商业欺诈的十二种解构：从造假技术到系统性崩塌

When you're waiting to find out whether you're seriously ill, a faster answer means something. Elizabeth Holmes built Theranos around that hope. She promised a wide range of blood tests from just a few drops, with results delivered quickly and cheaply. It was an easy future to want. Rupert Murdoch invested, and Walgreens began offering the testing service in its stores. Patients could actually walk in and use it.

But in 2015, journalist John Carreyrou began reporting that the technology wasn't doing what people had been led to believe. Theranos was running much of its testing on other companies' machines. Its own device could perform only a limited range of tests and had serious accuracy problems. Presumably, the revolution was just within the branding. Using another company's machine wasn't itself the fraud; Holmes was selling investors a story about capabilities her technology didn't have. Trial evidence showed that she knew about its limitations while concealing them. In 2022, a jury convicted her of defrauding investors; later that year, she was sentenced to more than 11 years in prison. People had wanted a less frightening way to find out what was wrong with them; that hope had helped sell a breakthrough the company couldn't deliver.

Volkswagen offered drivers an attractive compromise: a diesel car with good performance that also met strict pollution standards. And there was evidence the cars passed their emissions tests. Then researchers from West Virginia University tested them somewhere rather inconvenient—on the road, working for the International Council on Clean Transportation. They found a large gap between the laboratory results and what came out of the cars during actual driving. Regulators began asking why the same vehicles behaved so differently; the answer was in the software. Some Volkswagen diesel cars could recognize the conditions of an emissions test. When they detected one, their pollution controls worked with full effectiveness. During ordinary driving, the software changed how the engine operated, reducing those controls and allowing much more pollution. The cars had learned to behave themselves when someone important was watching. That meant the test was measuring a performance the software had deliberately arranged for the occasion. Drivers were buying cars whose clean credentials didn't reflect their normal behavior. Volkswagen pleaded guilty to related criminal charges in the United States in 2017; buybacks, repairs, and financial penalties followed the scandal. A company selling clean diesel had built cars that could tell when to put on a good impression.

In 1998, Waste Management corrected its accounts, wiping out roughly 1.7 billion dollars of previously reported profits. A rubbish collection business had managed to create a very expensive mess of its own. Part of the explanation was sitting in its truck fleet. A truck is expensive, but a company normally spreads that cost over the years it expects to use it. Imagine one costs 100,000 dollars, lasts five years, and has no resale value. That means recording 20,000 dollars a year as an expense. Now claim it lasts 10 years; the annual expense falls to 10,000. So the reported profit rises. You haven't collected more rubbish or spent less money; you've changed an estimate. Waste Management repeatedly extended equipment lifespans and exaggerated what trucks and containers would be worth at the end without proper support. Both changes reduced the costs in its accounts. Unfortunately, telling a truck it has another five years left does very little for the gearbox. Alongside other accounting tricks, those estimates helped create false profits. The company eventually had to correct them, and its former finance chief was found liable for securities fraud. Shareholders had been judging the business using profits that its ordinary rubbish operations hadn't actually produced.

In 2002, Cynthia Cooper and her internal audit team were examining WorldCom's accounts when they encountered enormous entries that needed explaining. Senior finance executives resisted the investigation. Cooper's team kept looking. What they found helped expose a telecom company that had been making ordinary bills look like investments. WorldCom paid other companies to carry calls over their networks. Those payments were running costs, much like a phone bill. They should have reduced the profit reported for the period they belonged to. Instead, WorldCom moved billions of dollars of costs into the category used for things that provide value over several years. That let it spread the expense into the future, making the current profit look better. By that logic, paying your electricity bill makes you a proud investor in the national grid. A real investment in equipment might benefit a company for years; these were current charges for access to other companies' networks. Changing the label didn't buy it anything extra. The audit team found that the entries lacked proper support. WorldCom disclosed the problem in June 2002 and filed for bankruptcy the following month. The company had been able to report stronger results because its accounts postponed costs that the business had already incurred. Once the auditors followed those costs, the apparent improvement no longer held up.

