Could a transactions tax be good for capitalism?
Many of those who work in banks and financial institutions equate support for a European financial transactions tax with a vindictive desire to see the City of London shrink.
But the argument, accepted by the Treasury (although not in quite so emotive terms), that if the tax were implemented in Europe alone, there would be an exodus of important firms to homes outside the EU - which would be particularly damaging to the City as Europe's leading financial centre - may not be quite as clear cut as it seems.
An EU financial transactions tax would be both desirable and feasible, argues an influential economist, Avinash Persaud, who has worked for a fair number of leading banks in his time and is currently chairman of Elara Capital.
Slightly embarrassingly for HM Treasury, his argument is based on the long-term impact of the UK's Stamp Duty Reserve Tax, which levies 0.5% on transactions in UK shares.
Although the London Stock Exchange has long campaigned against this stamp duty, the levy has been around in its current form for 25 years - and for longer in other incarnations - and hasn't been associated with the mass departure of equity trading away from the UK.
In fact, the London Stock Exchange has been remarkably successful in persuading international companies to list their shares in London: it has probably been the most successful stock market in the world, in that respect.
What's more, stamp duty raises an invaluable £3bn a year for the exchequer. And the Office for Budget Responsibility expects the take from the tax to rise by a third over the next three years.
So tax avoidance does not appear to have undermined this very British transaction tax.
As for the proposed European tax, it would be at a rate of 0.1% on shares and bonds, a fifth of Britain's stamp duty levy, and at 0.01% on derivative transactions, so - in theory - would create less of an incentive to emigrate from London for avoidance purposes than stamp duty has done.
In Persaud's view, the relative success of stamp duty is in its design. It is levied on any London listed shares, regardless of the nationality of those trading the shares. The test for liability to stamp duty is not the residence of the trader or investor.
The important point, he says, is that a transfer of shares is not legally enforceable if the tax hasn't been paid and the deal "stamped". So even if investors and banks relocate to Singapore or Geneva, if they trade in UK shares they have to pay this duty - and the consequence is that foreign residents pay 40% of British stamp duty.
Persaud would therefore argue that a similar rule should apply to the EU's transactions tax: so long as the transactions were in shares, or bonds or derivatives relating to money raised by companies based in the European Union, then the tax should apply.
On this scenario, there would be no point in huge banks, investors and traders quitting London, because it would not allow them to escape the tax.
There would only be acute damage to the City of London from a new transactions tax if huge companies, such as BP, or Unilever, or Vodafone stopped issuing bonds and shares in Europe or ceased to hedge their financial transactions in Europe.
The prospects of enormous "real" companies relocating to avoid the transactions tax or doing all their financial business in other jurisdictions seems pretty remote.
That said, a transactions tax might reduce the volume of transactions - especially in derivatives - designed for purely speculative purposes by increasing their cost.
It is not clear that would be such a terrible thing, since there is some evidence those deals increase irrational exuberance and manic depression in markets, to the detriment of businesses trying to finance themselves, and are also devices for extracting excessive fees from more gullible businesses (see my earlier post for more on this).
With its marginal increase in the cost of financial trading, a transaction tax would disproportionately hurt those trading in securities on the slimmest of profit margins.
As a result, there would be a reduction in so-called high-frequency trading, viz massive automatic trading in shares and bonds carried out by computers that run algorithms looking for price discrepancies or trends.
Would that be so disastrous?
Less high-frequency trading could mean that overshoots and undershoots in markets would become less extreme, in that the new hegemony of the machines in stock markets seems to have been associated with a sharp increase in the volatility of markets (the extreme case being Wall Street's notorious flash crash).
If that's so, then a financial transaction tax would actually increase the efficiency of the allocation of capital: it would improve the functioning of capitalism, rather than undermining it.