No subject is more certain to raise the hackles of traders in financial institutions than a financial transaction tax, however small. Measured for unleashing a torrent of abuse, it even edges out, in the City of London, the question of continuing to be part of the European Union. The traditional banker’s response, something I initially bought into, is to adopt a rather superior air that suggests that the protagonists do not understand the realities of finance. We would, in patronizing tones, remark that it would be a wonderful tax if only it were feasible. But lamentably, finance is now conducted in cyberspace, and cannot be tracked down and taxed.

My views changed when I became a bank director, jointly liable for the bank’s compliance with anti-money laundering and counterterrorism financing rules. I quickly realized that banks are expected to know where the funds behind every transaction come from and go to. They can and must track all their transactions. Failure to comply leaves banks and their directors vulnerable to substantial fines. Between 2007 and 2014, cumulative fines and legal settlements on banks for contravention of rules on controls were in excess of $300 billion.¹ I began listening more discerningly to the arguments my banker colleagues were using against financial transaction taxes and found them both factually incorrect and disingenuous.

The Cassandras tell us that even a small tax, say 0.1 percent of the value of the purchase of a financial security as proposed by the European Commission, would usher in a nuclear winter for financial markets. It seems odd that the value added of our financial centers—and their highly paid bankers—is called into question and billions of dollars of business would drain away were we to impose the tiniest of transaction taxes. Is all the cleverness, superior expertise and connectivity worth less than ten basis points? If financial services are supposedly as internationally substitutable as the decision of whether to pull into this gas station or one across the street, why are the bankers not also on minimum wage?

It is worth remembering that the economic and market impact of a transaction tax would be no greater than any other transaction cost. Consumers of finance already absorb costs that in total exceed the proposed transaction taxes and there is little industry disaffectation. These costs include broker commissions; trading spreads; the price impact of trading, clearing, settlement, and exchange fees; transaction-related research; risk management costs; and other trade administration costs. Traders and fund managers may make these “transaction” costs opaque so they are more easily passed on to consumers either directly or indirectly through lower returns.² In the United States, home to a competitive investment industry, transaction costs were estimated to range from 1.15 percent to 1.44 percent of assets under management per annum between 1995 and 2006.³ If a 0.1 percent turnover tax will cause tumbleweeds to roll across a deserted Wall Street or City of London, then we should be on high alert about the effect of these existing charges. Total transaction costs are annually equivalent to more than ten times the incidence of the proposed European Union Financial Transaction Tax and sometimes more.⁴

Recent financial crises, and a career in the most liquid financial markets, has taught me that while low transaction costs are generally good, zero costs may ironically be bad. Systemic dangers lurk where there is no hindrance to activities that, through rapid velocity, give the impression of great citadels of value. When, for example, transaction costs are low, banks are incentivized to engage in a large amount of offsetting transactions. They earn commissions on the gross but are only exposed to the net. In quiet times, it is the net that is vital. However, in a crisis where everyone exits simultaneously and fear rages, many counterparties appear to be bust it is the gross exposures that will overwhelm the system. In earlier chapters, I discussed the Lehman Brothers default. At the time of its bankruptcy, when markets were in turmoil and credit counterparty risk⁵ was high, Lehman seemed to be facing a $400 billion loss on its gross credit derivative exposures. This itself generated new fears throughout the industry. Once governments had stepped in to underwrite the financial sector, and counterparty credit risk had abated, many of the contracts netted out. The eventual credit loss could end up being less than $6 billion.

Interest in new financial taxes has recently been revived in a number of different places for different reasons. In Europe, after the GFC turned private loss into public debt, campaigners are keen on the revenue-raising potential of transaction taxes, one-off bonus taxes, and mandatory self-insurance funds. In the EU, eleven member states have agreed to enhanced cooperation in placing a 0.1 percent tax on share transactions and a 0.01 percent tax on derivative transactions. They hope to collect up to $25 billion per annum.

In the United States, Michael Lewis’s Flash Boys, a tale of the abuse of ordinary investors by their high-frequency cousins, has garnered interest in the consumer protection potential of charges or taxes on the churning of portfolios. Countries such as China, India, and Brazil have experimented with financial transaction taxes or other short-term capital inflow taxes. They view these measures as a way of calming overheated financial markets. When markets have been calmed, they have lowered the tax.

In their zeal to oppose these new proposals traders have suggested complicated off-balance-sheet transactions that might evade the tax. This reveals both their cleverness and their complete disregard for their status as licensed, regulated entities. Supervisors require capital to be set aside for activities that create risks. Given the size of fines, bank shareholders should be deeply concerned when bankers show this lapse in recognition of their regulatory obligations. However, little is gained by debating the virtues of a policy that cannot be implemented. We begin by looking at the broad feasibility of financial transaction taxes in a world of unwilling participants. Having considered what is possible, we turn to issues of liquidity and stability. Risk-sensitive capital adequacy requirements, mandatory reserve, and insurance requirements have a similar economic impact to a tax and we have already covered these mechanisms in previous chapters. We use a similar analytical framework to consider the systemic implications of financial transaction taxes. My conclusion on the desirability of financial transaction taxes is contrary to the current opinion among financial market participants and regulators. But it is simple and based on our previous analysis of systemic liquidity. Because some of the ideas I raise tend to be dismissed out of hand with false specificity, we look at the practical challenges of using fiscal instruments to regulate finance at a greater level of detail than we have devoted to other instruments.

The Feasibility of Financial Transaction Taxes

We can clearly see the persuasive power of repetition in the case of financial transaction taxes. The trading industry’s stock response is that unless transaction taxes have both global coverage and enforcement they will be sidestepped. The industry repeats this maxim ad nauseam until it becomes the first thing non-experts spout. Given the obvious difficulty of getting over two hundred countries and territories to agree on anything, most also assume that financial transaction taxes must be the preserve of abstract theory. Yet national stamp duties on legal transactions are the most ancient, least avoided and toughest of all taxes to evade. There is no other major tax with a higher compliance rate by non-residents. In Europe, stamp duties stretch back at least to the middle of the sixth century.⁶ In the UK, following an earlier Dutch example, stamp duties were first established on June 28, 1694, to help finance the war against France. Almost three hundred years on, the UK’s 1986 Finance Act, among other things, lowered the rate of the stamp duty on share transactions to 0.5 percent and extended it to paperless transactions.⁷ The act also introduced a penalty rate of 1.5 percent for deposits into depository-receipt, clearance, or nominee accounts, as these schemes could lead to tax avoidance—an approach we return to later. This most ancient of taxes has not stopped the UK from being the home of the fourth-largest stock exchange by market capitalization—worth $3.4 trillion in 2013. The London Stock Exchange is one of the oldest⁸ and most international with a listing of three thousand companies from seventy different countries.

The UK’s opposition to the European Commission’s proposal for a financial transaction tax has been blunted by the UK’s own experience—as having one of the oldest and most successful examples of a financial transaction tax. In order to sharpen its opposition to the EU proposal, and despite a large budget deficit, the UK’s Conservative-led coalition government tried to show its dislike of the tax by granting additional exemptions to the UK stamp tax for share transactions.⁹ Prior to that, the tax collected was over €6 billion (or $8 billion) per annum, or approximately 0.8 percent of all UK tax revenues or 9.9% of UK Corporation Tax receipts.¹⁰

Financial transaction taxes are often seen as ensuring that the financial industry—where value-added taxes (VAT) are not imposed¹¹—makes a more equitable contribution to public finance.¹² More than 30 countries already collect approximately $38 billion per year through stamp duties on financial transactions.¹³ Other types of financial transaction taxes yield further revenue. Interestingly, many in the United States who passionately rail against the feasibility of financial transaction taxes are unaware that since its formation in 1934, the US SEC has been financed by a small financial transaction tax. The so-called Section 31 fee is paid by all the national securities exchanges based on the volume of securities sold on their markets. Over the decades, the fees have been raised and trimmed without warranting a murmur. The fee currently amounts to $18.40 per million dollars.

