What's all this Financial Transaction Tax fuss about?
EU leaders are at it once again; putting Financial Transaction Taxes (FTTs or Tobin Taxes as they are also called) back on the agenda while they are forced on the back foot by the unresolved Euro crisis. At a time when citizens are losing faith in the ability of our leaders to solve the crisis, talking about FTTs, which remain heavily popular with the public, almost always earns political brownie points.
But what can FTTs really achieve? And is the current approach, presented by the European Commission, designed to succeed? If not, should be abandon the idea altogether or is there another tax design that will work better?
One thing is for sure FTTs will not change the world, nor democratize global finance. Nor will they raise the hundreds of billions of Euros of revenue that is sometimes attributed to them. But, approached sensibly, a well-designed and flexible regime for financial transaction taxes can deliver a lot of benefits.
- Raise substantial revenue (in the tens of billions not the hundreds of billions)
- Deliver a highly progressive form of revenue compared to alternative sources such as VATs
- Help tackle the problems brought about by high frequency trading
- Help engender incentives to track transaction level information
- Help reduce some of the spurious liquidity that afflicts most financial markets
- Help create incentives and a policy tool to tackle systemic risk and procyclical behaviour
- Help tackle some of the excessive short-termism that afflicts financial markets
We have dealt with these issues in previous publications that can be found Here (EU do it) Here (A FIT Proposal) Here (FTTs: Tools for Progressive Taxation and Improving Market Behaviour) Here in our testimony to the German Bundestag and Here (Tobin or not Tobin?)
The European Commission’s proposal for a Financial Transaction Tax
The European Commission has proposed a Pan-EU Financial Transaction Tax regime. While there are parts of the legislation that are good and its scope is ambitious, it suffers from a serious design flaw that can and must be corrected.
What is good about the proposal is that it aims for a very broad scope. It applies to capital markets, money markets, derivatives, UCITS, AIFs and securitizations – namely most financial instruments of note. It is also good in capturing all forms of trading venues and also has a broad definition of trade including a transfer of risk.
Under the EC proposal, the tax becomes chargeable when a financial institution ‘resident’ in the EU engages in a financial transaction. So a tax will be payable if an EU institution trades American financial assets but not if an American institution trades European assets. This is a fatal flaw and this proposal will likely fail unless it is universally applied, not just in the EU but also internationally across all current and potential financial centres. This is the model the failed Swedish tax followed (Scroll down to the section about the UK vs. the Swedish Model). Much of the trading of non-EU originated financial assets will migrate out of the EU and a substantial proportion of EU-financial asset trading will also migrate, or at least have strong incentives to.
Another problem with this model is that it rewards countries with tax revenue that is proportional not to their GDPs but to the percentage of financial trading that takes place in these countries. On current form, 60%-80% of the total trading across all instruments is concentrated in Germany and the UK which makes for a highly inequitable distribution of FTT revenue. It makes it politically very difficult for these countries to agree to share this revenue more equitably across the EU.
A better design
Levying the tax on the basis of the source of the original financial asset makes much more sense. Essentially a transfer of ownership of the asset or any legal rights will not be recognized unless a ‘stamp duty’ has been paid – the UK model in the section below. This makes avoidance much more difficult and the tax is payable no matter where you trade so minimises incentives and possibilities for relocating trading. It also means that since your legal right is only as good as the title of the previous owner in the chain – this creates strong incentives for monitoring compliance amongst market makers and those seeking to avoid the tax will be shunned by market actors.
Another advantage of this is that the tax revenue will be allocable roughly in proportion to the GDP of the country as the stock of financial assets is far more uniformly distributed than trading in these assets. This ‘UK-inspired’ design will also allow a unilateral/coalition-of-willing approach to work if the pan-EU/G-20 approach fails. This generates incentives for country to make them work.
How to impose an FTT and how not to impose an FTT?
The best point of reference would be to look at the FTT regime in the UK. This is relevant, as London is one of the most important financial centres in the world; home to top investment companies, banks and other financial institutions.
The UK has had a 0.5% (one of the highest in the world) tax on all share transactions in UK incorporated companies. It is chargeable whether the transaction takes place in the UK or overseas, and whether either party is resident in the UK or not. Even in 2000-2001, the tax raised £4.5 billion in revenue and had the lowest cost of collection of all major taxes (0.09% as opposed to an average collection cost over all major financial taxes of 1.11%). Now it raises more. This shows that the FTT regime in the UK is thriving and raises significant revenues (2000-2001 revenues were 14% of all corporation tax collected in the UK) at a time when some of the other major economies have dismantled their FTT regimes.
The Stamp Duty on shares in the UK was introduced in 1963 at a rate of 1% and in 1974 was revised upwards to 2%. It was reduced in 1984 to 1% and then again in 1986 to the current level of 0.5%. This level of 0.5% on the market value of a security is payable by the buyer. There is at present no tax on derivative securities but the transfer of a share to a nominee owner (to be re-issued as non-taxable depository receipts) attracts a triple rate of 1.5%.
Instead of focussing on both successful and unsuccessful examples of FTT regimes, opponents of FTTs have chosen to talk just about the Swedish experience, which was a failure by most measures. In 1984, the Swedish government introduced a tax of 0.5% on both the sale and purchase of equities. This was payable only in the case of a Swedish Brokerage service being used. The tax went through many changes of rate but the basic structure remained the same. In 1989 a similar albeit smaller tax was imposed on the trade of fixed income securities, including government debt and associated derivatives, such as interest rate futures and options.
