Gold on Paper, Nothing in the Vault: The Making of a ₹4 Lakh Crore Business That Didn’t Exist

A SEBI interim order dated June 3, 2026, raised concerns regarding the financial reporting and governance practices of Rajesh Exports Limited (REL). According to the findings discussed in the article, SEBI prima facie observed that REL attributed substantial revenues to overseas subsidiaries, while audited accounts of a key subsidiary reflected significantly lower revenue figures. The regulator also examined transactions recorded in the company’s standalone accounts and noted that large sales and purchase entries had been booked against a stockbroker that reportedly denied having any transactions with the company. SEBI further reviewed fund transfers from the company to the personal accounts of the chairman, his son, and another entity associated with the chairman, finding that several transactions had not been disclosed as related-party transactions. The order also highlighted accounting issues, including the treatment of foreign exchange gains and interest income as revenue from operations, the non-elimination of intra-group balances in consolidated accounts, and the handling of long-outstanding trade receivables. The article additionally discusses concerns regarding the use of unaudited subsidiary financial information in consolidation, the absence of audit trails in accounting software, the lack of identified Key Audit Matters in audit reports, and the issuance of unmodified audit opinions despite various risk indicators. It notes that SEBI’s findings are prima facie in nature and that legal proceedings remain ongoing.

Gold on Paper, nothing in the Vault: How to Build a ₹4 Lakh Crore Business That Doesn’t Exist

Let me start with two sentences.

The first was written on August 5, 2025, by M/s BSD & Co., Chartered Accountants, in their audit report on Rajesh Exports Limited: “The aforesaid standalone financial statements give a true and fair view.”

The second was written ten months later, on June 3, 2026, by a Whole Time Member of SEBI: “REL has prima facie misrepresented approximately ₹15,15,385 crore representing 99.80% of its revenues attributed to subsidiaries.”

Same company. Same financial statements. Completely opposite conclusions.

One of these sentences was written by the people whose entire professional existence is built around answering the question: are these numbers real? The other was written by a regulator who had to step in because nobody else had asked the question loudly enough.

That gap between what the auditors said and what turned out to be true is what this article is about. Not just the fraud itself, as extraordinary as it is. But the ecosystem that allowed it to survive, and the lessons that every auditor, investor, and company director must take away before the next one begins.

Before the Fall

For a while, everything about Rajesh Exports Limited (REL) looked like a success story. The Bengaluru-based gold company had been listed on BSE and NSE for years. It had the ‘Shubh Jewellers’ retail brand that ordinary Indians could walk into. And through a chain of overseas subsidiaries REL Singapore, then Global Gold Refineries AG in Switzerland, then Valcambi SA, it claimed to be connected to one of the world’s most prestigious precious metal refineries.

On paper, the numbers were staggering. Consolidated revenue of ₹2.58 lakh crore in FY 2020-21. Then ₹2.43 lakh crore. Then ₹3.39 lakh crore. Then ₹2.80 lakh crore. Then ₹4.23 lakh crore in FY 2024-25. For comparison, ₹4.23 lakh crore is more than the entire annual revenue of most companies in South Asia. It made REL sound like it was operating at the scale of a sovereign entity trading gold for the world. Markets believed it. The share price climbed to ₹1,028.40 on February 6, 2023. The company’s market capitalization touched ₹30,365 crore at its peak. Shareholders poured in from 22,472 in March 2020 to over 2,06,942 by September 2025. These were real people, making real investment decisions, based on financial statements that had been certified as true and fair. Then someone filed a complaint.

In March 2024, a shareholder sent SEBI an email. Not an elaborate whistleblower tip. Just a simple observation: REL had enormous trade receivables that had been sitting unpaid for more than two years. What, the shareholder wanted to know, was going on?

What followed was a forensic investigation that would unravel five years of alleged financial misrepresentation on a scale that, frankly, is difficult to fully comprehend even after reading the SEBI order several times.

What Actually Happened

The Illusion Factory. Here is the simplest way to understand what SEBI prima facie found. REL told its shareholders it was a ₹4+ lakh crore business. The engine of this scale, REL always explained, was its overseas subsidiaries principally Valcambi SA in Switzerland, one of the genuinely great precious metal refiners in the world.

