Subex Says FY27 Will Convert 24% Order Growth

Subex Limited — Q4 FY26 Earnings Call (Quarter & Year ended Mar 31, 2026) | Call held May 13, 2026

1. Overall Tone of Management: Optimistic

• Management repeatedly emphasizes “reset is done”, “foundation is built”, and “FY27 is where we will convert it.”
• Strong confidence language: “pipeline is real”, “organized and resourced to deliver”, “very bullish”.
• Acknowledges underperformance vs expectations (“could have moved faster on the top line”) but frames it as execution timing rather than fundamental weakness.

2. Key Themes from Management Commentary

• Transformation completed / fundamentals rebuilt
• Claims “first phase of that transformation has been delivered” via exiting non-core businesses, freeing capital, and rebuilding the foundation.
• Order intake momentum + conversion narrative
• “order intake grew 24%” and pipeline conversion is the focus for FY27.
• Explains revenue timing due to implementation cycles and multi-year contracts.
• Profitability consistency + liquidity improvement
• EBITDA margin “14.5%” (quarter) and PAT “13.6%”.
• “Positive EBITDA is 9 out of the last 10 quarters” and “three successive quarters of positive PAT.”
• Liquidity strengthened by “INR 70 crores over the year,” enabling investment “with conviction” rather than defensive management.
• AI traction becomes central
• “included more AI deals than any other prior year” and “more AI models into production this year than ever before.”
• GenAI investment: “embedding GenAI across our entire MCLC” to improve developer productivity and release velocity.
• Geopolitical risk acknowledged, mitigated operationally
• Middle East exposure described as unpredictable, but delivery has been offshored and renewals are expected to be controlled.
• Market messaging / valuation gap
• Management calls the stock “materially undervalued” and expects the market to “catch up” through execution.

3. Q&A Analysis

Theme A: Why revenue growth lags order intake; timing of conversion

• Core questions
• Orders up (and large order announcements previously), but revenue only ~3% sequential—“is it really not converting into revenue? How or when will it get converted?”
• Will Q1 revenue grow faster?
• Management response
• Revenue recognition depends on contract structure: multi-year + implementation phase before subscription/support kicks in.
• Expects conversion to improve “as you move from quarter to quarter” once implementation cycles complete.
• Assessment
• Not evasive; explanation is consistent with prior quarter’s “implementation → subscription” model.
• However, management avoided giving a quantitative conversion schedule beyond qualitative “uptick” expectations.

Theme B: Geopolitical impact (Middle East / Iran war) and demand outlook

• Core questions
• How do geopolitical issues affect Subex, especially Middle East and AI-related competitive dynamics?
• Impact of Iran war given “31% of the revenue is coming from that region.”
• Management response
• Middle East risk: delivery exposure mitigated by offshoring and partner support.
• Renewals: “don’t see a major risk” unless war escalates; new purchase decisions may face delays.
• AI: framed as “tail end” for their business stack (System of Record embedded in telco), not a threat.
• Assessment
• Strong mitigation narrative on delivery; more cautious on new deal timing.
• Competitive AI model discussion is somewhat dismissive (“tailwind”), but not directly quantified.

Theme C: Product commercialization metrics (FraudZap, “team of LLMs/agents”)

• Core questions
• How many paid customers for FraudZap and for the “team of LLMs” solution?
• When will “team of LLMs” become commercial?
• Management response
• FraudZap: “one paid customer” earlier, then clarified “three customers” via onboarding through an intermediary (telcos onboarded to the product).
• Team of LLMs: POCs completed; “in conversations with at least two or three customers” to commercialize “in this quarter,” but delayed by GPU/hardware procurement constraints.
• Workaround: exploring “agent in a box” and CPU-based approach.
• Assessment
• Mixed: clear on customer count, but commercialization timing is still conditional (“should be making it commercial”).
• GPU constraint is a concrete operational blocker (credible), but it also implies a dependency on customer capex/procurement.

