Financial Shenanigans

Financial Shenanigans — Forensic Risk Read

Figures converted from INR at historical FX rates — see data/company.json.fx_rates. Ratios, margins, and multiples are unitless and unchanged.

The reported numbers at eClerx Services Limited are broadly a faithful representation of economic reality, but the past two reporting years carry yellow flags that require active monitoring rather than dismissal. The auditor elevated revenue recognition related to unbilled revenue to a Key Audit Matter at both the standalone and consolidated level for FY25 [1] [2]; receivables (billed plus unbilled) grew 59% in FY25 against revenue growth of 15%, the largest revenue-to-receivables gap in eight years of disclosed data; and management itself publicly attributed a softer Q1 FY26 net operating cash flow of just $2.6 million partly to a "significant" gratuity-fund contribution it admitted had been paid pay-as-you-go for the prior four to five years [3] [4]. Against these signals sit several genuinely clean tests — zero net debt, mandatory (not voluntary) audit-firm rotation with no qualifications carried forward [5], an unqualified Price Waterhouse Chartered Accountants LLP opinion for FY25 [6], and a three-year cash conversion (CFO/Net Income) of 1.17x that is genuinely structural rather than working-capital lifelined.

Forensic verdict — Watch (35/100)

Forensic Risk Score (0–100)

Red flags

Yellow flags

Clean tests called out

3-yr CFO / Net Income

1.17

3-yr FCF / Net Income

0.99

FY26 Accrual Ratio

-4.9%

FY25 Recv− Rev growth (pp)

-44.2

Top two concerns. First, unbilled revenue is now an auditor-flagged Key Audit Matter, sitting at $34.4 million at the consolidated level on March 31, 2025, with the audit procedures explicitly testing aging and post-period invoicing of old items [1]. Second, management is publicly transitioning to a new non-GAAP "operating EBITDA" metric that excludes other income, telegraphing at the FY26 close that "we will probably gradually shift our commentary to refer to the operating EBITDA metric" [7] — a metric-redefinition the analyst needs to track even though, in this case, it makes the comparison more conservative than headline EBITDA, not less.

Cleanest offsetting evidence. Net debt is structurally negative (cash of $76.6 million against zero borrowings on March 31, 2026) [8], the auditor was rotated under the mandatory ten-year cap with no qualifications carried over from the outgoing firm [5], and the three-year cash-conversion ratio (CFO/Net Income at 1.17x) is supported by a negative accrual ratio — cash receipts have been arriving ahead of accounting earnings, not behind them.

What would change the grade. A single disclosure: an aging schedule of unbilled revenue that shows material balances older than one quarter without subsequent invoicing. If the FY26 audit report retains the unbilled-revenue KAM and the aging or write-off rate deteriorates, the grade moves to Elevated (45–55). Equally, if the buyback-driven capital return pattern continues into FY27 without operating-cash-flow growth keeping pace, the working-capital lifeline thesis on CF4 strengthens.

How the income statement and cash flow disagree

The single most informative test in this name is the gap between revenue growth and receivables growth — and that gap blew out in FY25.

FY25 is the standout: total trade receivables (billed plus unbilled) grew 59% against revenue growth of 15%, a 44-percentage-point gap that has no analogue in the prior six years. The mechanism is partly visible. The standalone balance sheet at March 31, 2025 carries billed trade receivables of $43.9 million and unbilled receivables of $24.7 million (a unbilled-revenue line that the auditor explicitly identifies as a Key Audit Matter at the standalone level) [2]; the consolidated figure is billed $57.9 million and unbilled $34.4 million [1]. FY26 then carries the stretch forward — billed receivables rose another 35% to $73.3 million and unbilled receivables reached $38.3 million [8].

DSO, computed off year-end receivables (consolidated, including unbilled), confirms the deterioration.

The CFO explained the spike on the Q1 FY26 call with unusual frankness, citing "system and process changes in some of our large clients, resulting in invoices being hold for some time" in BFSI and Retail/Emerging clients [9]. The plausibility of that explanation hinges on whether the held invoices are being written off, billed, or simply rolled forward as unbilled — which is exactly what the auditor's FY25 KAM tests for [1]. Watch the FY26 audit report for whether the KAM persists, and watch the FY26 balance sheet for the aging of unbilled.