Stamp taxes on share and nonbearer bond transactions are near impossible to avoid because the legal title of these securities has to be registered. Transfer of title is not legally binding unless the certificate (electronic or otherwise) has been stamped indicating that taxes have been paid. Relocating to a non-tax jurisdiction does not avoid the tax if you want to have legal ownership of a share registered in a tax jurisdiction. It is estimated that more than 50 percent of those people who pay the UK tax on share transactions are non-UK residents, the highest proportion of any major UK tax.¹⁴ The UK collects the tax mainly through CREST—the paperless electronic-settlement and share-registration system administered by Euroclear in Brussels. As of 2014, Belgium did not itself have a financial transaction tax so this tax is being collected in a non-tax jurisdiction.¹⁵

There are different ways in which a financial transaction tax liability can arise. The stamp duty reflects the “issuance principle” where the tax arises not as a result of where the share is traded or the purchaser resident. Rather the key is where the security is issued or deemed to have been issued. A non-resident who chooses not to pay the tax can do so, but they will lack legal ownership of the share they have paid good money to purchase. The company issuing the share could move its residency. But many other salient factors drive the residency of a company including corporation taxes, the legal environment and commercial advantage. Saving a tax that investors in the secondary market for shares pay and which is a tiny fraction of their total transaction costs is not on par with these other factors. It is one of the reasons why, despite the tax being around since 1694 in the UK, giving ample time for lever financiers to innovate ways around it, it still raises a large chunk of total corporation taxes. The feasibility of levying and collecting financial transaction taxes is not in doubt.

Transaction Taxes, Financial Stability and Economic Growth

I argued earlier that financial crashes invariably follow financial booms.¹⁶ The bigger the boom, the deeper will be the crash. One of the big issues in crisis prevention is therefore limiting the size of booms. In a seminal paper, DeLong and colleagues¹⁷ show that asset market booms can become large and self-sustaining if there is a preponderance of “noise traders” in the market. They suggest that we have essentially two kinds of traders. The first are the so-called fundamental traders, pursuing value and who tend to be sellers during booms when financial prices rise above historical metrics of value. In crashes they become buyers. Finding value is a long-term goal and can lead to short-term losses. They resemble investors managing a long-term pension fund, life insurer or sovereign wealth fund.

Not all investors can cope with short-term losses. In a world of uncertainty concerning where long-term value lies, capital constraints and risk limits, clients have become increasingly risk averse when markets are going against them. Consequently, there is second type of trader, who DeLong et al call noise traders. Noise traders focus on momentum strategies, buying when markets have risen in the past and selling when they have fallen. Many hedge funds and High Frequency Traders resemble noise traders. Rather than stubbornly selling into a boom, they will buy from the fundamental sellers and gain short-term profit if prices continue to rise. Once prices are rising and noise traders are profiting, more noise traders will enter the market to buy. As prices rise above their fundamental value, the more fundamental traders will sell to noise traders and exit. This creates a self-feeding bubble, pulling markets further away from their long-term value. This process has a natural limit, as there comes a point when there are no more fundamental traders left to sell to the trend followers. DeLong’s work as well as subsequent experimental and empirical studies, reveal a threshold above which the bubble collapses upon itself.

The presence of a significant number of momentum-driven, short-term noise traders makes a market susceptible to bubbles. Bubbles lead to crashes so the more noise traders there are, the greater the systemic risk. Small transaction costs create a return hurdle for short-term traders—one that is far less relevant to long-term traders. Consider the following simple example using a 0.1% tax on security purchase and sale. The long run, average, annual return to equities is approximately 7.5 percent per annum.¹⁸ A long-term investor turning over his portfolio once every three years—the average of most pension funds—can expect a return of about 22.5 percent over these three years (approximately 7.5 percent x 3) His effective annual tax rate would be less than 1 percent of his return (0.1 percent x 2 ÷ 22.5 percent). To be clear, this is 1.0 percent, not percentage points, meaning that he is left with 99 percent of the return. A short-term momentum trader, turning over his investment portfolio weekly, can expect to generate an average return of 0.14 percent over seven days (7.5 percent ÷ 52). He would pay more than that in a 0.1 percent transaction tax levied on both sides of the trade. His effective taxation for a week is over 100 percent and so he would not trade. The incidence of the transaction tax rises as trading behavior shifts from long term to short term. That is the intention.

Given the link between the number of noise traders and systemic risks, this is a classic Pigouvian tax—a tax applied to an activity that is generating negative externalities (costs for others). Realistically we cannot always calculate negative externalities precisely so governments try to at least make Pigouvian taxes proportional to the amount of the activity they are trying to offset.¹⁹ The more pollution a company produces the more taxes it can expect to pay.

Why is a Pigouvian tax needed in financial transactions? Assume a broker with two customers. One customer, more akin to a fundamental trader, turns over her entire portfolio (a pension fund or life insurance institution) once every three years while the other, similar to a noise trader, turns over his portfolio perhaps more than 52 times annually. Which of the customers generates more commissions, more market information, and greater financial flows for the broker? Often modestly sized hedge funds of a couple billion dollars generate more profit for banks, brokers and exchanges than the mammoth pension fund with hundreds of billions of dollars under management. US stock exchanges report that high-frequency traders accounted for approximately 50 percent of US equity trading volume—a number that in times of stress, as in 2009, rises to between 60 and 73 percent.²⁰ It is little wonder that, as discussed in earlier chapters, it has been alleged that banks were allowing their high-frequency trading clients to take advantage of their institutional clients in dark pools of liquidity. This is exactly why we need a Pigouvian tax. Untaxed, the natural order is for the system to favor high-frequency trading over low-frequency trading with potentially serious systemic risks. Predictably, however, banks are vehemently opposed to financial transaction taxes believing it will probably kill off an important income stream.

In other areas of economic life, taxes based on the value of an activity are considered more efficient than taxes based on volumes and transactions. The IMF, among others, has proposed taxes on bank balance sheets to disincentivize banks growing into a systemically important institution, or taxes based on the amount of lending activity.²¹ However, given the inducements for traders to be involved in churning of their clients’ portfolios coupled with the role of short-term, “noise”, trading in asset market bubbles, it is hard to find a more direct measure to deal with this key source of financial instability.

Transaction Taxes and Financial Liquidity

In response to arguments that a financial transaction tax will strengthen financial resilience, bankers have responded with research purporting to show that high-frequency trading improves liquidity. It must, the argument goes, therefore reduce volatility and, by extension, financial transaction taxes will cause liquidity to fall and volatility to rise. This reflects unfamiliarity with what liquidity actually is. Indeed, many of the young econometricians doing these exercises have had limited experience of markets and crisis.

While high turnover is a common symptom of liquidity, financial market liquidity is actually about diversity.²² A market with only two participants would be highly liquid if whenever one wants to buy, the other wants to sell. A market with a thousand participants who each use the same model to value assets and have the same trading strategies, so that when one wants to buy so do the other 999, would be highly illiquid. During trendless times, high-frequency traders tend to adopt contrarian strategies. They buy when the market is moving down and sell when the market is moving up. At these times, when data is plentiful and the econometricians run their tests, high-frequency trading adds to liquidity. But they are adding to liquidity when it is already in abundance. When volatility is low, turnover is high, and therefore bid-ask spreads (the difference between the buy and sell levels of traders) are also depressed. In times of crisis or sharp market moves, the short-term trading models become momentum driven and high-frequency traders try to run ahead of the trend.²³ They are draining liquidity and doing so when the system needs it most. A graphic illustration was the “Flash Crash” on May 6, 2010.