So, SEBI did the obvious thing. It looked at Valcambi SA’s own audited accounts prepared under Swiss law, audited by KPMG SA. And Valcambi’s revenues for the five-year period FY 2020-21 to FY 2024-25? A combined ₹3,027 crore. Not ₹3,027 crore per year. ₹3,027 crore across five years. REL’s subsidiaries, over the same period, were credited with ₹15,18,413 crore in consolidated revenues. The gap between those two numbers, ₹15,15,385 crore, is what SEBI prima facie calls misrepresentation. It represents 99.80% of all revenues that REL attributed to its overseas subsidiaries. The business REL presented to the world was, prima facie, almost entirely imaginary.

The mechanism was simple but obscured by structural complexity. Between Valcambi SA (the real operating entity) and REL India (the listed parent) sat a holding company called Global Gold Refineries AG (GGR), also in Switzerland. GGR had, by REL’s own admission, no day-to-day operations. It was a shell. Yet GGR prepared internal “consolidated accounts” not subject to statutory audit, not compliant with any recognized international standard, prepared under a homemade “Group Accounting Manual” that showed revenues vastly higher than Valcambi’s audited numbers.

REL then used GGR’s unaudited, inflated figures as the basis for its own Indian consolidated statements. It chose the fictional unaudited number over the real audited one and it did this while already possessing Valcambi’s KPMG-audited accounts, knowing exactly what they showed.

The Stockbroker Who Never Knew

If the consolidated fraud is bewildering in its scale, the standalone fraud is almost darkly comic in its brazenness.

When SEBI analyzed REL’s standalone sales and purchase data, it found something peculiar. About 66% of standalone sales ₹11,487 crore and 67% of purchases ₹11,488 crore during FY 2021-22 to FY 2023-24 were booked against a single counterparty: Affluence Shares and Stocks Private Limited, a stockbroker. The net difference between these sales and purchases over three years? ₹1.82 crore. On ₹11,000-crore-plus of transactions. No genuine business in the world has that profile. You do not buy and sell over ₹11,000 crore of gold to generate ₹1.82 crore of margin.

SEBI then did something simple: it asked Affluence what was going on. Affluence said, under oath, that it had never dealt with REL. No agreement. No transactions. Nothing. What Affluence did have was a trading relationship with Mr. Rajesh Mehta REL’s chairman, promoter, and managing force in his personal capacity. Mr. Mehta had traded gold derivatives through Affluence on 102 trading days during those three years. REL had funded these trades through his personal account. The net losses from the trading were absorbed. The residual was returned.

REL had taken its own chairman’s personal gold speculation and booked it as the company’s own sales and purchases. ₹11,487 crore in entries, symmetrically constructed, appearing to show a thriving trading business that existed only on paper. No board approved it. No audit committee saw it. No related party disclosure mentioned it. The shareholders who owned this company had no idea.

The Money That Went Wandering

Beyond the revenue fabrication, SEBI found a pattern of fund flows that reads like a case study in how not to run a public company. During April 2020 to September 2025, REL transferred ₹338.90 crore to Mr. Rajesh Mehta’s personal bank accounts. Of that, ₹232.44 crore came back. The purposes, where explained at all, included “maintaining confidentiality of REL’s bank accounts,” “parking funds for court proceedings,” and this is the remarkable one, routing money “without revealing the bank account from which the funds had come.”

That last phrase is worth sitting with. A listed company’s chairman routed its money through his personal account specifically to hide where it originated. When SEBI asked REL about this, REL confirmed it. And then called it operational.

There was also Siddharth Mehta, the chairman’s son, who REL’s own KMPs said had “no role in REL.” During the same period, REL transferred ₹21.25 crore to him. Only ₹5.79 crore came back. The explanation: he was paying company expenses through his credit cards because of “convenience, better pricing, and credit card benefits.”

