Who killed Saab Automobile?

An obitury of an automotive icon… After briefly pointing readers to the Holweg/Oliver report on April 3, 2015, we return to it now with a fuller reading of the figures, decisions, and missed opportunities behind Saab Automobile’s collapse.

On December 19, 2011, Saab Automobile AB was declared bankrupt in the Vänersborg district court, ending 62 years of car production and putting roughly 3,700 Trollhättan workers out of a job. Within weeks, Matthias Holweg of Cambridge Judge Business School and Nick Oliver of the University of Edinburgh Business School published “Who Killed Saab Automobile? Obituary of an Automotive Icon,” a detailed report that tried to settle, with production figures and financial statements rather than forum arguments, who actually bore responsibility.

Most SaabPlanet readers know the broad outline of the final year: the supplier stoppages in spring 2011, the Pang Da and Youngman memoranda, the GM veto, the December press conference. What the Holweg/Oliver report adds is the documentary backbone – a year-by-year financial table from 2000 to 2010, a production chart going back to 1947, and a sober comparison with the GM portfolio and with MG Rover’s 2005 collapse. This piece walks through the report’s main arguments and adds the context that a Saab-literate reader will want.

The number that explains everything: 150,000 units

The report’s central technical point is simple, and most enthusiasts have heard a version of it before, but the report quantifies it precisely. Saab’s annual production never exceeded 134,000 units, a peak reached in 1987. Under GM, output stabilized around 120,000 to 130,000 units a year through the late 1990s and 2000s – essentially Porsche-scale volumes. But Saab was priced and positioned against Audi and BMW, brands that by the time of Saab’s collapse were each building over a million cars a year and benefiting from platform sharing across the Volkswagen and BMW group portfolios.

The report frames this as “Porsche volumes with Audi prices,” and argues this combination is structurally unsustainable. Porsche survives on volumes similar to Saab’s because its customers tolerate genuinely high margins. Saab’s customers – loyal, but not wealthy in the Porsche sense, and not numerous in the Audi sense – could support neither the margin nor the scale. The report’s appendix table for 2000–2007 shows after-tax losses every year except 2001, with losses peaking at -4,767 MSEK in 2003 against revenue of roughly 23,613 MSEK – an 18 to 20 percent margin in the wrong direction, year after year.

Study is very detailed, with lots of background data and the recommendation is to read, for all true fans of Saab brand.

Twenty years of GM, and the Vectra problem

Ask around any Saab forum or comment thread and one verdict comes up more than any other: GM killed Saab. The reader comments under SaabPlanet’s own 2015 version of this article make the point bluntly – one reader writes simply “The GM killed Saab,” another argues GM “ruined” the brand the way Ford nearly did with Jaguar and Land Rover, and a third pins it on “American companies” not knowing how to handle European ones. It’s an understandable instinct: GM owned Saab for the 20 years in which the brand’s reputation for distinctiveness eroded most visibly, and it was GM’s 2011 veto that ended the last realistic Chinese rescue.

The Holweg/Oliver report takes that sentiment seriously but doesn’t let it stand as the full answer – the authors explicitly call the “brand mismanagement by GM” reading “too simplistic.” GM owned 50 percent of Saab from 1990, took full ownership in 2000 after Investor AB sold its remaining stake for $125 million (a quarter of what GM had paid for its first half a decade earlier), and supported the brand through losses in 17 of those 20 years. To GM’s credit, the report notes, it kept funding a brand that lost money in nearly every year of its ownership – hardly the behavior of a parent simply trying to strip-mine a name.

What the report does fault GM for is something more specific than neglect: a structural mismatch. GM Europe was configured as a high-volume producer of mid-range cars, and that configuration sat uncomfortably against Saab’s premium, low-volume positioning. Saab, for its part, resisted the platform-sharing and standardization that might have made the relationship work financially – so the friction ran in both directions.

But the report is equally clear about what went wrong technically. The 1994 900, based on GM’s Vectra platform, was poorly received – the report cites a “Which?” magazine assessment describing reliability problems, and notes the new 900 was reported to have done poorly in Swedish crash testing. The 1997 9-5, developed on GM2900/Vectra-derived architecture, carried the same credibility problem with purists who felt the cars no longer had the body stiffness of earlier-generation Saabs. By 2003, Saab’s engineering department was merged into GM’s operations at Rüsselsheim, costing the company 1,300 engineers and designers – a detail the report flags as directly contradicting later claims by Spyker that Trollhättan retained full vehicle-development capability.