A significant amount of cross-sectional²⁴ evidence of high-frequency traders and their program and algorithmic cousins trying to run ahead of sell orders exists. However, the time-series data has breaks in it where data is simply unavailable. The econometricians don’t run their models over these time periods because at these points time-series data is sparse and unreliable.²⁵ Traders post wide bid-ask spreads to turn away customers and volatility is high. But we know that unless high-frequency traders deliberately act as contrarians in these events and have the capital to carry large losses, they cannot be adding to liquidity. We can safely conclude that high-frequency trading adds to liquidity when it is already plentiful and takes it away when it is most needed.

Another disingenuous argument that banks and traders use against financial transaction taxes involves computing the economic cost of increasing transaction costs and lost business they might experience. Implicit in their calculation is that transaction costs that go to the private sector generates GDP and transaction costs that go to the state falls into a botomless pit. The European Commission originally estimated that their proposed 0.1 percent financial transaction tax will lower GDP by approximately 0.2 percent.²⁶ All taxes lower GDP if you fail to take into account either the reduction of negative social externalities or that the cash is used in some way to lower corporation taxes, labor taxes, reduce debt or any other activity that actually boosts GDP. In any event, the figure is likely to be an overestimation. The economic forecasting model used by the European Commission also assumed that the cost of capital rises by the same amount for everyone in the economy. This assumes away that only 15% of investment is financed by the issuance of equity and debt securities²⁷ and the difference in the incidence of the tax caused by different holding periods. A transaction tax will raise the cost of capital for long-term investors by far less than for short-term traders.

Even if we ignore these issues, once we reflect on the economic costs of financial crisis, transaction taxes only have to contribute a tiny reduction in the risk of asset market bubbles in order to have a net positive effect. Reinhart and Rogoff estimate that, from peak to trough, the average fall in per capita GDP resulting from major financial crises is 9 percent.²⁸ The UK’s Institute of Fiscal Studies suggest that, based on its comparison between the real median household income in 2009–10 and that of 2012–3, the impact of the GFC was a 7.4 percent decline in real incomes. European countries like Greece, directly hit by the sovereign debt crisis, experienced a greater decline in GDP and incomes. During the Asian Financial Crisis, initial GDP losses were also far larger, at around 20 percent. Consequently, even assuming revenues are buried in a hole in the ground, the tax only has to limit noise trading to the extent that it creates a 2 percent reduction in the probability, or size of, a major bubble and bust that occurs once every seven years or so. Estimating the costs, causes and probability of crisis is fraught with difficulty. Yet the tiny impact on the probability of crisis required to produce a meaningful difference to GDP, the small estimated costs involved, and the ability to reverse course if necessary, means that it is worth trying.

Avoiding Financial Transaction Taxes

The traders’ argument against transaction taxes has a curiously religious slant. They present these taxes as uniquely evil even though the issues of implementation, avoidance, and evasion are common to many taxes and charges. Consequently, there are some general observations to make when considering the argument that the tax will simply be avoided. Inherent in all taxes is the incentive and potential for avoidance and evasion, occurring proportionally to the size of the tax. Therefore, as with all taxes, it is important to set the right tax rate. Bear in mind too that modest cost variations caused by different taxes or other costs are sustainable. Within the OECD, for example, corporate income and profits taxes range from in Denmark’s 60.9 percent to the Slovak Republic’s 17.9 percent. The magnitude of this divergence has been fairly constant over a period where Denmark enjoyed one of the highest levels of GDP per capita in the world, and, incidentally, measures of happiness. Fear over the relocation of trading as a result of a small transaction tax is overstated. Despite the UK having a financial transaction tax while several stock markets do not, the London Stock Exchange has emerged as one of the largest, most liquid stock markets in the world. Low-tax Switzerland has a financial transaction tax. A financial transaction tax has long coexisted with a vibrant and rapidly growing stock market in places like Hong Kong, India, Taiwan, and South Africa. Three-quarters of the members of both the G-8 leading industrialized economies and the G-20 grouping of the largest economies in the world levy some form of financial transaction tax.

To require transaction taxes to be 100 percent avoidance proof is to set a bar not required for any other tax. The reality is that stamp taxes have much higher rates of collection and lower collection costs than almost all other taxes. Ordinary shares and nonbearer bonds have centralized registries of ownership, which are usually located where the issuer is headquartered or the shares listed.²⁹ You will recall that in stamp duty jurisdictions, unless the share certificate (these days an electronic document) has been stamped to prove that all taxes have been paid, the transfer of title is not legally enforceable. This process guarantees that stamp duties on financial transactions are some of the least avoided taxes—with nonresidents taxpayers lining up to pay rather than risk not having legal title to a share they purchased.

Derivatives

Not all securities have centralized registries where shares are listed. Anyone can write a contract where value is derived from the price of anything else, including a share or bond, without ever having title to that share or bond. We can make an agreement that if the price of BP shares rise above £5.50³⁰ before June 2020,³¹ I will pay you £1,000.³² In return for this contingent liability, you agree to pay me a consideration, determined by our assessment of the likelihood of the payout, of maybe £250.³³ Simple options like these, known colloquially as “vanilla” options, are essentially insurance contracts where one party is insuring against a financial event taking place and paying a premium for that insurance while the counterparty is underwriting the insurance and receiving a premium.³⁴ The person with the contingent liability is the insurer, otherwise known as the writer or seller of the option, and the person paying the insurer the premium is the insurance holder or option buyer.³⁵

An importer of Venezuelan oil, for example, on reading the international financial press, may become anxious that the Venezuelan oil company PDVSA is in financial difficulty and will renege on its commitment to supply it with oil. If this happened the importer would in turn find it impossible to fulfill supply contracts to his customers, leading to penalties amounting to, say, $1 million. An advisable course of action would be for the importer to buy a credit default swap issued by an investment bank that would pay it $1 million if PDVSA declared a default. Residency in Venezuela is not a requirement for issuing or buying this insurance. Indeed, derivative contracts can be written on any verifiable event, taking place anywhere, and can be issued by anyone competent and sold anywhere. The issuance principle of taxing transactions cannot capture a tax on such derivative transactions.

Opponents of a financial transaction tax jump up and down with glee at this point. Before we assess how to address this loophole, it is worth asking why, after three centuries of financial innovation in the UK and elsewhere, all transactions subject to a tax on a title transfer have not shifted to the derivatives market. There are two main reasons why the failure to place taxes on derivatives has not, and never will be, fatal to transaction taxes. First, the stock market is the market for corporate control. Large institutional investors, investor activists, socially responsible investors, and corporate raiders—who together account for more than two-thirds of all investors in equities by assets—own shares so that they can influence corporate decisions. Unless they have good title to the shares, they cannot attend the annual general meeting, demand board membership, exercise corporate social responsibility, earn dividends, benefit from a scrip or bonus issue,³⁶ or protect against dilution of interest in a rights issue.³⁷ Given the range of corporate actions that are possible, such as rights issues, mergers, spin-offs, and acquisitions, owning good title to shares provides essential investor protection of both interests and returns. The majority of shareholders need to own their shares rather than only benefiting from short-term price movements.

The other key reason why the failure to tax derivatives is not a fatal blow to a financial transaction tax is that derivative contracts mainly give the right or obligation to purchase/sell shares at some future date. Unless on expiry the derivative contract becomes worthless, one party will purchase or sell the shares on which the derivative contract was written. Even when a derivative is settled with a cash payment, current market practice will hedge shifts in this potential cash payment or receipt through transactions in the stock market. It is the most efficient hedge of the potential exposure. If the likelihood increases that a holder of a derivative will have to buy a share at a higher future price and deliver it at a fixed lower price (having sold a call option), she will hedge that potential loss by buying some shares today. Holding derivative contracts often spurs one, sometimes multiple, transactions in the underlying cash market. A tax on transactions in the underlying market will reduce the frequency of these hedging derivative transactions in that market but hedging will still take place, as the potential costs of not doing so will more than offset the benefit of saving a little tax. In the case of the proposed European Union financial transaction tax this saving is just 0.01 percent tax on the notional value of a derivative or 0.1 percent of its economic value.

