For a British start-up in the physical sciences to establish clear global leadership in its market and a solid presence on the stock exchange is a rare thing indeed. It has taken more than 40 years for Oxford Instruments — the University of Oxford's very first start-up company — to get where it is today, with 1,200 employees and annual sales of £156 million (US$276 million). But having made a name for itself as perhaps the world's leading supplier of high-specification magnets for research equipment, the company has suffered a few setbacks.
The firm was established in 1959 by a physicist, Martin Wood, who had just built Europe's first superconducting magnet. It grew steadily and was listed on the London Stock Exchange in 1983. But analysts say that its recent woes illustrate the pitfalls for companies whose scientific ambitions get in the way of profitability.
"They were always regarded as an extremely good business, but the results began to decline in the late 1990s," says one London-based analyst who has followed the company closely but declined to be identified. "Oxford Instruments was battling against the limits of its capabilities. Customers would decide on a number of specification changes as the system was being put together."
The problems originated largely in the company's superconductivity division, which produces superconducting magnets and low-temperature bench-top cryostats. It also manufactures superconducting wire, magnets for magnetic resonance imaging (MRI) and systems for a technique called ion cyclotron resonance, which determines the chemical composition of molecules.
The division's clients range from the healthcare industry, which uses MRI for medical scans, to researchers at major laboratories, who use it to study materials and molecules.
Particle physicists, meanwhile, have been among the most demanding customers for Oxford Instruments' high-specification magnets, says Michael Cuthbert, sales manager for physical sciences at the company's US offices in Concord, Massachusetts.
The division's difficulties began after it took on a number of particularly demanding contracts on a fixed-cost basis, the London-based analyst says. "Because they were very challenging contracts, it was very appealing to take them on for the kudos of doing them," she explains.
The main projects involved building magnets for a physics experiment at CERN, the European particle physics laboratory in Geneva, and for an MRI machine at the Pacific Northwest National Laboratory in Richland, Washington state.
CERN was planning an experiment called Compass, which sought to study the structure of hadrons using an existing synchrotron. The project incorporated several smaller studies, including one looking at the structure of protons and neutrons by scattering high-energy muons from them. This required the spins of all the particles to be polarized, using the field produced by a high-performance magnet.
Quenched field
L. GUIRAUD/CERN
Damp squib: a magnet supplied by Oxford Instruments for CERN's Compass experiment struggled to produce a high enough field strength.
Compass ordered the magnet in 1996, but physicists were never able to get it working at the required field strength, according to a source close to CERN. "The problem was that the magnet kept on quenching — the magnetic field could not be held for any length of time," said another source familiar with the project, referring to the release of liquid helium and subsequent loss of the field.
Ultimately, the CERN source said, the company gave up and made a substantial indemnity payment to the project. A number of former Oxford Instruments employees, working independently, eventually managed to get the equipment to work, the source added.
Around the same time, Oxford Instruments faced significant delays in delivering a high-specification MRI machine to the Pacific Northwest's Environmental Molecular Sciences Laboratory. The machine, which was envisaged as the most powerful of its type in the world, was intended to help scientists study the structures of larger molecules and watch the interaction of molecules and cells at high resolutions (see Nature 383, 375; 1996).
But it wasn't delivered until six years later, in 2002, and even then the laboratory held back a $1.2-million final payment to be made once it was fully operational. Several current and former employees of the laboratory contacted by Nature declined to speak about the contract. An Oxford Instruments spokeswoman also declined to comment on the CERN and Pacific Northwest contracts.
"In trying to be customer-focused and really challenging the technology, while perhaps not understanding all of the risks, we probably too often pushed to keep the customer satisfied," Cuthbert says.
As these contracts unfolded, the operating profit of the superconducting division fell from £10 million in 1997 to just over £1 million in 1999, followed by losses of £5.3 million in 2000.
In the wake of these problems, the division has tried to improve the way it manages exposure to risk, Cuthbert says. "We have regular management engineering risk reviews and assess all enquiries for risk well in advance of quotes," he said. "It's no good for the final customer or for any business to take on very high-risk projects and not succeed."
Over the past few years, the company has sought to rein in some of its technical ambitions and focus on the bottom line. But after staging something of a recovery, it reported some more problems last year. Quality issues had arisen at a plant making superconducting wire in the United States, and there had been a loss of business from a key client, Varian Medical Systems of Palo Alto, California. In March, it appointed a new chief executive, Jonathan Flint — a trained physicist with management experience in other British high-tech firms.
A brand-new generation of magnets is a key part of the company's future growth strategy, says Adrian Philips, an analyst at the UK investment company Williams de Broe in Birmingham. The superconducting division has two particularly promising magnets in the pipeline, both of which look to extend the limits of its existing technology.
Molecular distinctions
A 950-megahertz, 22–23-tesla magnet will be delivered in October for use in MRI systems, Cuthbert says, allowing researchers to make more detailed distinctions between molecules. The division is also developing an 800-megahertz magnet that is actively shielded to restrain any stray field, improving safety and leading to more efficient use of lab space.
At the superconductivity division, analysts say, there will be less emphasis on ambitious, one-off projects to meet special customer requirements and more on standard products that play to the company's undoubted strengths.
Since Flint's appointment, the company's share price has remained relatively strong, and investors have high hopes for the future, says the London-based financial analyst who spoke anonymously to Nature.
"The company has always had undoubted technical expertise, but it has never been translated into consistent, profitable growth," she adds. "The arrival of a new chief executive from outside gives the hope that he will convert the technology into earnings growth, and deliver returns to shareholders."