Theme D: Margins by revenue component; EBITDA bridge

• Core questions
• Margins for license vs implementation vs managed services; why gross margins don’t translate directly to EBITDA.
• Any further cost optimization opportunities?
• Management response
• Provided gross margin ranges: license very high, implementation ~40–45% (up to 60%), MS 50%+, AMC 70%+.
• Explained EBITDA differs due to below-COGS expenses (R&D, GTM, G&A).
• Claimed EBITDA improvement came from cost control and AI reducing engineering cost.
• Assessment
• Direct and detailed; no major evasiveness.
• Still, it’s mostly gross margin and qualitative EBITDA drivers, not a full bridge.

Theme E: Guidance / targets / willingness to provide numbers

• Core questions
• Analysts asked for growth targets, “stretch targets,” and when INR 100 cr/quarter revenue could happen.
• Management response
• No formal quantitative guidance; management reiterated bullishness and “trust the process.”
• For INR 100 cr/quarter: “Inshallah… this is the year we go for it” and revenue uptick expected as implementation cycles finish (implied mid-year).
• Assessment
• Strong confidence but no hard guidance; answers are more narrative than measurable.

Theme F: Customer concentration, revenue mix, and contract economics

• Core questions
• Top 10 customer share; key clients; revenue split product vs service; milestone payment visibility.
• Management response
• Top-tier customers: “I think about 60% comes from the top tier customers.”
• Key clients: logos disclosed in investor presentation.
• Revenue model: license + implementation + support (MS/AMC); not purely subscription.
• Milestone payment disclosure: declined due to confidentiality and competitive sensitivity.
• Assessment
• Reasonable refusal (confidentiality/competition), but it limits investor ability to model cash conversion.

Theme G: Balance sheet / legacy items (goodwill, Sectrio)

• Core questions
• Status of Sectrio receivables; goodwill impairment status.
• Management response
• Sectrio: one contract closed; second is being pursued legally; prior write-off already provisioned (“no risk… already provided for”).
• Goodwill: annual impairment exercise completed; auditors concluded carrying value fine; “no need to invest.”
• Assessment
• Clear closure status; reduces uncertainty on goodwill impairment risk.

4. Guidance / Outlook

Explicit guidance (quantitative)

• FY27 total income: “6% YoY” (management stated FY27 total income came in at 6% YoY—note: this is retrospective for FY27, but it was discussed in the call as part of the FY27 framing).
• Q4 sequential growth: “3% sequential growth.”
• Revenue target (qualitative/aspirational):
• INR 100 crores/quarter: “this is the year we go for it” and “Inshallah” (no formal numeric timetable beyond implementation-cycle logic).

Implicit signals (qualitative)

• Conversion expectation: revenue conversion should “look better and better” quarter-to-quarter as implementation cycles complete.
• AI commercialization: FraudZap is live; “team of LLMs” expected to become commercial “in this quarter” subject to GPU constraints.
• Investment posture: liquidity enables investing “with conviction” and moving away from defensive management.
• Risk posture: renewals expected to be stable; new deal decisions may be delayed in Middle East if conflict escalates.

5. Standout Statements (most revealing)

• Transformation/positioning
• “the reset is done. The foundation is built and FY27 is where we will convert it.”
• Liquidity as strategic enabler
• “We can now invest with conviction… stop managing defensively.”
• Market valuation claim
• “stock is materially undervalued related to our fundamentals and our trajectory.”
• Revenue conversion mechanism
• “once we complete the implementation phase… subscription revenues will kick in.”
• Middle East risk split
• “delivering on the project…” mitigated via offshoring; renewals controlled, but “new purchase decisions” have uncertainty.
• AI commercialization bottleneck
• “hardware challenge… requires some GPUs… customers… reluctant to procure GPU.”
• No formal guidance stance
• When pressed: “Curiously… no forward-looking guidance” (analyst observation); management responded with bullish narrative rather than numbers.

6. Red Flags / Positive Signals

Positive signals
– Strong profitability consistency: “Positive EBITDA is 9 out of the last 10 quarters.”
– Liquidity improvement: “INR 70 crores” strengthened.
– Concrete operational mitigation for geopolitical risk (offshoring delivery).
– Clear explanation of revenue mechanics (implementation → subscription/support).