While it may be unnecessary, there is scope to tax derivative transactions, but not through the issuance principle. The issuance principle is not the only or even the main principle of taxation used in finance. The current international practice is to levy income and capital gains taxes on the proceeds from shares held by residents. The incidence of this tax falls on residents whether the shares are registered locally or overseas, purchased using a local or overseas broker, or held by a local or overseas custodian. Tax residency rather than issuance is key. A financial transaction tax based on both the issuance and residency principles, as proposed by the European Commission³⁸ and introduced by the French and Italian authorities in 2012 and 2013, respectively, will capture residents’ transactions in derivative or collective instruments including those that are “off market.”

Where the tax is due, but the trade has taken place in a foreign jurisdiction that does not levy a tax of its own and so is not automatically deducted by the clearing agent, residents would be liable for reporting the transaction and paying the tax. This can be done through annual tax returns as is currently done for the assessment of capital gains tax³⁹ or more frequently if the jurisdiction desires. In the UK, for example, if the relevant agent has not automatically deducted the tax at the point of clearing or settlement, residents must, within one month, report transactions and are subject to interest and penalties for any delay. Other countries could easily adopt this practice.

Given that transactions are not legally enforceable in tax jurisdiction countries if the tax is outstanding, clearing and settlement houses would be incentivized for the sake of legal certainty to collect relevant transaction taxes at the point of clearing. Requirements introduced after the GFC that required all vanilla derivatives to be centrally cleared would make this the most likely route for collection—even under the residency principle. We discuss central clearing in greater detail later. However, assuming that clearinghouses⁴⁰ always act in their best self-interest, they are likely to collect the tax without any regulatory encouragement—as they have done with the UK stamp duty. The trading industry argues that this would incentivize traders to evade central clearing and undermine financial stability. This is an oddly desperate argument, evocative of an age when the industry routinely threatened bad behavior in order to get its way. The tax would be due irrespective of whether the instrument is cleared or not. The capital charge and management risks of uncleared instruments would anyway make such evasion both pointless and expensive.

The New International Tax Environment

Up until 2008, the opportunities for avoiding or evading transaction taxes levied on the residency principle were far greater than those levied on the issuance principle. Tax residents in one country could set up a shell company in another that did not levy a tax and did not have a legal or enforced requirement to disclose the residency of beneficial owners. They could then purchase derivatives on shares registered in a financial transaction tax jurisdiction without paying the tax. Although it was not a major cause of the crisis, regulators trying to manage bank failures during the GFC often found themselves caught in a morass of legal entities established largely for the purpose, and I am being kind, of tax minimization. In the Lehman bankruptcy, administrators had to grapple with about three thousand different subsidiaries. It is foreseeable that those familiar with this mode of operation believe that a transaction tax based on the residency principle would simply push derivatives trading elsewhere. The trading industry has made several governments worried about this prospect. But times have changed.

Finance used to be presented as something ethereal—materializing momentarily before disappearing again and impossible to pin down, report, and tax. Whether that was ever true, events over the past 15 years have changed that perception. The 9/11 tragedy spawned new, tough anti-money laundering and antiterrorist financing measures and rules. Then came the GFC, which reinvigorated the role of the tax, licensing, and regulatory authorities. There will be much skepticism on the efficacy of international tax assistance, especially after what took place before the financial crisis and what has since been revealed about the absurdly low taxes paid by major nonfinancial corporations such as Apple, Amazon, Google, and Starbucks. These corporations utilized several schemes to avoid corporate tax, including tax inversion-driven mergers, base erosion, profit shifting, the “Dutch sandwich” and the “double Irish”⁴¹ schemes.

However, several recent developments collectively suggest that in the future we can rely far more on the residency principle for the taxation of financial activity in general, and derivative instruments specifically, in a way we could not prior to 2008. One of the most important occurrences must be the new anti-money laundering regime sponsored by the 36-member Financial Action Task Force and its eight associate regional task forces. A measure of the effects of this new regime is found in the fieldwork of Michael Findley and colleagues, who have looked at the ease with which shell companies can be set up across the world.⁴² This work, initially done in 2010 and updated in 2012, shows that in many of the international financial centers of small states, where it is often thought that compliance is problematic, it is no longer possible to establish shell companies. This is true for the island of Jersey, the Cayman Islands, the British Virgin Islands, Monaco, Gibraltar, Luxembourg, the United Arab Emirates, the Seychelles, Bahamas, the Isle of Man, Barbados, and Bermuda. These jurisdictions no longer find bearer bonds admissible as vehicles for corporate ownership or for any financial purpose such as collateral for loans.

Ironically, the worst performers in this field experiment were actually the major jurisdictions that were the most vocal about the role of small states in international tax evasion—the United States, the UK, Canada, and Australia. There is further work to be done on eliminating shell companies, but Findlay and colleagues suggest that much has already changed in the outposts of international finance. What needs focus is at home, in countries that delight in publicly berating foreign companies and small jurisdictions, and falsely boast that they are leading the fight against international tax avoidance. This should not be difficult to correct but these issues are more political than most realize and relate to a point we introduced in Chapters 3 and 5 on the desire of regulators in big financial centers to promote their particular centers and national firms.

The failure of several larger economies to “walk the talk” on eliminating shell companies makes important the G-20’s call at its April 2009 London Summit to amend and extend the OECD Convention on Multilateral Assistance in Tax Matters. In 2010, the convention was significantly amended to provide for all possible forms of administrative cooperation between states in the assessment and collection of taxes, including automatic exchanges of information and the recovery of foreign tax claims. To date, 70 countries have signed the convention, including all major financial centers.⁴³

In March 2010, another major event occurred with the passing of the Foreign Account Tax Compliance Act (FATCA) in the US Congress, which requires US citizens, including those living outside the United States, to report their financial accounts held outside US jurisdiction. FATCA also requires foreign financial institutions, under threat of severe sanctions for noncompliance, to report to the IRS.⁴⁴ There does not seem to be an American translation for “extraterritorial.” Thirty countries, including all European and G-7 countries, have already established local rules mandating their local institutions to comply with FATCA. The reason this is especially significant is that based on the United States establishing this principle and model, and getting it expensively complied with abroad, the UK and the EU have openly discussed replicating it, using the growing network of compliance agreements. A European FATCA is probably on its way. The European Commission is waiting for details of the automatic tax-information-exchange model that is part of the OECD Multilateral Convention.⁴⁵ If that is unsatisfactory, it might consider extending the administrative cooperation directive to cover all tax administration. The repercussions of FATCA and some of the European initiatives must not be underestimated. Had they been in place in 1963, the London Eurobond market would not have developed.

The GFC has pushed us to a point where the reporting of exposures of licensed financial institutions’, on and off balance sheet, is mandatory, and as mentioned earlier, central clearing and settlement is required of the most heavily traded financial products.⁴⁶ The measures that have created this situation include, most importantly, the European Parliament and Council’s Regulation on OTC) Derivatives, Central Counterparties, and Trade Repositories (EMIR).⁴⁷ It is estimated that by 2015 the notional value of over-the-counter (OTC) derivatives that are centrally cleared will be in excess of $470 trillion.⁴⁸ Institutions failing to comply will suffer severe penalties and will be eliminated from crucial access to funding, payment systems, and licensed activities. To be non-compliant would amount to a financial death penalty.

Related to such measures is a newfound aggression by regulatory authorities in fining institutions and pursuing those they suspect of criminal actions. Credit Suisse was caught in this net and in 2014 agreed to pay a $2.6 billion fine and plead guilty to helping US citizens evade taxes due based on the residency principle. That guilty plea could escalate the eventual total cost to Credit Suisse. A number of its counterparties are forbidden by their internal rules to work with convicted felons. The United States also fined BNP Paribas more than $10 billon and barred it from dollar-clearing facilities for a period in order to settle allegations that it violated trade sanctions by disguising transactions with Iran, Sudan, and Cuba. On the announcement of the fine, a further $10 billion was wiped off the share value of BNP through fear of the impact the temporary removal of dollar clearing would have on its business. HSBC was earlier fined $1.9 billion for routinely handling money transfers from countries under sanctions and from Mexican drug traffickers.

