Revenue too soon bends time. Revenue that isn’t invents substance. Schilit distinguishes four techniques: transactions lacking economic substance (round-trips), transactions lacking arm's length (related parties), non-revenue receipts booked as revenue, and revenue inflated by gross-up (booking what you merely pass through). In every one, the "customer," the "sale," or the "amount" isn't what it appears.
(a) Round-tripping: your own money, returning in a costume
The move. Money leaves the company as a loan, an advance, an investment, or a prepayment to a "vendor," travels through one or more entities, and comes back as payment for "sales." Revenue is conjured and cash appears to collect. DSO can even look healthy. There are variants. Two companies swap identical "sales" to inflate both top lines (the energy-trading "wash trades" of 2001–02). A company funds its distributor-of-convenience. Or capex paid to a promoter-linked EPC contractor returns as orders or deposits.
The entries. Both legs, and learn them cold:
Leg 1 (money out): Dr Loan/advance to entity X / Cr Cash
Leg 2 (money back): Dr AR — entity X / Cr Revenue
Dr Cash / Cr AR — entity XAdvertisement
Consolidated view: cash is unchanged, an intangible "loan to X" now sits where cash was, and revenue plus profit have been created from nothing. If leg 1 is disguised as capex, the sediment is CWIP instead of loans.
The residue. Loans/advances/investments in unconsolidated or promoter-linked entities grow alongside revenue. Receivables from related parties appear, and so do standalone-vs-consolidated gaps. CFO sits chronically below EBITDA because net cash never really arrives. And the interest income on the "cash" is oddly missing.
The tests. Read the related-party note (Ind AS 24 / ASC 850) line by line for counterparty names, amounts, terms and balances. Compare loans & advances growth with revenue growth. Use India's structural disclosures in promoter related party transactions (Rule 11(e) conduit representations, CARO loans-to-related-parties clauses). Then run the cash-yield test. An interest income ÷ average cash & deposits far below prevailing deposit rates says the cash isn't there or isn't the company's.
On the record. Luckin Coffee (2020). An anonymous short report published by Muddy Waters (January 2020), built on thousands of hours of store-traffic video across ~1,000 store-days and tens of thousands of customer receipts, alleged fabricated store sales. In April 2020 Luckin's own board investigation confirmed that its COO and others had fabricated roughly RMB 2.2bn (~$300m) of 2019 revenue, routed partly through related shell companies, with expenses inflated to match. Nasdaq delisted it, and the SEC settled for $180m (December 2020). Two lessons last. Primary observation (counting customers) beat audited numbers. And the fabrication sat between the IPO and a follow-on raise, the incentive window. There's an epilogue worth knowing, too. Under new ownership and management the underlying business later recovered. The fraud was in the reporting, not the coffee.
(b) Related-party sales: and Satyam, the India anchor, worked properly
The move. Sell to entities you control or that your promoter controls, at inflated prices, with no intention of collection, or with a side agreement to unwind. At the limit, skip the counterparty entirely and forge the invoices.
The entries. Dr AR (related party) / Cr Revenue. When collection never comes, the receivable ages, gets provisioned in a bad year (the big bath), or is "settled" against some new advance (round-trip).
Satyam Computer Services (2009): the case every Indian analyst must be able to narrate. On 7 January 2009, chairman B. Ramalinga Raju confessed in a letter to the board that Satyam's accounts had been falsified for years. Per the letter, the balance sheet carried ₹5,040 crore of non-existent cash and bank balances (against ₹5,361 crore shown) and ₹376 crore of accrued interest that did not exist. It also carried an understated liability of ₹1,230 crore (funds Raju had arranged personally and routed in), and debtors were overstated by ₹490 crore.
Look at the September 2008 quarter alone. Reported revenue of ₹2,700 crore was actually ₹2,112 crore. Reported operating margin of ₹649 crore (24% of revenue) was actually ₹61 crore (3%). Over 90% of the quarter's reported operating profit was fictional. It was built on thousands of fake invoices (investigators later counted ~7,500) fed into a doctored invoicing system, and supported by forged bank confirmations that the auditors (PwC affiliates) accepted.
There was a near-miss tell. In December 2008, Raju attempted to have Satyam acquire Maytas Properties and Maytas Infra, his family's companies, for ~$1.6bn. That deal would have converted the fictional cash into real (family) assets and closed the hole. Institutional investors revolted within hours and the deal died. The confession followed within three weeks. A special CBI court convicted Raju, his brother, and eight others in April 2015. The government-supervised rescue auction handed the company to Tech Mahindra.