Cash flow — strong, but name the mechanism

CFO genuinely outpaces net income, and on a three-year basis FCF roughly equals net income — but the acceleration in FY26 deserves attention.

The FY26 consolidated cash-flow statement reconciles a $103 million pretax profit to $96.1 million of CFO, but the bridge contains material non-operating reversals: a $3.28 million gain on sale of current investments is added back (it sits in Other Income, not operating performance) [10], share-based payment expense of $4.74 million is non-cash, and dividend income (now $0.24 million, up from nil) flatters the year-on-year comparison. The cash flow is real, but the analyst should resist the temptation to multiply CFO/Net Income by the headline trend to project forward: a meaningful share of the year's CFO outperformance versus net income is the working-capital math (employee-benefit-obligations build of $3.7 million, trade-payables build of $2.5 million) plus the mechanical add-back of share-based compensation.

A separate disclosure cuts the other way: in Q1 FY26 the CFO publicly stated net operating cash flow was just $2.6 million in the quarter and the EBITDA-to-CFO conversion was 9.5%, "much lower than usual," and attributed roughly half to a gratuity fund top-up that captured 4–5 years of deferred funding [3] [4]. That admission cuts both ways: management was transparent (good); but the fact that gratuity funding was paid pay-as-you-go for half a decade is exactly the kind of EM5 "hiding expenses or losses (under-reserving)" pattern that the auditor's question forced into the open.

Other income is becoming structurally larger and more diverse

Other income as a share of pretax profit averaged 9–10% from FY21 through FY24, then drifted up to 12.0% in FY25 and 10.8% in FY26 — a smaller share than the FY19 baseline of 16% but with a more diverse composition that includes more one-time items.

The FY25 Note 23 decomposition shows the granular pattern. Items that are clearly recurring (interest income on fixed deposits $2.94 million, FX gain $0.90 million) sit beside items that are sector-specific and lumpy (government grants $1.94 million — nearly double FY24's $1.04 million; profit on sale of current investments $3.28 million — 2.4x FY24's $1.39 million; interest on income-tax refund $0.20 million — a true one-off) [10]. Management itself attributed the Q4 FY25 jump in other income to $2.14 million of "apprentice benefits from government skill development initiatives" in its FY25 earnings call [11].

No Results

Page / 2 10 of 13 records

For valuation work, strip the lumpy items (government grants, profit on investments, gain on PPE, gain on lease modification, FX, tax-refund interest) — that takes ~$6.5 million out of FY25 other income, leaves a recurring base of roughly $3.5 million, and converts the FY25 PBT margin from 21.3% to about 20.5% on a "core" basis. A reasonable buffer; not a thesis breaker.

Capex and capitalization — no signs of soft-asset bloat

Capex has tracked depreciation closely on a multi-year basis, with the capex/D&A ratio averaging 0.66 over FY22–FY26. Soft-asset growth (goodwill plus intangibles) has been slower than revenue growth — goodwill drifted from $51 million in FY21 to $49.4 million in FY26 (a near-flat number despite FX translation effects) while revenue compounded at 21.4%.

R&D is expensed: the FY25 standalone Note 23 reports revenue R&D expenditure of $4.77 million and capital R&D expenditure of nil [12]. This is the right test for EM4 — capitalization of customer-acquisition costs, software development, or contract costs — and the policy is conservative. There is no sign of operating costs being parked as assets to inflate margins. Right-of-use assets have grown with new offices (Mumbai, Chandigarh, Pune, Lima, Cairo, Mohali) but Ind AS 116 mechanics — not aggressive capitalization — drive the line.

The standalone parent records that 66.99% of total purchases in FY25 and 68.71% in FY24 went to related parties, while sales to related parties were 5.08% in FY25 versus 7.50% in FY24 [13]. This sounds alarming until you read Note 31: the counterparties are eClerx Limited (UK), eClerx LLC (US), eClerx Private Limited, CLX Europe S.P.A., eClerx Canada Limited, eClerx Australia PTY, eClerx ME, and eClerx B.V. — i.e., the company's own overseas subsidiaries that act as the parent's onshore sales force [14]. The headline line is "Sales and marketing services by subsidiary to the Company" at $56.1 million in FY25 (up from $47.9 million in FY24) [15], and these are eliminated in consolidation. Two caveats: (i) the BRSR note explicitly states the FY24 figure was global whereas FY25 is restricted to India locations only, so the year-over-year drop in the sales-to-RPT percentage is partly a definition change and partly arithmetic [13]; (ii) the AOC-2 disclosure asserts that all such transactions are at arm's length with omnibus Audit Committee approval. We find no evidence of round-tripping, undisclosed promoter affiliates, or revenue inflation through related entities — the structure is conventional for an Indian export-services group.