HealthSouth ran hospitals and other healthcare facilities. But when its profits fell short of what investors expected, some of its finance staff had a solution that involved very little healthcare. They simply added the missing earnings to the accounts. That solved one problem and created another: if the company had earned all this extra money, where was it? According to America's securities regulator, staff supported the invented earnings with things the company claimed to own, but didn't. Some entries described equipment at individual facilities. These were assets an investor might reasonably expect a healthcare company to have. The equipment was remarkably cheap to maintain, largely because it didn't exist. A single enormous purchase might attract an auditor's attention, so staff divided false additions into smaller amounts that fell below the thresholds used to select entries for questioning. If an auditor challenged an item, an invoice could be altered to support it. The fraud became public in 2003, and several finance employees pleaded guilty. HealthSouth later agreed to a 100 million dollar civil penalty. The deception had worked backwards from the result management wanted: instead of reporting what the business had earned, staff invented the earnings and then built the paperwork underneath them. Scattered across the accounts, small false entries had added up to a very large lie.

Satyam's employees were doing real software work for real customers. From outside, there was a business you could point to, with accounts showing profits and money in the bank. Then its chairman sent a letter that made those accounts almost impossible to believe. In January 2009, B. Ramalinga Raju admitted that much of the reported cash didn't exist. He described a gap between the company's actual profits and the profits it had announced growing over several years. So how had the accounts kept adding up? According to America's securities regulator, senior managers created thousands of false invoices. These were bills for work that hadn't generated the sales being claimed. Bogus bank statements then made it appear that customers had paid. The imaginary customers were apparently excellent at settling their bills. Each false document supported another: the invoice explained where the profit came from, and the bank statement explained where the money went. Together, they made the business look more successful than it was. The Indian government replaced the board, and a search began for a new controlling shareholder. The people writing the software still had a real business to keep running; they now had to do it after their chairman had admitted that a substantial part of its financial success existed only on paper.

You could walk into a Luckin Coffee shop and buy a perfectly real cup of coffee. That made its growth story easier to believe. Investors could see a business serving customers, and the sales figures suggested it was doing rather well. Then questions raised during its annual audit led to an investigation. In 2020, the company's own investigators found that its 2019 sales revenue had been inflated by about 300 million dollars. Some of those apparent customers weren't ordinary customers at all. Money for fake transactions had come through businesses linked to employees or other people connected to Luckin. A customer who exists only in a spreadsheet is unlikely to complain about the coffee. But inventing sales created another problem: more coffee sales should usually mean more spending on supplies and other costs. A sudden jump in sales without the supporting expenses could attract attention. So, according to America's securities regulator, employees inflated expenses too. They also altered bank and accounting records to support the deception. The fake growth needed an increasingly elaborate set of documents to make it look normal. Luckin corrected its results, removed senior executives, and agreed to a regulatory settlement. The shops hadn't been imaginary; their existence had made it much easier to believe sales figures that included business they hadn't done.

Enron had a troubled power project in Brazil. Selling part of it helped the energy company report a profit and keep the project's debt out of its main accounts. To an investor, that could look like management had dealt with a problem. But look at who was buying: as regulators later described, the buyer was a business controlled by Andrew Fastow, Enron's own finance chief. There was an unwritten agreement that made the sale rather less reassuring: Enron would buy the investment back at a guaranteed profit for the buyer. The buyer had an excellent safety net; unfortunately, Enron shareholders were holding it. Normally, selling an investment means the buyer takes on the risk that it might disappoint. That secret promise meant Enron hadn't simply walked away from the problem; it still had an obligation to the business on the other side of the deal. This was one of several arrangements used to make Enron's results look healthier than they were. Investors could see the apparent benefit of the transaction without being told the full cost of the promises behind it. Enron filed for bankruptcy in 2001, and its shares became virtually worthless; Fastow later pleaded guilty to fraud-related conspiracies. The reassuring accounts had concealed obligations that selling the project hadn't made disappear.

Michael Woodford was chief executive of Olympus, a company known for cameras and medical equipment. In 2011, he questioned whether some of its past deals had been properly explained; that October, the board dismissed him. Being chief executive evidently didn't come with unlimited access to explanations. The following month, Olympus admitted it had been hiding losses on investments stretching back to the 1990s. The deals Woodford had questioned were part of a much older financial problem: when investments lost value, Olympus moved them into outside funds instead of acknowledging the losses in its accounts. That kept the damage out of view, but someone still had to pay for the arrangements holding those losses. Later, Olympus bought other companies. Parts of the purchase payments, including unusually large adviser fees, were routed through funds to help cover the old losses. That changed what the spending appeared to mean: someone looking at a company purchase could think Olympus was paying to expand its business. Part of the money was actually helping clear up investment losses from years before. An investigation traced the connections, and Olympus had to confront the history those deals had concealed. The camera and medical businesses were real, so were the old losses, however many transactions had been placed between them and the shareholders.