The detection lessons, because they generalize. There are four. (1) The cash-yield test. Satyam's reported cash pile earned implausibly little interest, and analysts who divided interest income by average deposits found a yield no real treasury would tolerate. (2) Margins vs peers. Satyam reported near-tier-1 margins with a weaker business mix than TCS/Infosys, and peer benchmarking flags that straight away. (3) Revenue per employee vs peers sat oddly, because fake revenue needs no engineers. (4) The promoter's behavior: a low and falling promoter stake (~8%), pledged shares, and a sudden related-party acquisition attempt.
When a company hoards "cash" yet tries to spend it only on the promoter's own assets, believe the behavior, not the balance.
Also on the record. At DHFL, the administrator's transaction auditor (Grant Thornton) reported ~₹14,000 crore of fictitious retail loans kept in a separate "Bandra Books" ledger. Hundreds of thousands of fake borrower accounts generated fictional interest income there, alongside large fund flows to promoter-linked entities. The CBI's 2022 chargesheet (~₹34,600 crore of alleged consortium fraud) was the largest bank-fraud case filed in India to that date. DHFL was resolved under IBC to Piramal (2021).
Cox & Kings defaulted on commercial paper in mid-2019, weeks after reporting a large cash balance. Lender-commissioned forensic reviews alleged fictitious sales and debtors, and the promoter was later arrested in a related money-laundering case. Treat both stories as "per investigators' findings/allegations," since litigation continues. But the patterns are the ones you've just been studying: fake granular assets, cash that can't be touched, promoter-linked flows.
(c) Non-revenue receipts booked as revenue
The move. Cash arrives that isn't earned from customers (a vendor rebate, an insurance recovery, a loan dressed up, a grant) and lands in the revenue line. Dr Cash / Cr Revenue where the credit belongs to a liability, a cost offset, or other income. Watch for it wherever "other operating income" swells. India's Schedule III P&L makes this line visible, so use it. There's more in operating vs non-operating income and promoter related party transactions.
(d) Grossing up agency revenue: size without substance
The move. An agent, one who arranges a sale someone else fulfills, must book only its commission (net). A principal controls the good or service before transfer and books the full ticket (gross). The tell-tale indicators are primary fulfillment responsibility, inventory risk and pricing discretion (ASC 606-10-55 / IFRS 15 B34–B38 / Ind AS 115). The shenanigan is an agent booking gross. Revenue multiplies and profit doesn't change by a rupee, but growth-hungry markets pay for the top line anyway.
The entries. Agent truth: Dr Cash 100 / Cr Revenue (commission) 7 / Cr Payable to principal 93. Gross-up lie: Dr Cash 100 / Cr Revenue 100 and Dr Cost of sales 93 / Cr Payable 93.
The residue. You see a huge revenue line with a wafer of gross margin, and a balance sheet far too small for the "business" (no inventory, no fulfillment assets). Sometimes revenue triples on a "policy reassessment."
The tests. Look for gross margin near zero on a large line, and read the principal-vs-agent policy note. Ask the three control questions yourself: who holds inventory risk, who sets price, who is on the hook for fulfillment? Then recompute the company on net revenue and see if the multiple still makes sense.
On the record. Groupon's IPO filing (2011) initially reported the entire deal value as revenue. After SEC review it restated to net, and 2010 revenue fell from ~$713m to ~$313m at a stroke. It later revised Q4 2011 results over refund reserves, disclosing a material weakness.
Vakrangee (2018), in India, was a technology/kendra-network company whose revenue was swollen by low-margin bullion and jewellery trading. Its auditor (a PwC affiliate) resigned citing inability to obtain information about that business. The stock fell more than 90% from its January 2018 peak, and SEBI later penalized disclosure violations. Always ask: why is a franchise-tech company trading gold? Revenue mix that exists to be big is the gross-up in spirit even when the trades are real.
What happened at Wirecard?
The terminal case of revenue that isn’t: Wirecard (2020). Germany's DAX-listed payments champion reported half its revenue and profits from "third-party acquiring" partners in Asia. Then came years of Financial Times reporting (Dan McCrum, from 2015) and a KPMG special audit that could not verify the TPA revenues. EY refused to sign the 2019 accounts. €1.9bn of cash in Philippine escrow accounts, roughly equal to all the net profit the company had ever reported, likely never existed. Insolvency followed within a week (June 2020). The COO fled and remains a fugitive, and the CEO was arrested and put on trial.
Wirecard is revenue that isn’t at end-state. Fabricated revenue requires fabricated assets (the sediment law). The asset finally chosen was cash, the one auditors are most expected to verify, and with forged confirmations they still didn't. Same plug as Satyam, eleven years and one continent apart.