Exceptional items — the FY24 Personiv pattern is borderline

In FY24, the Group's Personiv subsidiary "entered into an agreement with one of its clients to transfer its personnel to the client's subsidiary," received a one-time fee of $2.48 million, and in the same period recognized an impairment charge of $2.70 million on the related customer-relationship intangible. Both items were shown net as exceptional items, for a P&L impact of just ($0.22) million [16]. Management framed the impairment on the Q3 FY24 call as "an impact on customer relationship of about INR 225 million or so" and described the deal as a client-led personnel transfer [17].

The pattern is a textbook EM7 candidate to test: a one-time receipt arriving in the same period as a discretionary impairment of similar size, leaving the optical bottom line nearly unchanged. Either the impairment is genuinely event-driven (the personnel transfer means the future cash flows that the intangible represented are gone) — which is the company's case — or the impairment was opportunistically sized to absorb a one-time gain. The audit firm signed off, the receipt and the impairment are reasonably linked in disclosure, and no further "exceptional" items have been booked in FY25 or FY26 [18] [19]. We grade this yellow, not red: the symmetry is convenient but not prima facie improper.

Non-GAAP framing — the new "operating EBITDA"

On the Q4 FY26 call, the CFO announced that "we are now showing both operating EBITDA, which is excluding other income and EBITDA, including other income. We will probably gradually shift our commentary to refer to the operating EBITDA metric, which, as many of you have pointed out, is a more meaningful number to look at" [7]. This is a new top-line management KPI. Two judgments:

The direction makes the metric more conservative, not less — pulling other income out of EBITDA lowers headline margin and forces investors to value the operating business separately from treasury income. That is generally investor-friendly.
The rollout still counts as a KPI definition change and should be tracked. Each quarter going forward, reconcile "operating EBITDA" to reported EBITDA and check that the reconciling items match the Note 23 Other Income line — not a moving subset.

A second non-GAAP overlay is "core EBITDA margin." On the Q2 FY25 call, in response to an analyst question, the CFO confirmed that the 26% Q2 EBITDA margin contained 85 bps of one-offs (sign-on bonuses, higher 401k contributions, and a change in calculation of leave provision) and that "core" margin was therefore ~25% [20] [21]. This is exactly the kind of "recurring one-off" that the analyst should track for repetition — and it has, in the Q4 FY25 deck, recurred in the form of an apprentice-benefit other-income line [11].

A revenue-bucket reclassification happened in Q4 FY25: management restated the "analytics and automation" segment "by removing revenue which we felt did not belong in this bucket" [11]. Yellow: segment growth-rate comparability now requires care.

The 13-shenanigan scorecard

This is the standardized comparable view. Every category has been considered; "no clear evidence" is stated where the test passes.

No Results

Page / 2 10 of 13 records

Breeding ground — governance, incentives, audit

Several structural factors dampen the accounting risk:

Mandatory auditor rotation. S.R. Batliboi & Associates LLP completed the second 5-year term and was replaced by Price Waterhouse Chartered Accountants LLP for FY25–FY29 under the statutory 10-year cap. The FY24 directors' report explicitly states "there are no qualifications, reservations, adverse remarks or disclaimer made by M/s. S.R. Batliboi & Associates LLP, Statutory Auditors in their report for FY2024. The Statutory Auditors have not reported any incident of fraud" [5]. The first-year Price Waterhouse opinion for FY25 is unqualified [6].
Zero financial leverage. Short-term borrowings nil, long-term debt nil, net cash position of $76.6 million on March 31, 2026 [8]. There is no debt covenant or refinancing pressure that creates a structural incentive to flatter earnings.
Capital return is dominantly buybacks, not dividends (FY25 buyback $44.9 million vs dividend $0.55 million) [10], and management has publicly stated that promoters and promoter group will not participate in the FY26 buyback — a credible signal that buybacks are intended as per-share-value-accretion rather than promoter exit.

