For students of perverse incentives created by tax, it is a bonanza week. Apple has raised $17bn in bonds to buy back shares, despite having $130bn sitting in cash overseas, and Pfizer wants to turn itself into a UK-domiciled company by acquiring AstraZeneca for £60bn.

So great are the distortions created by the US corporate tax rules that it was better for Apple to borrow than to use its abundant cash. Meanwhile, the UK has succeeded so well with its plan to become a respectable kind of offshore tax haven that it has turned its pharmaceuticals industry into a takeover target.

General Electric’s bid to use some of the cash it has amassed overseas to buy the energy unit of Alstom of France for $13.5bn is a mere footnote; corporate taxes are just as high in France as the US, and GE will remain an American company. But fiscal engineering is becoming as important a factor in multinational mergers as corporate logic.

There is surely something wrong with this. Investors can of course benefit from big mergers involving ‘tax inversion’, whereby US companies break the surly bonds of the Internal Revenue Service. But this does not create stronger enterprises or bring wider economic benefits. The notion of ‘shareholder value’ becomes distorted when it means ‘tax breaks’.

“If tax distorts business decisions, shareholders may be better off but someone else will be worse off. The global economy will be smaller and incomes lower,” says Michael Devereux, professor of business taxation at Oxford university. Such is the size of the potential corporate tax arbitrage between the US and the rest of the world at the moment, the risk is exactly that.

The Pfizer bid, which follows pharma groups such as Valeant and Actavis becoming Canadian and Irish companies respectively through mergers with Biovail and Warner Chilcott, is a stark illustration of the mathematics. Pfizer might want to buy AstraZeneca anyway to improve its drugs pipeline but the transaction stands or falls by tax.

Pfizer’s bid would not work financially if it remained a US company because it would shift AstraZeneca’s domicile from the UK, where the corporation tax rate is 21pc, to the US, where it is up to 35pc. Instead, Pfizer wants to turn itself into a lower-taxed UK company (although its headquarters and stock-market listing would remain in New York).

The second tax advantage is that Pfizer would rid itself of the main perversity of the US tax regime - that in theory, the IRS taxes the global income of corporations but in practice it does not touch offshore earnings. These are only taxed if they are brought home to invest or distribute to shareholders.

As a result, cash-rich US companies such as Apple, Pfizer and Microsoft do not repatriate cash but leave it offshore, sometimes making a provision for its future taxation (Pfizer is holding a $20bn tax liability on its balance sheet). This creates huge war chests of cash, which US companies are tempted to spend on buying foreign businesses.

Pfizer cannot simply turn itself, for example, into an Irish company to release this cash because the US changed the law in 2004 to curb tax inversion. Instead, it needs to reverse into a foreign company, leaving its investors with less than 80pc of the combined equity. AstraZeneca both provides a good use for Pfizer’s cash reserve and prevents the offshore earnings problem from recurring.

All this would make AstraZeneca attractive enough but there is a third tax benefit - the ‘patent box’, which was introduced by George Osborne, the UK chancellor, to match schemes in the Netherlands and other countries. It permits companies that research and develop new products in the UK to pay only 10pc tax on the income.

By becoming UK-domiciled, Pfizer could channel many of its new drugs through the UK patent regime. Even if it made the bulk of the revenues by selling these drugs in the US, it could exploit the 10pc tax rate by arranging for its US sales arm to pay large royalties for the UK intellectual property.

Indeed, Pfizer is eager to discover whether it could go further than this and place some of the drugs it is already developing into the patent box if it acquires AstraZeneca. Given that Ian Read, its chief executive, is a Scottish-born chartered accountant, one expects a keen focus on all the potential tax benefits.

Mr Read is doing his job for his investors and a merger might work on industrial terms, even ignoring the tax benefits (although big mergers carry big risks). But the enormous gulf between the US and UK corporate tax regimes means that shareholder value is being defined in a strange way.

There is no reason for this wave of fiscally engineered mergers to subside. Merck & Co, Pfizer’s rival, has plenty of offshore cash - about $57bn, according to Audit Analytics, the research group. If Pfizer manages to acquire AstraZeneca, why would Merck not run the numbers on GlaxoSmithKline? Having lobbied for a patent box, the UK drugs industry might vanish into it.

This would not be a tragedy if the scientific expertise and the jobs remain in the UK. The identity of multinationals is more fluid than in the past - most of their revenues often come from outside their home territory and their senior executives are of different nationalities. Nonetheless, it is telling that Pfizer’s head office will stay in New York.

Companies put together purely for tax reasons - with no industrial or cultural logic behind them - will be fragile. Of course, no company will admit to it when it suggests a deal. But as long as the financial benefits of tax arbitrage remain so great, so does the incentive.