moat narrative.

Theme E: Wealth management—AUM, user growth, revenue classification

Core question(s):
Break up wealth AUM (Groww Prime vs W); where wealth income sits in revenue mix.
Clarify active user reporting differences due to wealth inclusion.
Management response:
First-time publishing wealth AUM; MF Prime small; Prime fastest-growing.
Wealth income in “other operating income”.
Active user count differences may be due to inclusion of Fisdom users.
Assessment:
Some refusal to disclose too much (“don’t want to disclose too much into those numbers”).

Theme F: Brokerage metrics—market share vs volumes; order intensity

Core question(s):
Why revenue flat despite lower volume; order per ADTV changes.
Why cash/derivatives market share doesn’t reflect NSE active market share.
Management response:
Revenue depends on orders and ticket size/mix, not ADTV directly; commodities/equity derivatives priced per order.
Cash market share convergence expected long-term as customer age increases and MTF/cash turnover rises; may take “10 years” (and “infinity” framing).
Assessment:
Clear explanation of pricing mechanics; long-term convergence admits near-term mismatch.

Theme G: US stocks product & pipeline

Core question(s):
Initial feedback; beta→full timeline; how product pipeline evolves.
Management response:
Feedback “really good”; go slow; no fixed timeline.
Pipeline includes US stocks, MF Prime, and wealth products post-acquisition integration; some products still being figured out.
Assessment:
No timeline = conservative; “feedback good” is positive but not quantified.

Theme H: Other operating income—drivers & reclassification

Core question(s):
What drove other operating income growth? IPO-led? BSDA?
Management response:
Reclassified IPO income + BSDA income into other income.
“About 30 IPOs” in the quarter; large IPOs contributed.
BSDA revenue from new demat account additions.
Assessment:
Strong specificity (IPO count) and accounting clarification.

4. Guidance / Outlook

Explicit guidance (quantitative)

MTF book additions: expect INR600–700 crore quarterly addition (trajectory “should continue”).
MTF-related cash turnover contribution: MTF portion “slightly above 5%” of cash turnover; expected to go up as book grows (but not every quarter).
Cost to Grow / Cost to Serve annual spend range (contextual):
Cost to grow spend: INR400–500 crore per year; average around INR125 crore/quarter; flexibility depending on market.
CAS normalization: qualitative timing—normalization expected faster after applicability date (no numeric date given).

Implicit signals (qualitative)

Regulatory: derivatives impact from CAS expected limited; normalization expected rather than structural impairment.
Wealth: Prime is the growth engine; MF Prime small; management prefers limited disclosure while scaling.
AI: management frames AI as both productivity and customer experience lever; also claims in-house stack reduces per-token cost.
US stocks: “feedback really good” but “go slow” and no fixed launch timeline → cautious execution.

5. Standout Statements (direct / revealing)

Cost to Serve one-time hit: “one-time hit of around INR10 crores to INR13 crore” due to expiry-day file delay.
CAS mechanism framing: derivatives “will either be like 100% based on VWAP or will have a weighted average” → expected limited impact.
MTF scaling expectation: “trend should continue” and “add like INR600 crores, INR700 crores on a quarterly basis.”
Risk control example: “in some cases like 80% as haircut… only funding 20%.”
US product stance: “feedback is really good” but “we go slow” and “I don’t think we have the timeline like this.”
Wealth disclosure restraint: “we don’t want to disclose too much into those numbers.”
Cash market share convergence admission: “probably it will take 10 years… but it will converge one day.”

6. Red Flags / Positive Signals

Red flags
– Limited disclosure / deflection on key metrics:
– Wealth AUM split not provided; “don’t want to disclose too much.”
– Net inflows not reported (asked explicitly; not answered in transcript).
– No hard timelines for major product milestones (US beta→full).
– Regulatory dependence remains a recurring uncertainty (CAS effective date; UPI MDR framework outcome).

Positive signals
– Operational transparency on costs with quantified one-time impact.
– Clear risk management articulation for MTF (haircuts + approved list changes).
– Product momentum evidence: US feedback “really good”; wealth user growth “Prime, yes.”
– Accounting clarity on other operating income reclassification and IPO/BSDA drivers.

7. Historical Comparison & Consistency Analysis (vs prior calls provided)

a. Change in Tone Over Time

Current (Q2 FY27): Neutral vs Q4 FY26: Optimistic
Q4 FY26 emphasized “inflection point how AI will start impacting” and “focus will continue to be on scaling wealth” with more confident forward framing.
Q2 FY27 adds more hedging around regulatory timing and volatility-driven variability (CAS effective date, market volatility affecting volumes).
Shift classification: More Cautious
More “waiting for circular/date” and “no timeline” language.

b. Tracking Past Commitments vs Outcomes

AI inflection / productivity narrative (Q4 FY26): “inflection point how AI will start impacting… productivity side”
Outcome in Q2 FY27: AI is now tied to specific cost drivers (AI-related cost, compute cost) and in-house stack/token cost reduction.
Flag: ✅ Delivered (narrative matured into cost/operations impact).
Wealth scaling after Fisdom acquisition (Q4 FY26): “six months… learnings… continue scaling wealth”
Outcome in Q2 FY27: wealth AUM publishing begins; Prime drives user growth; still early on profitability/scale specifics.
Flag: ⏳ Delayed / Partially delivered (progress visible, but limited quantitative disclosure and no detailed scaling targets).
Algo strategy (Q4 FY26): management previously said no strong algo focus until regulatory clarity.
Outcome in Q2 FY27: still downplays 915/algo contribution as “not significant” today.
Flag: ✅ Consistent (no contradiction; still cautious).

c. Narrative Shifts

From “growth compounding” to “volatility + regulation mechanics”:
Q4 FY26 leaned on market opportunity and AI inflection.
Q2 FY27 spends more time explaining cost variances, MTF volatility drivers, and CAS/UPI regulatory mechanics.
Wealth reporting becomes more formal:
Q2 FY27 introduces wealth AUM disclosure and clarifies revenue classification (“other operating income”)—a step toward more structured reporting.

d. Consistency & Credibility Signals

Medium credibility
Strength: consistent explanations of volatility drivers (market direction, trading days, security eligibility).
Weakness: recurring refusal to provide timelines/precise splits (wealth AUM breakdown, net inflows, some product launch timing).
No obvious contradictions, but precision is selectively withheld.

e. Evolution of Key Themes

Demand / volumes: Stable long-term optimism, but short-term framed as volatility-driven.
Margins/costs: More granular cost bridge in Q2 FY27 (IPL absence, AI compute, one-time hit).
Wealth: Transition from “launch/acquisition” (Q4) to “scaling + reporting” (Q2), but still early on detailed targets.
Regulation: Increasing emphasis on implementation/execution and effective dates.

f. Additional Insights (Cross-Period Intelligence)

Risk is becoming more explicit in cost language: Q2 FY27 quantifies an operational one-time hit and discusses MTF haircut strategy in detail—suggesting management is preparing for scrutiny on risk-adjusted performance.
Convergence narrative for market share is long-dated: “10 years” framing implies near-term market share gaps may persist longer than investors might expect.