This new regime of increased reporting and closer supervision has bit hard, and several institutions have vacated whole sectors where they are unsure of compliance. Concerns have been voiced that these enforcement measures are being used politically and thus inefficiently. A level playing field does not exist. Large countries make the rules, which they do not apply to themselves, but are not shy to apply to countries or companies that cannot object or retaliate.⁴⁹ What is clear, however, is that there are mechanisms in place to make financial taxes based on residency and issuance work, if policy makers are prepared to use them.

The trading industry’s opposition to a tax that they say will not work is immense. The strategy they employ is similar to the tobacco industry in the 1970s, namely, obfuscation of the issues, ridicule of its proponents and repetition of falsehoods to such an extent that listeners feel they must be right. Although we have already set out the case for financial transaction taxes, we now turn to assess some of common arguments used against these taxes. Following that we will look at the issue of tax rates and instruments and then conclude.

Lessons from the 1984 Swedish Tax

The requirement to disclose the beneficial owner of a company was not in place in 1984 when Sweden initiated a 0.5 percent financial transaction tax, raised to 1.0 percent in 1986. It was levied entirely on the residency principle and collected by local brokers. Swedes wanting to evade it established nonresident accounts in London and traded in Swedish stocks from there. Tax revenues were lower than expected as a result. Had the tax also included the issuance principle rather than relying solely on the residency principle, the result would have been different. It would have ensured that all purchasers of Swedish shares, regardless of location, would have to pay the tax to secure legal title to the shares.

This poorly designed tax also suffered from being introduced in an age when residents could evade taxes by going offshore and establishing nonresident entities with the active encouragement of their brokers, bankers, and sometimes the foreign jurisdiction itself. London’s current position as one of the world’s largest offshore financial centers started with turning a blind eye to the tax status of its foreign clients and the creation of an offshore bond market.⁵⁰ It has continued with favorable tax treatment of income and capital gains for those in the hedge-fund and private-equity sectors. Today the Swedish tax could not be so easily evaded. The Swedish beneficial ownership of the London entities would have to be declared in order for the entities to be established. Without declaring their beneficial ownership, Swedes cannot open a bank account from which to trade or have an account with a counterparty licensed to broker Swedish shares in London or Stockholm. Once the beneficial owners are established, with an automatic exchange of tax information, a tax demand would follow with penalties for late payment. Failure to comply with either the Swedish or British tax authorities would leave the directors of the corporate service companies, banks, or brokers open upon conviction to five years imprisonment, fines in excess of $500,000, or both. The times they are a-changin.⁵¹

No tax is watertight. Estimates suggest that between 20 and 30 percent of income tax is either evaded or avoided. But this loss has not lead to an outcry that income tax should be scrapped. The standard of compliance being required of financial taxes is excessive. Tax authorities exist to make tax evasion, money laundering, the financing of terrorism, corruption of public officials or any other illegal activity a high-risk, low-return game so that illegal conduct is kept to a minimum. The impact on banking practice of recent developments is clear, with banks weighing up the new balance of risks very differently from how they did in the 1980s when Sweden imposed its tax. Recently, for instance, citing the new legal environment and heavy financial sanctions for inadvertently assisting any illegal activity, J. P. Morgan’s board decided to withdraw banking services to any (non-American) on a list of politically exposed persons. RBC has joined in stating that the cost of anti-money laundering processes and sanction for breach of those processes was a factor in deciding to withdraw its wealth management business from certain jurisdictions.⁵²

A strong disincentive for tax evasion and avoidance is to make all untaxed, taxable instruments null and void—even where this is limited to financial transaction tax jurisdictions. While a derivative transaction could initially take place between non-residents of a tax jurisdiction, a significant part of the value of an instrument, far in excess of the value of a 0.1 percent tax, is its wide marketability and transferability. An untaxed instrument that cannot be transferred or marketed to a resident in a financial transaction tax jurisdiction has a severely reduced value to the point where it would be better to pay the tax. This is especially so if the instrument is based on a security issued in the tax jurisdiction where its natural buyers would reside.

Moreover, financial trades including derivative contracts are essentially zero-sum games. If I win it is because you lose. The potential winner has a strong incentive to verify that the loser cannot cancel their loss (and the winner’s gain) by moving their tax residency, or the tax residency of the beneficial owner, to a financial transaction tax jurisdiction and letting the instrument become null and void. At the start of life for a derivative contract, each side thinks it will win. Both parties are therefore incentivized to pay the tax upfront—a tax significantly lower than the profits each hopes to make. It is possible that even residents outside of financial transaction tax jurisdictions who are trading instruments related to, but not issued in, a tax jurisdiction, would want to voluntarily pay the tax. It is a simple and inexpensive insurance against the risk of the contract becoming null and void in the future. One party to the contract should be able to ensure that the instrument never becomes null and void, whatever the other parties do, by paying all taxes due.

The trading industry will strenuously object to the potential legal uncertainty of a null-and-void rule. Yet, in this instance, uncertainty for untaxed instruments is precisely the incentive we need to secure compliance. We can also address this issue less combatively by encouraging the development of a standardized amendment to the documentation of these contracts (ISDA/FIA⁵³) that provides for the automatic payment of the tax if one or both parties is a resident, or the beneficial owner is a resident, of a financial transaction tax jurisdiction. Amendments to ISDA contracts have already been introduced to deal with exceptional matters, such as bond instruments with collective action clauses or for those following Sharia law.⁵⁴

Relocation of Business

Financial sector lobbyists routinely threaten that banks will aggressively relocate if national governments impose significant taxes on their activities or employees. In 2008, Terry Smith, head of Tullet Prebon, grandly stated he would allow any of the company’s 950 London-based staff to move overseas before the UK’s 50p top tax rate came into force. The Guardian reported on April 14, 2010, that so far “none . . . have taken him up on the offer.” Relocation is much harder than the financial sector suggests.

The residency incidence of a financial transaction tax is determined not by where the trade takes place but by the “tax residence of the financial institution or trader.” Residents, or owners of instruments issued in a tax jurisdiction, are being taxed, not the trading venues. With regard to the European Union financial transaction tax, the Commission states that a financial transaction would be taxed where an EU resident is involved—even if the transaction is carried out outside the EU.⁵⁵ In the past, some financial institutions would try to confuse the identity of the trading entity through devices like shell companies. However, as already discussed, in today’s world where there is comprehensive reporting, disclosure of beneficial ownership and, with over 70 countries have signed up to cooperate on tax matters, this is becoming increasingly difficult, costly, and risky.

Bankers have been making new threats. The claim is that they will move their derivative trading operations from those European countries that are considering a financial transaction tax to London subsidiaries where there is no transaction tax. On scrutiny, the impracticality of such a move makes it less menacing. If the trade is being conducted on behalf of a resident of a tax jurisdiction, then the tax would still be payable according to the residence principle just described. And where banks are transacting on their own account, they are required to put aside capital that absorbs losses and protects against the riskiness of their exposures. Pre-GFC, this capital could be easily shifted between locations and therefore trades could be distributed across jurisdictions to minimize tax. Following Lehman’s collapse, capital is now ring-fenced within countries making this practice more costly. Increased capital adequacy and margin requirements on derivative transactions established for tax avoidance would offset the saving of a 0.01 percent tax on the face value of derivatives or a 0.1 percent on its economic value.

A Tax on Consumers?

