Just over two years ago, I took a train from Berlin to Wolfsburg, to visit a senior executive at Volkswagen.

From the station platform the famous chimneys of VW’s vast car plant loomed up. Inside the gate, the company’s vehicles were everywhere, catching the sun like a metallic sea. It crossed my mind that it would take a fair bit of courage for an employee at the plant to drive a non-VW car.

However, the cars were just a distraction. I was there to talk to the executive about hedging transactions. The business of manufacturing and selling cars globally created huge cash flows and financing pressures with risks that needed to be managed.

From my perspective at the time, Volkswagen was just a customer (albeit a very big one) of the financial services industry ““ it bought derivatives linked to currencies, interest rates and commodities from major dealers. Since the financial crisis, the executive had grown concerned about VW’s counterparty credit exposure to banks and was in the midst of diversifying it.

Seeing the world through his eyes I shared his concern about banks. When my contact told me that Volkswagen had forced its derivative dealers to accept one-way collateral agreements (so VW received collateral but didn’t have to post any in return), I saw this as prudential risk management. I didn’t imagine that banks might ever have a reason to worry about the credit risk of Volkswagen. I certainly didn’t think about VW as a source of risk to my health and that of millions of others.

It wasn’t that I didn’t have the education to see it in that way.

In my twenties, I attended Harvard graduate school where a professor, Steve Wofsy, explained to me the photochemistry of nitrous oxide and ozone. In a fascinating chemical detective story, ‘NOx’ was fingered by Wofsy and others as a contributing villain to stratospheric ozone depletion and acid rain. Although the focus back then was on industrial pollution, I certainly was aware that diesel engines were the other major source.

All that got relegated to the back of my mind after I left Harvard and became a financial journalist. Instead of photochemistry, I spent more than a decade learning how financial innovation created its own ‘pollutant’ – hidden leverage ““ that ultimately led to bank meltdowns and government bailouts. I saw how the industry lobbied regulators, who allowed them to calculate their own ‘pollution level’ and use models to give a misleading impression of health.

The Volkswagen scandal contains similar elements. Like banks, the auto industry lobbied regulators to be permitted to use standardised emission tests. Like banks, it was tempting to construct cars with engine computers that gamed the tests in the same way that risk models gamed capital requirements. Desperate for profit, VW (and perhaps other car companies) succumbed to the temptation.

As with Basel 2 banking rules, regulation inadvertently played a part in this fiasco. Committed to fighting global warming, European governments veered a decade ago towards favouring lower-carbon diesel over petrol engines. But there was a catch ““ dirty diesels emitted particulate matter (PM) that was linked to lung cancer and other health problems. So the industry proffered ‘clean diesels’ which supposedly burned off the soot in catalytic filters. Like Alan Greenspan arguing that derivatives made banking safer, regulators lapped it up.

Again there was a catch. To burn off the soot at low enough temperatures, the converters produced nitrogen dioxide (a component of NOx) as a by-product, particularly in the zippy turbo-diesels that car companies were pushing to consumers. Aware of this, regulators progressively tightened the rules, reducing the allowable emissions of NOx and PM measured in standard tests.

Meanwhile, a research team led by chemist David Carslaw at Kings College London began developing roadside tests that measured the actual concentrations of NOx in urban locations. They found something shocking: while official government emissions figures showed steady declines in NOx, the roadside data from London showed NO2 increasing¹⁾the chart in this post uses Carslaw’s roadside data for greater London as a whole, in which dramatic NO2 increases for inner London are offset by declines in the suburbs ““ a smoking gun pointing to diesel vehicles. This pattern would be replicated across Europe where similar tests were done.

What bothered Carslaw and his colleagues was how reduced emission levels in standardised tests resulting from improved regulations was diverging further and further from reality. Without realising it, they had detected the impact of VW ‘defeat devices’. They published their work in 2011 and delivered a report to the UK government.

Their warning was ignored, and nothing happened until three years later, when Volkswagen’s push to convert the US to diesels prompted new research. A contradiction was found between the mandated test results and actual emissions performance of VW cars leading to the company’s devastating admission to regulators. Meanwhile, separate research has established the long-term health impact of NO2 exposure, equivalent to 6,000 additional deaths in London alone according to a KCL study.

It’s hard to believe that Volkswagen wasn’t aware of this research, which makes its persistence in selling defeat-rigged cars so damning. Like UBS, Goldman, Citigroup and other banks selling similarly-rigged collateralised debt obligations on the cusp of the financial crisis, will we now see internal VW emails referring to ‘shitty deals‘, “dogshit” or ‘vomit‘?

I have some sympathy for my treasurer contact. He will have his work cut out now, reducing hedges against cash flows that VW will no longer have, while banks buy default protection on his company to hedge their VW credit exposure (since they don’t get collateral). I have less sympathy for European governments that rolled over in the face of the evidence, and left it to the Americans to do something. In the auto sector as in banking, regulatory capture is poison.