And then there was Elest Pvt Ltd, a company incorporated by Mr. Rajesh Mehta to manufacture lithium-ion batteries. REL transferred ₹565.88 crore to Elest between 2020 and 2025. ₹350.03 crore came back. The net outflow of ₹215.85 crore remains, in SEBI’s words, “inadequately explained.” The MD and CFO of REL told SEBI in their depositions that they were not aware of most of these transactions.

None of it, not the ₹338 crore to the chairman, not the ₹21 crore to his son, not the ₹566 crore to his battery company appeared as a related party transaction in REL’s annual reports.

The Accounting That Made It Prettier

On top of all this, REL had a habit of using accounting treatment to make its numbers look better than they were. Foreign exchange gains on trade receivables, ₹866.60 crore worth, were stuffed into “Revenue from Operations” instead of recognized separately in profit and loss, as Ind AS 21 requires. This inflated the top line while making the gains look like operational performance.

Interest earned on fixed deposits and mutual funds, ₹204 crore, was similarly included in operating revenue, despite REL’s own accounting policy stating that interest goes to “Other Income,” and despite Ind AS 115 being unambiguous that interest is not revenue from contracts with customers. Intra-group investments of ₹2,501 crore and intra-group payables of ₹1,457 crore remained in the consolidated balance sheet uneliminated, inflating the apparent size of the group. Ind AS 110 requires their removal.

And trade receivables of ₹2,914 crore, money owed to REL, long overdue, from overseas counterparties, were quietly netted against payables through backdated arrangements, informal telephone understandings, and confirmation letters obtained only after SEBI started asking questions. One of the “debtor” entities’ email addresses returned a permanent delivery failure when SEBI tried to communicate with it. Another entity’s commercial license, when checked, belonged to a different company entirely.

When this netting reduced the receivables balance, REL presented the reduction as collections. The investors saw the number come down and felt reassured.

How It Stayed Hidden

Fraud of this scale does not stay hidden by accident. It needs architecture. REL’s architecture had six components.

The Swiss Fortress. By placing the inflated revenues within GGR and Valcambi SA in Switzerland, REL put them beyond the easy reach of Indian regulatory verification. When SEBI asked for customer-wise sales data for these entities, REL invoked Swiss data protection law. SEBI examined the law and found that it protects natural persons, not corporate data and that it explicitly permits disclosure to regulatory authorities. The shield was a bluff. But it worked for years.

The Unaudited Pipeline. This is the technical masterstroke. Valcambi’s real audited accounts showed tiny revenues. GGR’s internal, unaudited “consolidated accounts” showed enormous ones. REL chose GGR’s numbers for its own consolidation. The auditor accepted this because the alternative, admitting it could not verify 97% of the business, would have required a qualified opinion.

The Account less Accounting System. Here is a detail from REL’s own annual report that deserves a full stop of silence after it. The auditors’ report, for both standalone and consolidated accounts, contains this disclosure: “The Company has used accounting software for maintaining its books of account, which does not have a feature of recording audit trail (edit log) facility throughout the year.”

No audit trails. Meaning: entries can be created, modified, backdated, or deleted without any record of who did it, when, or what it replaced. For a company alleged to have fabricated ₹11,487 crore in sales to a stockbroker, the accounting system that has no memory of what was entered or changed is not a coincidence. It is infrastructure.

One Man, Total Control. REL’s MD stated in deposition that he had “no role to play in the day to day operations of REL and its subsidiaries” that everything was handled by Mr. Rajesh Mehta. When one person controls all information flows in an organization, there is no internal check. There are only that person’s choices. And when that same person sits on the Audit Committee that is supposed to oversee the very transactions he is directing, as Mr. Mehta did, attending all 51 Board meetings and all 21 Audit Committee meetings, the committee is not oversight. It is theatre.

Three Different Stories. When SEBI sought customer-wise sales data, REL provided three separate submissions at different times. Each materially contradicted the others. Entities appearing in one were absent from another. Numbers for the same customer varied by tens or hundreds of crores. When challenged, REL asked SEBI to discard all earlier submissions and rely only on the latest. This is a classic forensic technique: if investigators never have a stable dataset, they can never complete a verification. Every submission starts a new audit cycle.