Commentators argue that ultimately it is the customers who pay this tax. This is the case in a highly competitive market where firms are on the edge of breaking even. No one can characterize the banking industry in these terms. Leaving aside the periods of banking crashes, its returns to capital and labor are superior to other industries, so it is possible that part of this tax will be financed through lower profits. The top thousand banks in the world reported collective profits of £540 billion in 2008 before collapsing in 2009 and then rebounding to £267 billion. To preserve market share, banks may well decide to swallow a proportion of a tax that represents, in a good year, less than 10 percent of profits.

But what about the portion paid by consumers? Crucially, not all consumers of financial products will pay equally. The financial transaction tax is an indirect tax on companies and a direct tax on investors churning⁵⁶ portfolios. In the UK and elsewhere, the first issuance of a company’s shares is exempt from the tax. The amount of the tax paid by investors relates to the degree to which they churn, or turn over, their investments in the secondary markets. A pension fund that buys a stock and holds it for three years will have an annual average tax rate of 0.06 percent of assets under management (0.1 percent x 2 for purchase and sale ÷ 3 years). This is a tiny fraction of other annual transactions costs that equity mutual funds report are in excess of 1.0 percent per year.⁵⁷ A hedge fund that turns over its entire portfolio once every three months would have an average annual rate of 0.8 per cent or 12 times more. Given that regulatory agencies generally only allow hedge funds to market to high-net-worth individuals, this tax will be progressive with hedge fund profits tapped far more than future pensions.

Tax Rates

As with any tax, the rate and taxable base are vexed questions. The incentive to avoid or evade a tax is proportional to its size. Pitch a tax too high and it could be at a point of diminishing returns. The key test of proportionality to prevent tax avoidance is to consider the tax in relation to all other transaction costs. Total transaction costs include considerably more than the simple bid-ask spreads that the industry is fond of focusing on to the exclusion of other transaction costs. “Revealed preference” suggests that the 0.5 percent tax rate in the UK and the concentration of rates in other countries around 0.25 percent to 0.5 percent have not been a material impediment to the growth of major stock exchanges or economies. The rate of 0.1 percent for cash transactions or the economic value of derivative transactions in the proposed European Union financial transaction tax or the proposed 0.03 percent rate in the US⁵⁸ are likely to be well below the rate of diminishing returns and a good place to start.

The challenge is how to set tax rates on derivatives. All derivatives have a premium, which is the cash consideration paid for the contract. However, by overlaying different options, it is possible to have a derivative with a potentially large payout but no upfront premium. Corporate treasurers are easily seduced by these low-premium or even “zero-cost” options. But, as we have discussed in previous chapters free lunches are rare and reducing the premium is often only achieved by increasing the likelihood of an expensive payout.

There are several techniques we can use to determine the right tax rate for derivatives. The tax can be levied on the fixed or maximum size of the potential payout—the notional value of the option. This has the merit of being simple and transparent making it the preferred approach of the European Commission. In order to minimise the distortionary impact of any tax, it is best to make it proportional to the economic value of the activity. Yet the economic value of an option relates to the likelihood of it being “struck” as well as its value if struck. Consider two financial options—one that pays out $100 million if there is a tsunami tomorrow and the other paying $100 million if a tsunami hits at any time over the next ten years. Both will have the same notional value of “$100 million” but their economic value will vary. Since the second option is more likely to be struck, it is more valuable and should incur a higher tax take than the former. To do otherwise would put low probability options at a disadvantage. Yet, low probability catastrophe insurance, for example, is socially useful. The potential payout is large but the probability of any payout is small. Consequently, the tax may be better set as a levy on both the premium paid and the end cash settlement as suggested in US proposals. A further advantage is that it can then be levied at the same rate as for all other securities—0.1 percent of the premium and 0.1 percent of the payout. Inconsistency of tax rates is a common enemy of compliance. It is also easier to achieve tax compliance when taxes are being paid out of an existing cash flow as opposed to future revenue.

Revealed preference presents a respectable alternative route. Aligning the notional values of derivatives to economic value of derivatives has previously been addressed by clearinghouses that clear derivative contracts and need to find a way of charging for doing so. Relative to the push back from the industry over the financial transaction tax, there has been little resistance to the call for mandatory clearing and the imposition of clearinghouse fees. It is estimated that given these new requirements, clearinghouse revenues from fees charged on clearing OTC derivatives) will rise above $10 billion.⁵⁹ Since the economic and market impact of every euro of a clearing house fee on a derivative must be the same as the impact of every euro of a transaction tax, the current level of clearinghouse fees appear well below the level of diminishing returns for a tax rate on derivative contracts. At a minimum, the authorities could start by charging a transaction tax at the same rate as clearinghouse fees.

Today clearinghouse fees average at around 0.002 percent of the notional amount of all cleared derivatives, or 0.05 percent of the gross market value.⁶⁰ These fees vary by product to reflect the different economic value of each instrument and the risks attached to clearing them. For instance, the fee rises if a product is an over-the-counter option versus a listed derivative; is bespoke or vanilla; is complex; has low trading volumes; or is settled by physical delivery rather than cash. These are useful factors to consider and helpful to the wider project of financial stability if the tax creates an added incentive for simpler products that pose less systemic risk. While including these factors may seem to be adding unnecessary complexity, under the European Parliament and Council’s Regulation on OTC Derivatives, Central Counterparties, and Trade Repositories (EMIR), all clearinghouses must publicly disclose the prices and fees associated with clearing services. It would be quite manageable and transparent as well as helpful to systemic risk management to simply link the starting tax rates to the current menu of clearinghouse fees. They can be reviewed periodically to ascertain their impact and reflect how clearinghouse fees have themselves changed over the period.

Coverage

The European Commission proposes to exempt government bonds from the standard 0.1 percent tax on cash transactions. The economic argument for this exemption is unclear. Private borrowers will view this as distorting the playing field in favor of government borrowers. If corporate bonds are to be taxed, then surely government bonds should be as well. But governments all over the world want to make their funding instruments more attractive than others so this may be a step that comes later. Beforehand policy makers should consider extending the tax to all credit derivatives—whether derived from corporate bonds or tax-exempt government bonds. The unhindered churning of credit derivative paper and the explosion of gross credit exposures were sources of systemic risk in the last financial crisis and an avenue for the muddying of the true picture of government debt that caused difficulties for some sovereigns later.

American depository receipts (ADRs), global depository receipts (GDRs), and nominee accounts are an avenue for nonresidents to avoid paying the financial transaction tax on shares originally issued in a financial transaction tax jurisdiction. Like a nominee account, in the case of an ADR, a tranche of shares of a French company, for example, is put into a depositary bank in the United States. ADRs are then issued by the depository bank against the shares they hold, and the ADRs are listed on, perhaps, the New York Stock Exchange in US dollars. The same arrangement in the UK is referred to as GDRs. A French resident cannot avoid paying a French transaction tax by buying the ADR of a French company listed in New York, because they are liable for tax on any instrument they trade irrespective of venue or instrument by virtue of their French residence and would have to declare the transaction and pay the tax in their annual tax return. However, Americans trading the French share using the New York listing of ADRs would avoid a tax they would otherwise have paid if they purchased elsewhere by virtue of the share being issued in France.

There are a couple ways to limit this. Shares in depositary banks held for the purpose of backing ADRs, GDRs, or any nominee account programs, could face a higher tax rate when they enter into the program and this would be observable at the share registry. This is the case with the UK stamp duty. Alternatively, legal title to the shares held in the ADR program could be subject to an annual fee based on the frequency of turnover. This essentially causes the bank managing the ADR program to pass on the transaction tax.

Conclusion

Contrary to the banking lobby opposition, small financial transaction taxes are feasible and desirable. At least 30 countries already collect approximately $38 billion in taxes through stamp duties on share transactions that are in the range of 0.25–0.50 percent of the value of a transaction. New rules on reporting, anti-money laundering, disclosure of beneficial owners, and automatic tax information exchanges, make extending these stamp duties to derivative and other instruments easier to implement and harder to avoid.