A few factors amplify it:

Promoter-co-founder dominance. P.D. Mundhra and Anjan Malik remain reference shareholders and management influences; the related-party universe disclosed at IPO (DRHP/RHP) was extensive and the post-IPO Note 31 intra-group flows are large. We find no evidence of related-party leakage in current disclosure, but this is a single point of failure to monitor.
Recurring "one-offs." The Q2 FY25 85 bps benefit (sign-on bonuses, 401k, leap-provision change) [20] and the Q4 FY25 $2.14 million apprentice benefit [11] both fit a pattern where "non-recurring" items appear with quarter-to-quarter regularity. The new "operating EBITDA" metric helps the reader strip these out, but the analyst still needs to track repeats.
Gratuity funding admission. The fact that the gratuity fund had been paid pay-as-you-go for 4–5 years and required a catch-up "significant contribution" in Q1 FY26 [4] is a vivid reminder that the company can carry undisclosed liabilities that surface only when the auditor pushes. It does not prove there are others, but it is the kind of disclosure that argues for less benefit of the doubt elsewhere.

Material disclosure to underwrite forward. Q1 FY26 (management call, July 2025): "We had applied to shift our gratuity from LIC to HDFC way back in 2019. And then COVID happened and somehow that application got lost, we lost track of it. And when the auditors brought it up, then we realized that this hadn't been completed. And we did that process… For the last 4, 5 years, we've been paying as we go. So, this is a significant contribution that we have not made in the past." This is an admission that an employee-benefit obligation was under-funded for half a decade — a textbook EM5 signal that surfaced only because the rotated auditor asked the question.

What to underwrite next

The five highest-value items to monitor through FY27:

The FY26 audit report. Does the "Revenue recognition related to Unbilled revenue" KAM persist into the first full Price Waterhouse audit? If yes, the FY26 aging of unbilled becomes the load-bearing test. If the KAM is dropped, that itself is meaningful — it means the auditor judges the FY25 spike to have unwound or been confined to specific clients.
The aging schedule of unbilled revenue. Not disclosed at year-end FY25 in a single table; reconstruct from Note 8 footnotes and quarterly balance-sheet snapshots. Anything older than 90 days without subsequent invoicing is a red, not yellow, flag.
DSO normalization. Management has guided that DSO should "bring in line with what it has been on average in the previous quarters between 80 and 82" [9]. Year-end FY26 DSO was still 81 (per management) [7]. Compare each subsequent quarter against this floor; an upward drift is the signal to reassess.
Definition consistency of "Operating EBITDA." Each quarter, reconcile "operating EBITDA" to reported EBITDA and verify the reconciling items map to the Note 23 Other Income line items exactly. If the management bridges drift over time (e.g. including or excluding lease-modification gains), the metric is being moulded.
Other income — government grants and apprentice benefits. These are tied to specific Indian government skill-development schemes that can expire or be capped. Strip $1.8–$2.9 million from forward-year other income for prudent valuation work.

The signal that would downgrade the forensic grade (to Elevated, 45–55): the FY26 audit retains the unbilled-revenue KAM and the aging shows material balances older than two quarters, or a second auditor-prompted disclosure of a deferred employee-benefit liability emerges.

The signal that would upgrade the grade (to Clean, 15–20): unbilled revenue normalizes below $23 million on the FY26 consolidated balance sheet without a corresponding bad-debt charge, and the FY27 audit report drops the KAM.

For valuation and position sizing, this work argues for a modest haircut — strip lumpy other-income items (~$5.8–$7.0 million annually) before computing operating earnings, apply a small (~5–10%) discount to headline FCF to reflect non-recurring CFO add-backs, and treat any FY27 PE multiple expansion above ~25x on headline EBITDA with skepticism until the unbilled-revenue KAM is resolved. The accounting risk here is a valuation haircut and a position-sizing limiter, not a thesis breaker — eClerx Services Limited's underlying economics (zero debt, ~28% ROE, structural cash conversion) survive the forensic adjustments comfortably.

References

Generated Jun 20, 2026 00:08 UTC