The Receivables That Weren’t. The original complaint, aged receivables, was, in hindsight, both the first symptom and the last line of defense. REL’s standalone balance sheet at March 31, 2025, shows ₹2,643.92 crore in trade receivables. Of that, ₹2,613.23 crore, 98.8% had been outstanding for more than three years. The company’s explanation in the annual report: “Management believes that all receivables are recoverable and no material credit risk exists.” No provision was made. The auditor accepted this. And so, these phantom receivables sat on the balance sheet, classified as good assets, giving the impression of a business that was owed money by real customers who would eventually pay.

The People Who Were Supposed to Stop This

An Audit Report Worth Reading Very Carefully. On August 5, 2025, M/s BSD & Co. signed off on REL’s annual report with a clean opinion unqualified, unmodified, entirely positive on both standalone and consolidated financial statements. I have read that audit report carefully, because it is a public document, and because what it says alongside what it does not say is one of the most instructive things in this entire case.

Let me take you through it. On the consolidated accounts, the auditors disclosed, in an “Other Matters” paragraph that they had not audited the financial statements of REL Singapore or any step-down subsidiary. They had received, instead, “a copy of the consolidated financial statements of the above said entities, approved by the Board of Directors of Rajesh Exports Ltd.” And their opinion “in respect of above entities is solely based on such Board’s approved Financial Statements.”

Solely based on management-prepared documents. For entities that generated 97–99% of consolidated revenues. Then, in the very next paragraph, they issued a clean consolidated opinion. “True and fair view.” This is where audit theory and audit practice pull violently apart. SA 705 the standard that governs modifications to audit opinions exists precisely for situations where auditors cannot obtain sufficient appropriate evidence for material items. The subsidiaries that generated virtually all consolidated revenues were a material item. The auditors had no independent evidence for them. The standard required a qualified opinion or at minimum, an “except for” carve-out. What the investors received instead was a clean opinion with a disclosure buried in the body of the report that almost nobody reads.

Disclosure of an inadequate audit procedure is not a substitute for an adequate audit procedure. Telling investors “we relied only on what management gave us” while still saying “this is true and fair” is a contradiction dressed in professional language.

On Key Audit Matters, the auditor identified precisely none. In the standalone report. In the consolidated report. Zero. Key Audit Matters introduced specifically so that auditors must communicate the areas of greatest complexity and risk were apparently absent from an audit in which:

– 97% of revenues were unverifiable by the auditor’s own admission

– Trade receivables of ₹2,613 crore had been unpaid for over three years, classified as good

– The accounting software had no audit trail

– Related party transactions involved the promoter’s personal accounts

– Ongoing litigation existed with a major bank

If those conditions do not generate a Key Audit Matter, the KAM framework has been rendered meaningless.

On Internal Financial Controls, the auditor issued an unmodified opinion controls “adequate” and “operating effectively.” This, despite having disclosed, in the same report, that the accounting software had no audit trail. I want to be direct about this: these two positions cannot coexist. A company whose accounting software leaves no record of who entered what, when, or what was changed cannot have “adequate” internal financial controls over financial reporting. The entire point of IFC reporting is to tell users whether the numbers they are reading were generated by a system with appropriate safeguards against manipulation. An audit trail is not a feature, it is the minimum. Its absence is a material weakness.

On fraud, CARO 2020’s Clause xi(a) requires auditors to state whether any fraud has been noticed or reported. The auditor stated: none. Months later, SEBI’s investigators, working from many of the same underlying documents prima facie found ₹15 lakh crore in fictitious revenues, ₹11,487 crore in fake trades, and hundreds of crores diverted through personal accounts.

On related party compliance, the auditor certified that the company complied with Sections 177 and 188 of the Companies Act, the provisions governing Audit Committee oversight and related party transactions. SEBI subsequently found that ₹360+ crore had flowed through the promoter’s personal accounts without Audit Committee approval or disclosure.