Without intervention, the financial system disproportionately favors short-term trading. Short-term trading adds to liquidity when financial markets do not need it and takes liquidity away when financial markets are in most need. There is a trade-off. In the past, too much emphasis has been placed on liquidity in the good times at the expense of liquidity during periods of stress. Financial systems with a preponderance of short-term trading are more prone to economically and socially expensive boom-bust cycles. Financial transaction taxes are Pigouvian taxes that serve to limit these negative externalities and do so in proportion to the size of the externality, with long-term investors paying the least and high-frequency traders paying the most.

Financial transaction taxes also raise revenues. These revenues may not be as large as some hope but they are large enough to be significant to national budgets. The proportion of taxes raised from the financial sector currently appears disproportionately low given the sector’s exemption from value added taxes and the degree of state underwriting of the sector. In the US, a small transaction tax is already levied that pays for the cost of the Securities and Exchange Commission. In Europe, the proposed bail out funds as well as other public initiatives could be financed by revenues from the transaction tax, doubling up on its contribution to financial resilience. Resilient markets also attract more sustainable levels of investment.

___________________

¹I am including in this total the fines on J. P. Morgan for “egregious breakdowns in controls and governance” relating to its London whale-trading debacle in 2012 as well as fines on BNP Paribas, HSBC, UBS, Credit Suisse, and ABN Amro for sanctions and breaches of money laundering as well as aiding tax fraud.

²Hidden costs are estimated to be in the region of 50–60 basis points. See David Blake, “On the Disclosure of the Costs of Investment Management,” (Discussion Paper PI-1407, London: Pensions Institute, Cass Business School, 2014), www.pensions-institute.org/workingpapers/wp1407.pdf.

³In a study of approximately eighteen hundred US equity mutual funds from 1995 to 2006, aggregate trading costs were found to be 1.44 percent of assets under management, with hidden costs being around 0.55 percent. See Roger Edelen, Richard Evans, and Gregory Kadlec, “Shedding Light on ‘Invisible’ Costs: Trading Costs and Mutual Fund Performance,” Financial Analysts Journal 69 (2013), pp. 33–44. Similar results can be found in John C. Bogle, “The Arithmetic of ‘All-In’ Investment Expenses,” Financial Analysts Journal 70 (February 2014).

⁴The average pension fund turns over its portfolio once every three years, so a 0.1 percent turnover tax on the purchase of a security would be equivalent to a 0.033 percent per annum charge—one-twentieth of current annualized marginal costs.

⁵To understand credit counterparty risk, consider that you took out insurance against the default of a supplier. The market risk in this contract is the risk of the supplier defaulting. The credit counterparty risk is the risk that, given a default, the insurance company that promised to pay in the event of a default, has gone bust, perhaps because it had insured too many against the same risk.

⁶The existence of a form of stamp duty in Europe can be traced back to Roman times, when Emperor Justinian decreed that there must be certain inscriptions on legal forms in order for them to be enforceable.

⁷In reference to these transactions, it is called the stamp duty reserve tax.

⁸In England, stock exchange activities began in the coffeehouses in the City of London, like Jonathan’s in the 17th century and later Garraway’s. Long before the white socks and red braces, the coffeehouses had been banned from trading in the Royal Exchange for excessively raucous behavior. The London Stock Exchange was formally established in 1801.

⁹The UK government announced that it would abolish the stamp duty and stamp duty reserve tax on transfers of interest in exchange-traded funds, or ETFs, and on transactions in securities admitted to trading on a recognized growth market (like the AIM, or alternative investment market) provided the market is not also listed on a recognized stock exchange.

¹⁰See Stijn Claessens, Michael Keen, and Ceyla Pazarbasioglu, “Financial Sector Taxation: The IMF’s Report to the G-20 and Background Material,” IMF, September 2010, http://www.imf.org/external/np/seminars/eng/2010/paris/pdf/090110.pdf.

¹¹The banking industry argued that it would be hard to assess. No doubt it also said to governments, as it often does, that it is customers, as in voters, who will pay.

¹²See Claessens, Keen, and Pazarbasioglu, “Financial Sector Taxation.”

¹³This is a conservative estimate based on a narrow definition of securities transaction taxes. Brazil raises $15 billion each year from different types of financial transaction taxes, the UK $6 billion, and Taiwan $3–4 billion alone.

¹⁴See House of Lords, European Union Committee, “Towards a Financial Transaction Tax?” (HL Paper 287, London: Authority of the House of Lords, 2012), www.publications.parliament.uk/pa/ld201012/ldselect/ldeucom/287/287.pdf.

¹⁵Euroclear also provides central clearing and settlement services for Belgium, Finland, France, Ireland, Netherlands and Sweden across around nine hundred thousand different securities.

¹⁶See chapter 3.

¹⁷J. Bradford DeLong, Andrei Shleifer, Lawrence H. Summers, and Robert J. Waldmann, “Positive Feedback Investment Strategies and Destabilizing Rational Speculation,” Journal of Finance 45, no. 2 (1990), pp. 379–95.

¹⁸See Elroy Dimson, Paul Marsh, and Mike Staunton, Triumph of the Optimists: 101 Years of Global Investment Returns (Princeton, NJ: Princeton University Press, 2002).

¹⁹Arthur Cecil Pigou was a 19th-century Cambridge economist who made substantial contributions to the discussion of the divergences between the socially optimal outcome and the one left purely to market forces—negative externalities. See A. C. Pigou, “Divergences Between Marginal Social Net Product and Marginal Private Net Product,” in The Economics of Welfare (1932; Memphis: General Books, 2010).

²⁰See “Times Topics: High Frequency Trading,” New York Times, December 20, 2012.

²¹For a useful survey of alternatives, see European Commission, Taxation Papers: Financial Sector Taxation (Working Paper 25, Luxembourg: Publications Office of the European Union, 2010), http://ec.europa.eu/taxation_customs/resources/documents/taxation/gen_info/economic_analysis/tax_papers/taxation_paper_25_en.pdf. See also the earlier reference to the IMF’s 2010 report to the G-20—Claessens, Keen, and Pazarbasioglu, “Financial Sector Taxation.”

²²See Avinash Persaud, Liquidity Black Holes (Discussion Paper 2002/31, Helinski, Finland: UNU/WIDER, 2002).

²³This piece of reality is not part of the model in DeLong et al’s paper. However, their results would not change if the model was augmented with the simple idea that noise traders don’t have one strategy but learn quickly. They adopt contrarian strategies when market are directionless and momentum strategies when they exhibit strong movements in one direction or the other.

²⁴Studies and surveys of market behavior at a particular point in time.

²⁵This reminds me of the economist joke where one night a man comes across an economist looking for his lost wallet under a streetlight and offers to help. After some time looking, the man asks the economist if he is sure he dropped it there and the economist says, “No, I didn’t drop it here but this is where the light is.”

²⁶Based on the most recent data from the European Commission as well as the Bank for International Settlements, the sources of financing companies are assumed to be: primary equity issuance (10 percent), retained earnings (55 percent) and debt (35 percent). Among the total debt of nonfinancial corporations, the share of debt securities could be estimated at about 15 percent (or about 5 percent of total financing). This mitigating factor is now incorporated into the second version of the model (see Julia Lendvai, Rafal Raciborskiz and Lucas Vogel, Securities Transaction Taxes: Macroeconomic Implications in a General Equilibrium Model, Economic and Financial Affairs, Economic Papers 450, EU Commission, March 2012, Brussels), which has brought down the estimated growth effects of FTT to –0.2 percent of GDP.

²⁷Ibid.

²⁸See Carmen M. Reinhart and Kenneth S. Rogoff, The Aftermath of Financial Crises (NBER Working Paper 14656, Cambridge, MA: NBER, 2009).

²⁹Moving corporate headquarters is a major deal and when this is done merely for tax purposes, such as in the case of mergers motivated by tax inversions, the tax authorities can sometimes neutralize the tax effect of the inversion, as was the case with US-headquartered AbbVie’s 2014 proposed $54 billon takeover of rival drug maker Shire.