As a result of these taxes, a significant amount of trading in Swedish stocks migrated overseas. At the extreme only 23% of the trading in Ericsson, Sweden’s most actively traded stock, took place in Sweden in 1989. The average for the market as a whole was 57%. There is little evidence of a fall in the overall size of the market; most trading just migrated overseas in order to evade the taxes imposed on Swedish brokers. In the fixed income markets the effects were even more dramatic; bond trading fell by 85% and bills trading fell by 20% within a week of the tax being introduced. This was accompanied by a sharp increase in trading in untaxed fixed income instruments such as debentures, variable rate notes, forward rate agreements and swaps all of which served as close substitutes for the taxed instruments.
The Swedish FTT experiment came to an end in 1991 when all the transaction taxes were abolished. The markets in both fixed income and equities soon recovered to pre-tax levels.
What lessons can be drawn from these two regimes?
The major difference between the Swedish and the UK FTT regime was that the Swedish tax was a domestic tax on international capital whereas the UK tax is an internationally applied tax on domestically registered companies.
The lesson here for successful implementation of the FTT in the EU is that the tax needs to be imposed on the trade of all country-originated financial assets, no matter where in the world they are traded and no matter how the counterparties trade. Because if HM Treasury only taxes trades in the UK, similarly to Sweden just taxing their domestic brokers to levy their FTT, then it would be easy to legally evade through offshore migration. For a successful design of the FTT, therefore, the tax has to apply to British (country) originated assets themselves, and not to a jurisdiction or a channel of trade. If it operates in this way it would be next to impossible to evade through any legal channels.
The Swedish fixed income market saw a rapid migration of trades from taxed instruments to untaxed instruments that could act as close substitutes. This shows that it is important to cover all close substitute instruments when imposing an FTT. This being said, though equity derivatives in the UK are not taxed, the UK market has only seen a small migration of trades from the taxed share market to the untaxed derivative market. This shows that derivative instruments are not perfect substitutes for the underlying financial instrument. This is also partly as derivatives get less preferential treatment in the UK for accounting and reserve purposes.
So yes, we should aim for the tax to have as broad a scope as possible, but not get too worried.
The next lesson can be drawn by looking at the fact that no mass migration of shares to depository receipts has been seen in the UK market. The purchase of shares for the purpose of issuing non-taxable depository receipts, the creation of non-taxable bearer instruments and for the purpose of transferring into clearing services is all taxed at 1.5%, three times the standard tax rate. This high initial cost of transfer and the fact that these instruments are not perfect substitutes for owned shares in the UK have both helped deter a mass flight from shares into these untaxed instruments.
This means that even if we do allow an offshore migration of EU financial assets, an imposition of high exit costs would minimise any potential tax evasion.
The original stamp duty in the UK was payable whenever there was a transfer of the paper share certificate from one owner to the other. However, when an electronic settlements system, called CREST, was introduced in 1996, a new tax called the stamp duty reserve tax (also simply called stamp duty in this text) was introduced on the electronic transfer of ownership. This made collection much easier and less expensive; only 0.02% cost of collection as compared to the 1.11% average for other major financial taxes. This shows that the recent trend of financial markets towards increased electronic trading, centralized settlement and data repositories would make it much easier and cheaper for governments to administer and collect Financial Transaction Taxes.
Another relevant lesson that can be drawn from the UK stamp duty experience is that international co-operation on FTTs is possible. As the stamp duty is payable no matter where shares in UK incorporated companies trade, there is a need for international co-operation which has been forthcoming. Another good example of international co-operation is that between the Irish and the British governments who share the stamp duty revenue collected on the trade of Irish shares in the UK.
The UK stamp duty structure provides an excellent framework on which to develop a successful FTT regime. The fact is that the UK mobilizes upwards of Euro 5 billion in revenue every year through the stamp duty. As well, the London Stock Exchange remains the second largest exchange in the world despite having a stamp duty which is not payable in most competing exchanges. Both points are testament to the enduring success of the UK stamp duty regime.
Another attractive feature of this regime is that it creates a first mover advantage. For example, 60% or more of UK stamp duty revenue comes from non-UK tax payers. Any other country that implements a similar regime will enjoy a similar first mover advantge. Once a critical mass of countries adopt the tax, this relatuve advatge, the fear of losing which is the main reason the UK is opposed to the tax, will dissipate away. By contrast, the Swedish model suffers from a first mover disadvantage where trades migrate abroad.
The recently proposed EMIR legislation on the reporting, clearing and settlement of derivatives transactions and MIFID on the reporting and trading of all financial instruments help make the infrastructure much more conducive to implementing financial transaction taxes. They will make any form of avoidance and evasion much harder.
With these, and a change of the proposed European Commission design, the EU authorities would have all the three elements they need to levy financial transaction taxes 1) legitimacy 2) information and 3) an incentive structure that rewards compliance and makes avoidance very hard and costly.
If EU leaders are indeed serious about implementing financial transaction taxes, they need to change the proposed design. And yes, while these will generate revenue and help improve the functioning of the financial system, they should not be oversold to the public.