Buying shares in Adelphia meant investing in a cable television business. You might expect its money to help run the network; you probably wouldn't expect it to help finance a film made by the founder's family. John Rigas had built the company, and his family retained control after outside investors bought shares. The family also owned separate private businesses; their money and Adelphia's money flowed through a shared cash management system that blurred a boundary investors had every reason to care about. Funds went towards family debts and other private interests, including that film; the transfers weren't properly disclosed as loans or compensation. Shareholders had acquired a sideline in film financing without the inconvenience of being asked. There was a larger problem with borrowing: accounting entries placed debt with family-controlled businesses, even though Adelphia remained responsible for paying it. The company looked less exposed than it really was. In March 2002, Adelphia revealed about 2.2 billion dollars of previously unreported debts and other obligations; it filed for bankruptcy that June. John Rigas and his son Timothy were later convicted of securities fraud. The family's control had let them direct the company's money; it hadn't made that money theirs to use without telling the other people who owned the business.

Parmalat sold milk around the world, and its accounts offered lenders something almost as comforting: a large reserve of money. One company in the group, called Bonlat, supposedly held almost 4 billion euros in cash and investments at Bank of America. A lender could look at that reserve and think Parmalat had a substantial cushion if business went badly; there were even letters appearing to confirm the balance. Then in December 2003, the company acknowledged the billions of euros of reported assets weren't there. The supposedly reassuring account didn't exist either. It was an excellent bank balance, provided nobody involved the bank. The confirmation letters had been forged inside Parmalat's finance department; what looked like an independent check on the company's claims had actually been manufactured by the company itself. Bonlat's figures fed into Parmalat's overall accounts, so the invented fortune helped make the entire group appear financially stronger, even to someone who never looked at Bonlat separately. Parmalat collapsed and entered a restructuring process. Lenders and investors had assessed the company using money it couldn't produce when the story came apart. The trick had gone beyond inventing a number: it had supplied a convincing-looking answer to the very question that should have exposed it—had the bank confirmed the money was there?

As financial markets fell apart in 2008, Bernie Madoff's clients could still look at statements showing investment gains. Some had already withdrawn money successfully; that was a powerful reason to believe the figures. But when enough clients asked for their money back, Madoff couldn't pay. The balances on their statements were far larger than the cash available to meet their requests. The investment trades behind those balances were largely fictitious; staff created records of trades that hadn't happened, sometimes choosing them after they could see which prices would produce the desired profits. Picking winning shares is considerably easier once they've already won. That explains the statements, but how had clients been able to withdraw real money? Their deposits had gone into bank accounts; when someone wanted to withdraw, the business paid them using money deposited by other clients. A payment could arrive exactly as promised without any investment profit having been earned. That is the central deception of a Ponzi scheme. The financial crisis brought withdrawal demands that the remaining cash couldn't cover; Madoff pleaded guilty in 2009. For his clients, the painful discovery was that a successful withdrawal hadn't proved their investments existed; it had only proved there was still enough money from other people to pay them.

In November 2022, customers rushed to take their money out of FTX. The cryptocurrency trading platform couldn't meet their requests. For people who had been told their deposits were protected, that raised an urgent question: where had the money gone? Founder Sam Bankman-Fried had assured customers that their funds were kept separate from the company's own money. He also controlled a trading business called Alameda Research, which he said received no special treatment. That would have been reassuring had it been true. Changes to FTX's code gave Alameda privileges ordinary customers didn't have, allowing it to withdraw effectively unlimited amounts of cryptocurrency. Billions of dollars of customer funds were channeled to Alameda; money went towards repaying its lenders, making investments, buying property, and funding political contributions. Alameda was treated just like any other customer, apart from its extraordinary access to everyone else's money. Depositing funds on a trading platform hadn't meant agreeing to finance the founder's other business; customers had been exposed to risks that directly contradicted his assurances. FTX filed for bankruptcy that November; Bankman-Fried was convicted of fraud and conspiracy and sentenced to 25 years in prison in 2024. The cryptocurrency could be complicated; the broken promise wasn't. People had entrusted FTX with their money, and he had secretly used it for something else.

The unsettling thing is how much of this looked convincing. You could buy the coffee, drive the car, or receive an investment payment. None of that proved the rest of the story was true. And when someone asked for evidence, there was often a document ready to provide it. The paperwork could be excellent; reality was rather less cooperative.