On management representations, the auditor accepted written confirmations from management that no funds had been layered through intermediaries to ultimate beneficiaries, the standard benami/round-tripping declaration required under Rule 11(e). SEBI found that REL had done exactly this, through Rajesh Mehta’s personal account, Siddharth Mehta’s credit cards, and Elest’s bank accounts.

Each of these clean certifications sat in a document that was publicly available. Each of them told investors: this company is as it appears. Each of them was, prima facie, wrong.

Why Did a Professional Let This Happen?

I want to be fair to the auditors here, because fairness demands it and because the problem is bigger than any one firm. M/s BSD & Co. is a Bengaluru-based practice. They were auditing a company with operations in Switzerland, Singapore, and the UAE jurisdictions where independent verification requires international legal frameworks, foreign-language document review, and coordination mechanisms that a domestic firm may not have access to. The honest answer to this challenge would have been to qualify the consolidated opinion, or to appoint component auditors in each jurisdiction, or to refuse the engagement unless cooperation from subsidiaries was guaranteed.

But here is the structural truth of the Indian audit market: if a firm qualifies its opinion, it risks losing the engagement. If it insists on accessing Swiss subsidiary records and is rebuffed, it must either qualify or resign, both of which create complications for the client. And the client is the one who recommends the auditor’s appointment, sets their fees, and decides whether to continue the relationship. Statutory auditors are formally independent. Economically, they serve at the pleasure of management.

This is not BSD & Co.’s problem alone. It is the profession’s problem. And it is precisely why SEBI’s decision to refer the matter to the National Financial Reporting Authority is appropriate and why the NFRA must be seen to act meaningfully on it.

There is also the question of scale. REL claimed revenues of ₹4+ lakh crore. In context, this is auditing something with the top-line complexity of a major global bank. Most such entities in India are audited by large, multi-city, multi-national firms with the resource base to pursue complex overseas verifications. REL was not. The assignment was beyond what any domestic firm of limited resource base could credibly execute and the appropriate response would have been to say so, formally, in the opinion.

What This Means for You

If You Are an Auditor: This case should disturb you. Not because it is unique, but because the failures it documents are exactly the ones that your professional standards were designed to prevent and they happened anyway. The “Other Matters” paragraph is not an escape. You cannot accept management-prepared financials for 97% of the business and still issue a clean consolidated opinion. SA 705 does not offer a “disclosed-but-clean” option. Either you have sufficient, appropriate evidence or you do not. If you do not, your opinion must say so even if it costs you the engagement.

The absence of an audit trail is not a compliance observation. It is a material weakness that must appear in your IFC report, prominently and clearly, with the consequence that the clean IFC opinion cannot stand alongside it.

Key Audit Matters must be real. When the most significant risk in your audit, unverified revenues constituting virtually the entire consolidated business produces zero KAMs, you have either not identified the risk or chosen not to disclose it. SA 701 makes neither of those acceptable.

Management representations are the last resort, not the first. SA 240 tells you clearly that fraud involves intentional omissions and misrepresentations and that when fraud indicators are present, you cannot simply accept what management tells you. Three-year-old receivables classified as recoverable, symmetric buy-sell entries with zero value addition, revenues from entities you cannot access, these are fraud indicators. Professional skepticism requires you to push past the representation.

And fundamentally: your obligation runs to the investors who rely on your opinion, not to the management that pays your fees. That is the social contract of statutory audit. If you are unwilling to qualify, modify, or withhold your opinion when evidence demands it, if the pressure of client relationship overrides your professional judgment, then you are not providing assurance. You are providing comfort, and there is a crucial difference.

If You Are a Company Director or Audit Committee Member

Governance is not paperwork. REL had an audit committee. It held regular meetings. Minutes were recorded. Yet none of it caught ₹338 crore flowing to the chairman’s personal account, ₹566 crore to his privately held company, or ₹11,487 crore in fictitious trades. How? Because the audit committee included the very person orchestrating the fraud, and everyone else had been excluded from the information needed to raise questions.

The lesson is structural. Audit committees must operate with independence that is not merely nominal. They must have direct, unfiltered access to auditors, without management present. They must receive transaction-level data, not management summaries. They must question, specifically and in writing, every related party transaction, not ratify lists provided by the very executive whose conduct they are supposed to oversee.

