In business, as in life, breaking up is hard to do. Yet, that is precisely what many companies are doing today - known euphemistically as "restructuring".

Restructuring is often seen as defeat, a last resort, proof that all else has failed. Restructerers do not become corporate heroes. Mr Bill Gates's determination to keep his Microsoft empire intact may be driven as much by psychological instincts as business reasons.

In contrast, adrenaline-inducing, spectacular acquisitions are still regarded as the route to stardom for executives, though we know that takeovers have a dismal success rate, at least for acquiring shareholders.

Yet, despite its lack of popularity with executives, restructuring can benefit shareholders. Unlike the arithmetic of acquisitions, which claims synergistic benefits (2 + 2 = 5), restructuring arithmetic poses the equation 4 - 1 = 5.

Restructuring can take a variety of forms. One is demerger or splitting the company into constituent parts. Another is to divest some parts of the company. In the UK, BT is attempting to manage its colossal debt by doing both. It is demerging into two parts, BT Wireless and the remainder to be known as Future BT. Meanwhile, it will also dispose of some Future BT businesses, such as Yell, its yellow pages unit.

Divestitures of business units can take the form of sales to outside buyers, or to a management buyout (MBO), or spin-offs whereby the divested business becomes a quoted company in its own right. In this regard, the mobile phone company Orange has had an interesting history. Starting as an independent company, it was bought by Mannesmann and sold off to France Telecom by Vodafone when it acquired Mannesmann.

Eventually, France Telecom folded Orange into its own mobile business and spun it off in an IPO. The Orange name was retained as having a more dynamic customer friendly image than Itineris, France Telecom's former mobile subsidiary. A recent high profile MBO is the sale of Go, the budget airline launched by British Airways in 1999, to its management team, financed by the venture capital firm 3i.

Of course, restructuring is often synonymous with downsizing, itself a euphemism for laying off people. In really serious cases, the chief executive is often a casualty, as Mr Fran Rooney of Baltimore Technologies recently discovered.

Why restructure? As with acquisitions, restructuring should not be an end in itself. It should be a means to achieve strategic goals. Otherwise, it is a path to self-inflicted oblivion, such as we are witnessing in Eircom.

Research results on restructuring have been consistent on two points.

Firstly, restructuring is helpful if it is used to focus businesses, demerging units that do not go together, or divesting stray units that drain resources and attention from the company's core business.

Secondly, restructuring is not the final step; it must be accompanied by building and development of the remaining business.

Outright demerger is a dramatic event. In Britain, a noteworthy instance was the demerger of ICI in 1993. The strategic logic was compelling. The company had consisted of two main divisions - chemicals and pharmaceuticals, separated by a divide in technology, managerial demands and core competences. Chemicals were high volume, commodity, capital intensive businesses, with common needs in engineering and project management. Meanwhile, pharmaceuticals was geared towards differentiated, intensive R&D, high added-value products.

The scope for synergy and mutual support existed almost entirely within each cluster, and not across them. Thus, the demerger was able to create two smaller focused companies, each with a head office that was attuned and aligned to the strategy of its company.

Another compelling rationale for the ICI demerger was that it gave Zeneca, the ensuing pharmaceutical company, the flexibility, speed and higher stock market valuation to make deals with other companies. Thus, it could compensate for its relatively small size. In fact, Zeneca eventually merged with Astra of Sweden. Arguably this would not have been possible had Zeneca not been a freestanding company. Meanwhile, to be less cyclical and capital intensive, the ICI chemical rump has been restructuring itself to become a speciality chemicals company.

In the process, it has disposed of more than 40 of its traditional industrial chemicals units, while purchasing specialty chemical assets, most notably Unilever's for £4.9 billion sterling in 1997 - as Unilever itself restructures and refocuses on core businesses.

The demerger process itself is potentially fraught. Like divorce, demerger can be messy. It can involve real or imagined treachery, quarrelling over the division of the assets (not to speak of the children). The break-up of assets can be seen as a zero sum game with winners and losers, especially if the two sides are going to be competitors. All sides lay claim to what is most treasured - the best people, customers, property, valuable intangibles such as brands. A prime example is the way Arthur Andersen retained the right to the brand name Andersen Consulting, when its consulting arm broke away to form what eventually became Accenture. Of course, no one wants the debts or the unattractive assets.

The more integrated the assets, the harder to divide them up. In the case of ICI, the company was already operating as two largely separate independent divisions. Therefore, dividing the spoils was not contentious. Moreover, to ease the process, both the chairman and chief executive of the company were solidly behind the demerger.

Yet, for all its exemplary restructuring behaviour, it is doubtful that ICI would have split, had it not been forced by a threatened takeover by Hanson Trust (itself demerged and refocused in 1996). It was feared, at the time, that Hanson would break up ICI into many pieces and randomly sell off some of them.

It took a huge psychological shift for ICI, a traditional British company (as its name, Imperial Chemical Industries denotes), to break itself up. In its glory days, ICI was a symbol of British scientific progress, held together by technology, vertical integration and economies of scale.

In Germany, Bayer, like ICI, a chemical and pharmaceutical giant, and longstanding symbol of national pride, has been under pressure to separate its drugs and chemicals units. However, the company is resisting such a move. Its monolithic structure and the fact that its board executives have spent an average of 27 years at Bayer make any radical restructuring anathema, even if it would put the sum-of-the-parts value of Bayer at an estimated 20 to 25 per cent above its current valuation.

The various businesses are closely interwoven, even if offering few synergies. The lack of focus means that Bayer does not have the critical mass in research, marketing, and distribution to capitalise on the potential offered by promising new drug discoveries in its pharmaceutical business.

Given the modest size of Irish companies, instances of demergers are rare as small-cap companies are not of interest to shareholders.

This logic did not deter James Crean, which thought it could reverse its dismal performance by splitting into two businesses in 1999 - a print and paper business, renamed Oakhill, and a food business retaining the Crean name. However, in neither case, was there any considered basis on which to develop a sound company, in terms of size or competitive advantage. Thus, we have the spectacle of Crean shareholders, whose shares were valued at €1.03 at demerger, seeing a suspension of the trading in Crean shares in May at 18 cents. Last month, Mr Ray McLoughlin, chairman and CEO of Crean, agreed to buy the company at 27 cents a share through a vehicle, Monset. Meanwhile, Oakhill shares, which sold off its packaging arm, languish at 20 to 25 cents amidst profit warnings.

While the Crean case is not a warning against restructuring, it confirms that it cannot, in itself, turn a sow's ear into a silk purse.