The 2005 arrival of the 9-2X (a rebadged Subaru Impreza, quickly nicknamed the “Saabaru”) and the 9-7X (a Chevrolet Trailblazer with an eight-cylinder engine) compounded the brand dilution. The report’s framing is that GM tried two strategies at once – broadening Saab’s appeal for volume, and cutting costs through shared GM platforms – and “suffered the negative consequences of both without realizing the benefits of either.”

Spyker’s 93 days, and what Muller actually inherited

The report gives Victor Muller’s Spyker takeover a fairer hearing than much of the contemporary press did, while still concluding it was always a long shot. Muller’s own account – an email to GM director Bob Lutz that led to “93 days, working 20-hours a day” – is reproduced, along with his admission that he had “no idea” what the acquisition would entail and had never run anything close to Saab’s scale (Spyker had built 43 cars in the year before the deal; Saab was capable of over 100,000).

What Spyker inherited in February 2010 was a seven-year-old 9-3, an outgoing 9-5 at the end of its life, the still-unreleased new 9-5, developed on GM’s Epsilon II architecture shared with the Opel Insignia, and the 9-4X, a Cadillac SRX-based SUV to be built in Mexico. The new 9-5 launched in June 2010 to what the report calls a “positive, but somewhat lukewarm” press reception – reviewers liked the car but couldn’t shake comparisons to its Opel Insignia sibling. Sales forecasts told the story: 50,000–60,000 units projected for 2010 in March, cut to 45,000 by August, cut again to 30,000 – 35,000 by late October. Actual 2010 production came in at 32,048 units – barely half the original target, though still above 2009’s 20,791.

The China endgame, and GM’s veto

The report’s most detailed section covers the frantic search for Chinese partners in 2011: Hawtai’s $223 million deal collapsed in May when Chinese regulatory approval failed to materialize; the Pang Da memorandum brought in modest cash that briefly restarted production on May 27; the Youngman manufacturing agreement followed in June. By late October, a deal had Pang Da and Youngman buying Saab outright for $134 million (€100 million), with pledges to inject $855 million.

GM’s refusal closed the only rescue route that still looked financially large enough to restart Saab. On November 8, 2011, GM stated it could not support any ownership change that might damage its relationships in China – specifically its joint venture with SAIC, which shares GM technology embedded in Saab products and which objected to the deal in the Chinese press. A modified structure routing 49.9 percent of Saab through a Chinese bank was rejected on December 8. The report’s reading is unambiguous: it made no commercial sense for GM to hand Saab’s GM-derived technology to a potential SAIC competitor in GM’s second-largest market.

MG Rover, and the question of whether history repeats

The report draws a deliberate parallel with MG Rover’s 2005 collapse: both companies were disposed of by larger parents (BMW for Rover, GM for Saab), both landed with entrepreneurial owners who believed they could succeed where bigger players had failed, both had one newer model that the press liked but buyers didn’t buy in numbers (the Rover 75 and the new 9-5), and both ended with Chinese capital promised and ultimately not delivered in the form first announced. Rover at least “survives” in name through SAIC’s “Roewe” brand; the report is candid that a similar fate – production know-how absorbed elsewhere, the original marque gone – was the realistic outcome for what remained of Saab’s tooling and IP.

So who did the report blame?

The report’s actual verdict is less satisfying than a single name, and that’s the point. It explicitly clears Saab’s unions (who filed for bankruptcy reluctantly, to protect unpaid wages), administrator Guy Lofalk, and – more surprisingly – Victor Muller, whom the authors describe as naive but not the cause. GM is described as having “little choice” but to push integration, given its configuration as a high-volume European producer.

The report’s deepest answer points back decades: Saab’s missed opportunity was never becoming for GM what Audi became for Volkswagen – a fully integrated brand with a distinct premium identity. Both GM and Saab share that failure, GM for not understanding the brand and Saab for resisting integration that might have provided economies of scale. The authors call this Saab’s own contradiction: to survive economically it needed to grow, but growth would have diluted the very identity that made it Saab.

Closing

Sixty-two years and 4.5 million cars ended not with a single decision but with a structural mismatch that nobody – not Investor, not GM, not Spyker, not Pang Da and Youngman – ever solved: a brand built on doing things differently, operating in an industry that increasingly rewarded doing things the same way, at scale. The report is worth reading in full for anyone who still has a 9-5 or a 900 in the garage and wants the financial and historical record rather than the forum version.