Red flags
– No hard quantitative guidance despite repeated investor pressure; reliance on “trust the process.”
– Revenue growth still modest vs order intake; conversion timing remains a recurring theme.
– AI “team of LLMs” commercialization depends on customer GPU procurement—could delay scaling.
– Customer concentration: “~60% from top tier customers” (risk if any customer pauses spend).

7. Historical Comparison & Consistency Analysis (vs prior 3 calls provided)

a. Change in Tone Over Time

• Q2 FY26 (Nov 2025): tone was “reset/cleanup” focused; management emphasized returning to profitability and board reconstitution; growth still “remains a challenge.”
• Q3 FY26 (Feb 2026): tone improved to “steady recovery,” commercialization progress (FraudZap commercialization within a year), and faster deal flips; still conversion timing discussed.
• Q4 FY26 (May 2026): tone becomes more execution-forward and confident: “reset is done,” “FY27 singular focus,” “very bullish,” and explicit undervaluation claim.
• Classification: More Optimistic than prior calls.

b. Tracking Past Commitments vs Outcomes

• FraudZap commercialization
• Past (Q3 FY26): “successful commercialization of FraudZap… within just a year.”
• Current (Q4 FY26): FraudZap has paying customers; clarified to “three customers” via intermediary onboarding.
• ✅ Delivered (commercialization and paid adoption confirmed).
• Top-line recovery / INR 100 cr/quarter ambition
• Past (Q2 FY26): internal push for INR 100 cr quarter; timing uncertain.
• Current (Q4 FY26): “this is the year we go for it” but still framed around implementation completion and “Inshallah.”
• ⏳ Delayed / Not yet proven (no quantitative timetable or evidence of reaching the run-rate yet).
• Order conversion expectations
• Past (Q3 FY26): expected conversion dynamics; AI line faster.
• Current: still explaining conversion lag due to implementation cycles; no hard conversion schedule.
• ⏳ Partially delivered (profitability improved; revenue conversion still modest sequentially).

c. Narrative Shifts

• From “cleanup/reset” → “conversion/execution”
• Q2: heavy emphasis on board reset, legacy cleanup, and returning to profitability.
• Q4: emphasis shifts to “pipeline converting,” “liquidity to invest,” and AI production scaling.
• AI messaging becomes more central
• Q2: AI pivot and GenAI native direction.
• Q3: FraudZap commercialization and GenAI agent POCs in Europe.
• Q4: AI deals and production deployments “more than ever,” plus GenAI embedded across MCLC.
• Data center adjacency introduced as “starting to consider strongly”
• Q4 adds more explicit adjacency expansion (data centers) while still avoiding utilities due to receivables/slow decisions.

d. Consistency & Credibility Signals

• Consistent revenue mechanics across calls: implementation cycles delay subscription/support revenue.
• Credibility improved on profitability and operational discipline (positive EBITDA/PAT consistency).
• Still limited credibility on growth timing: repeated bullishness without quantitative guidance; revenue conversion remains the key investor skepticism point.
• Overall credibility: Medium (strong operational/profitability execution; growth guidance remains narrative and conditional).

e. Evolution of Key Themes

• Demand / pipeline: improving narrative from “robust order book” (Q2) → “healthy pipeline” (Q3) → “pipeline is real / market conversations strongest” (Q4).
• Margins / profitability: steady improvement and consistency (positive EBITDA/PAT) is the most stable theme.
• AI: evolves from “AI pivot” (Q2) → “commercialization + production deployments” (Q3/Q4) → “GenAI embedded across MCLC” (Q4).
• Geopolitical risk: acknowledged earlier as controlled exposure (Q3) and becomes more operationally mitigated (offshoring delivery) in Q4.

f. Additional Insights (Cross-Period Intelligence)

• Conversion remains the bottleneck: even with strong order intake growth (24% in Q4), management still leans on implementation-cycle explanations rather than showing acceleration in revenue run-rate.
• AI scaling may be constrained by customer infrastructure procurement (GPU issue), which could cap near-term growth despite strong product readiness.
• Management is increasingly willing to discuss valuation and market perception, suggesting confidence in fundamentals—but also potentially indicating they need the market to re-rate to validate the transformation story.