For companies with significant overseas subsidiaries: your fiduciary duty as a board member extends to the consolidated group. Accepting board-approved consolidated statements from management without independent corroboration is not governance. It is delegation to the very parties whose conduct requires oversight.

The audit trail requirement is not a technological nuisance. It is a fundamental governance tool. If your company’s accounting software does not record who entered what, when, and what was changed, you have no reliable history of your own financial data. That is not a compliance problem. That is an existential governance problem.

If You Are an Investor

The clean audit opinion on a complex listed company is not a guarantee. It is one professional’s judgment, formed under the conditions and incentives I have described above. It can be wrong. It can be incomplete. And when the company is complex enough that the auditor openly admits it could not access the underlying records of 97% of the business, the opinion is telling you something important: the assurance you think you are receiving is not the assurance being provided.

Read the “Other Matters” paragraphs in audit reports. They are where the important caveats live. Look at KAMs. If a company has significant overseas operations, complex consolidations, and aged receivables and the auditor has identified zero Key Audit Matters, ask why.

Do the basic arithmetic. When REL’s consolidated revenues were 45 times its standalone revenues, and REL itself admitted that Valcambi SA was the only real operating entity, the mismatch was screaming for an explanation. No explanation was offered. The question was never asked loudly enough. Until one shareholder, in a single email, asked it.

Check the receivables ageing schedule. It is in the notes to accounts of almost every listed company’s annual report. REL’s showed ₹2,613 crore outstanding for more than three years, classified as fully recoverable, with no provision. That alone should have prompted a question that the market never formally asked.

When a promoter-controlled company invokes foreign law to avoid disclosing the sales and purchase records of its core subsidiary to India’s securities regulator, as REL did repeatedly, that is not a legal technicality. That is a company telling you it does not intend to let anyone verify its numbers. The appropriate response is not to accept the explanation. It is to treat the refusal as the answer.

The Number Nobody Talks About Enough

Two lakh six thousand nine hundred and forty-two. That is how many shareholders REL had by September 2025. People who had bought shares in this company, based on financial statements certified as true and fair, showing a global gold giant with ₹4+ lakh crore in revenues and who had no idea what was actually in the books.

The market capitalization had peaked at ₹30,365 crore. By April 2, 2026, it had fallen to ₹2,365 crore. The public investor share of this wealth destruction is estimated at approximately ₹12,725 crore.

That is not a statistical abstraction. It is the retirement savings and the investment portfolios and the long-term bets of more than two lakh individual shareholders who trusted that the system designed to protect them, the auditors, the governance mechanisms, the disclosure requirements, was doing its job. It was not.

A Final Thought

Every few years, Indian capital markets produce a fraud of this kind, large enough, brazen enough, and structurally complex enough to briefly shock the system into a conversation about audit quality, corporate governance, and investor protection. Then, gradually, the conversation fades. The regulatory proceedings move slowly. The culpable parties engage their lawyers. New companies list. New investors enter the market. The lessons go unlearned or, more precisely, get learned by some and ignored by others.

What makes REL distinctive is not the scale alone, though ₹15 lakh crore in alleged fictitious revenues is, by any measure, extraordinary. What makes it distinctive is how clearly visible the red flags were, and for how long they were ignored by the people equipped and paid to see them.

The shareholder who sent that March 2024 email to SEBI was not a forensic accountant. They did not have access to Valcambi’s KPMG-audited accounts or GGR’s internal consolidation workings or Affluence’s contract notes. They just looked at the receivables schedule and noticed that the money was not coming in.

Sometimes the most important question in finance is not the sophisticated one. It is the obvious one that nobody wanted to ask.

Why hasn’t this money arrived yet? The answer, it turns out, was that much of the business that generated it never existed.

The writer is a practicing Chartered Accountant in Nepal. All findings referenced in this article are prima facie, drawn from the SEBI Interim Order dated June 3, 2026, and the publicly available Annual Report of Rajesh Exports Limited for FY 2024-25. Legal proceedings are ongoing. Views are personal.