³⁰This is known as the strike price of the option.

³¹This is known as the option expiry date.

³²This is the notional value of this option.

³³This is the option premium.

³⁴In Chapter 7, we noted that the connection between insurance and derivatives was not lost on the insurance firm AIG, especially where the capital requirements on the net insurance exposures of highly rated packages of credit insurance were lower than the capital requirements a bank would have needed for loans to those same credits. This provided an arbitrage opportunity across banking and insurance lines. AIG bulked up on both writing and buying insurance on highly rated packages of credit risks. This strategy exploded when the financial crisis hit, at which point AIG’s insured events coincided with the collapse of the credit quality of those who sold the offsetting insurance contracts. The corporation’s exposure moved from a tiny net position to the larger gross position.

³⁵Complex options are nothing more than the layering of a series of vanilla options. For instance, I may pay you if the price of BP shares rises above £5.50 as before, but now we might add an overlapping option where you will have to pay me the same amount if the price rises above £6.00, so that my exposure only lies between £5.50 and £6.00. Known as a collar, buyers of these options tend to be those with a view that prices will move within a small range and are really placing more of a bet on volatility being low than on a specific price being reached.

³⁶A scrip or bonus issue is one where a company’s cash reserves or part are converted into shares and distributed proportionally to existing shareholders.

³⁷A rights issue is one where a company issues new shares and gives existing shareholders a first right of refusal in taking up those shares, and therefore does not suffer a dilution in the proportion of the outstanding shares that they own.

³⁸See the European Commission, “Taxation of the Financial Sector: The Proposal of 14 February 2013,” European Commission, last updated April 14, 2014, http://ec.europa.eu/taxation_customs/taxation/other_taxes/financial_sector/index_en.htm#prop.

³⁹Note that in the assessment of capital gains tax on shares, in those jurisdictions that have capital gains taxes, information is required and already routinely disclosed on purchase and sale prices and times.

⁴⁰Clearing describes the activities that take place between the time a trade is agreed upon and the time when all the relevant payments are completed. Where millions of trades are agreed on daily, prices move around substantially through a day, and, as it takes 24–48 hours to settle trades, there is a potential for many a trade to slip twixt the cup and the lip. During clearing, one of the parties may go bust, leaving the other without the security it has paid for, or, having transferred the security, without the cash that is owed. Given that a small number of financial institutions are party to the majority of trades, it is safer and cheaper to have a centralized place—the central clearinghouse—where exposures between approved counterparties may be netted out so there only needs to be collateral and cash to cover the small net exposures and not the large gross exposures.

⁴¹The “double Irish” scheme is a tax minimization strategy that appears to rely on the fact that Ireland does not levy significant taxes on income booked in subsidiaries of Irish companies that are outside of Ireland.

⁴²See Michael Findley, Daniel Nielson, and Jason Sharman, Global Shell Games: Testing Money Launderers’ and Terrorist Financiers’ Access to Shell Companies, Griffith University, October 2012, www.griffith.edu.au/__data/assets/pdf_file/0008/454625/Oct2012-Global-Shell-Games.Media-Summary.10Oct12.pdf.

⁴³See OECD, “Convention on Mutual Administrative Assistance in Tax Matters,” last updated October 2014, www.oecd.org/tax/exchange-of-tax-information/conventiononmutualadministrativeassistanceintaxmatters.htm.

⁴⁴See IRS, “Foreign Account Tax Compliance Act,” last updated September 15, 2014, www.irs.gov/Businesses/Corporations/Foreign-Account-Tax-Compliance-Act-FATCA.

⁴⁵See, for example, PricewaterhouseCoopers, “Soon to be Released Common Reporting Standard Promises New FATCA-Type Obligations Around the World,” PricewaterhouseCoopers, January 9, 2014, www.pwc.com/en_US/us/financial-services/publications/fatca-publications/assets/pwc-tax-insights-common-reporting-standard.pdf.

⁴⁶The purpose of this requirement is to limit the systemic risk caused by the failure of a single counterparty. During good times, the counterparty may appear to present a small risk, but this may be because it is engaged in a large number of back-to-back transactions. If the counterparty fails and all of these underlying transactions fail, the system could fail. However, if there is an agreement through a clearinghouse on how this will be managed in the event of failure and how back-to-back contracts that net out will be handled and the net risk insured against, the risk of systemic failure could be avoided. The confidence this brings will spur activity.

⁴⁷The European Parliament and Council issued this regulation on July 4, 2012, and it entered into force on August 16, 2012. The implementing standards were published in the Official Journal of December 21, 2012. The main obligations of EMIR are the central clearing for certain (vanilla) classes of OTC derivatives; the application of risk mitigation techniques for non-centrally cleared OTC derivatives; the reporting of all transactions to trade repositories; the application of organizational conduct of business; prudential requirements for central clearinghouses; and the application of requirements for trade repositories, including the duty to make certain data available to the public and relevant authorities.

⁴⁸Bank for International Settlements, ISDA 2010 Market Surveys and Booz and Company Analysis (2009).

⁴⁹See Avinash Persaud, “Look for Onshore, Not Offshore Scapegoats,” Financial Times, March 4, 2009, and Avinash Persaud, “Power & Politics: How Small States Should Respond to OECD, FATF and G20 Initiatives,” Comsure (address at the IFC Conference, Barbados, September 2014), www.comsuregroup.com/power-politics-how-small-states-should-respond-to-oecd-fatf-and-g20-initiatives/.

⁵⁰The first Eurobond was issued in 1963 by Italian motorway network Autostrade. S. G. Warburg arranged the issue in London. By issuing US paper outside of the United States, the instrument attracted US investors but was free of national withholding tax.

⁵¹The title track of Bob Dylan’s 1964 album.

⁵²See Tom Brathwaite, John Paul Rathbone, and Gina Chon, “JP Morgan Shuts Foreign Diplomats’ Accounts,” Financial Times, May 6, 2014.

⁵³See ISDA, www2.isda.org/.

⁵⁴As we go to press, an even-more-ambitious ISDA clause is being discussed. Under pressure from the Financial Stability Board, the ISDA is considering including automatic standby clauses in derivative contracts when a bank is in trouble to stop a self-feeding panic about counterparty risk. Where there is a will, there is a way.

⁵⁵See European Commission, “Common Rules for a Financial Transaction Tax—Frequently Asked Questions,” last modified September 28, 2011, http://europa.eu/rapid/press-release_MEMO-11-640_en.htm.

⁵⁶Churning is the frequent turning over of an investment portfolio and, as is often associated with tax avoidance, the speculation on short-term events, awarding trading commissions and more. The average pension fund completely turns over its portfolio once every two years. Some high-frequency traders turn over their portfolios many times a week or even within a day.

⁵⁷See Roger Edelen, Richard Evans, and Gregory Kadlec, “Shedding Light on ‘Invisible’ Costs: Trading Costs and Mutual Fund Performance,” Financial Analysts Journal 69 (2013), pp. 33–44 and John C. Bogle, “The Arithmetic of ‘All-In’ Investment Expenses,” Financial Analysts Journal 70 (February 2014).

⁵⁸The proposed rate of the Bill proposed by Senator Harkin and Representative DeFazio and introduced to the Senate on November 2, 2011.

⁵⁹Estimate from Deutsche Börse Group, see: http://deutsche-boerse.com/dbg/dispatch/en/kir/dbg_nav/home. See also Amit Desai and Jussi Tahtinen, Getting Fit for Clearing: Pursuing the OTC Central Clearing Market, PricewaterhouseCoopers, 2011, www.pwc.co.uk/en_UK/uk/assets/pdf/pursuing-the-otc-central-clearing-market.pdf.

⁶⁰Estimates are from PricewaterhouseCoopers and Deutsche Börse Group.