The Financial Economists Roundtable Weighs in on Financial Transaction Taxes

This statement is the outcome of the Financial Economists Roundtable’s discussion at its annual meeting on 20–22 July 2013 in Napa, California. It reflects a consensus of more than two-thirds of the attending members. Larry Harris, Jay Ritter, and Stephen Schaefer are the authors.

Proposals to tax financial transactions are popular now. Eleven EU countries — Austria, Belgium, Estonia, France, Germany, Greece, Italy, Portugal, Slovakia, Slovenia, and Spain — recently agreed in principle to implement such a tax in January 2014. Bills to tax financial transactions are introduced regularly in the US Congress, where Senator Tom Harkin and Congressman Peter DeFazio introduced the latest such bill — the Wall Street Trading and Speculators Tax Act — on 28 February 2013.

Many concerns motivate these proposed taxes. Some people want the financial sector to pay for the tremendous economic costs that they believe it imposed on everyone in the global financial crisis of 2008. Others believe that the financial sector should bear its “fair share” of taxes; such beliefs are particularly strong in countries where financial transactions are exempt from value-added taxes and therefore “escape” taxation. Still others simply want new revenues to support additional government spending, reduce the deficit, or provide tax relief to other sectors. Finally, some commentators believe that too much financial activity focuses on short-term rather than long-term goals. They believe that a transaction tax would force investors and businesses to focus more on long-term values and less on short-term activities that they perceive to be wasteful.

Although many members of the Financial Economists Roundtable recognize and respect these concerns, we believe that governments should be extremely wary of imposing or increasing financial transaction taxes, which can have highly negative consequences with respect to the revenue raised. The dangers of damaging capital allocation mechanisms and of hurting investors in publicly traded securities are large.

Benefits from Financial Transactions

Financial transactions serve many important purposes in market-based economies:

• Investors use financial transactions to save money for future uses, with public, private, and personal retirement savings being the most important.
• Household borrowers use financial transactions to pay for large current expenditures, of which financing home purchases and tuition costs are two of the most important.
• Companies use financial transactions to fund new projects, many of which generate new jobs and produce new, beneficial products.
• Producers use financial transactions to hedge risks of promising but risky projects.
• Governments use financial transactions to raise money for infrastructure investment and social programs.

New or higher taxes on financial transactions would make these transactions more expensive and thus reduce their volume. Some investors would invest less and spend more now. Some borrowers would forgo current expenditures that would otherwise make them better off. Some companies would not invest in desirable new projects. Some hedgers would not undertake promising, risky projects that would otherwise benefit the economy. Losing these transactions would hurt both the broad economy and the millions of people who use or depend on well-functioning capital markets.

Taxing Financial Transactions Is Costly

Taxes on financial transactions make financial markets less liquid by imposing additional costs on traders and on the dealers who allow investors and companies to trade when they want to trade. Higher transaction costs would make stocks and bonds less valuable to investors; asset values would be lower and the cost of capital higher.

We are particularly concerned about the effect that a tax on equity trading would have on the market for new equity issues. Investors are willing to buy new issues in part because they value the opportunity to sell them at any point in the future: Investors know that they may need cash earlier than expected or that they may lose confidence in the issuer. Well-functioning equity markets make buying new securities attractive because such markets assure investors that they can generally sell their investments at low cost when they no longer want to hold them.

A financial transaction tax that exempts new issues, as is sometimes proposed, would still adversely affect the new-issue market, with the following chain of effects on the economy:

• A transaction tax would make equity markets less liquid.
• Both the expected lower liquidity and the taxes that would be paid by current and future owners of a security would make new equity issues less attractive to investors and thus lower their prices.
• The lower new-issue prices would result in higher costs of capital for companies.
• Some otherwise good investment projects would not be undertaken, resulting in a less productive economy and lower wages because there would be less capital per worker.

Financial Transactions Are Different

The central role of financial transactions in market economies makes financial transactions different from transactions in other goods and services. Unlike the taxation of trades in most goods and services, whose effects are usually direct and obvious, the taxation of financial transactions would produce many negative effects extending far beyond the financial markets. A tax on financial transactions would generate indirect effects that would impose costs throughout the economy.

Unintended Consequences.The introduction of a financial transaction tax would have many unintended consequences:

• Despite the best efforts of those framing the tax rules, trading would shift to markets and countries where transaction taxes are not imposed. The proposed adoption of transaction taxes by the 11 EU countries would undoubtedly shift much trading to London, Hong Kong, and New York City, as well as a variety of tax havens, benefiting countries that do not collect such taxes.

Sweden’s experience with financial transaction taxes from 1984 to 1991 is particularly instructive: Following the imposition of various taxes on equity trades, option trades, and fixed-income trades, a large proportion of trading moved out of the country — primarily to London — and Sweden collected far less tax revenue than expected.Equally instructive is the tremendous growth in the Eurobond market following the adoption of the US interest equalization tax in 1963. This transaction tax on foreign stocks and debt obligations effectively ended trading in those securities in the United States and transferred a large fraction of the market to London.

• Convoluted financial structures would be developed as financial intermediaries designed products and procedures to avoid transaction taxes. For example, a fund might invest in an unconsolidated special purpose entity domiciled outside the tax jurisdiction to avoid a transaction tax. This entity could then trade securities outside the tax jurisdiction, thereby avoiding the transaction tax.
• Because proposals to tax transactions generally exclude bank loans and time deposits, commercial banks would become increasingly important in the financial sector. This result should give pause to those who seek transaction taxes to penalize the financial sector for the excesses that led to the global financial crisis because much of the alleged bad behavior occurred in commercial banks.
• Large distortions would occur if the tax rates were set such that the tax payment on a derivatives contract trade differed from the tax payment on a trade with an equivalent amount of risk exposure in the underlying instrument (as the 11 EU countries currently propose). Such differences would cause traders to shift trading from the high-tax vehicle to the low-tax vehicle. For example, the derivatives contract-for-differences market in the United Kingdom exists in large part to allow traders to avoid the UK stamp tax.

These substitutions would not occur if the tax rates for equivalent risk exposures were all equal. However, equalizing these rates would require very complex rate schedules for derivative securities, such as options, for which the relation between the security value and the underlying risk is nonlinear.

• Large distortions would also occur if the tax rates were set to reflect the average holding periods of similar financial instruments. If a flat tax were applied to all instruments, short-term contracts, such as daily repos, would become much less attractive than long-term contracts, such as term loans, because repo traders would pay the transaction tax every day whereas traders in long-term contracts would pay it much less often. For example, a tax of 1 bp on a daily repo would be the equivalent of a 2.52% annual tax because repo traders would have to pay the tax on each of the 252 business days in a year. This rate, which is higher than current interest rates, would effectively stop trading in the repo market. A transaction tax could thus have a major — and, we suspect, unintended — effect on the funding of the financial system.

These distortions would not occur if the tax rates were set to equalize the rate of tax collection over time. However, equalizing these rates over time would be impossible for such contracts as options, whose contract termination dates are uncertain. Too high a rate on such contracts would cause traders to avoid using options; too low a rate would cause traders to shift into options to avoid the tax.

• Attempts by practitioners to avoid financial transaction taxes and corresponding attempts by tax authorities to secure financial transaction tax revenues would complicate the financial system. Financial risk managers trained in systems engineering recognize that complexity is an important cause of systemic risk. The additional complexity of new taxation systems and the inevitable efforts to avoid those taxes would work against efforts to reduce systemic risk.

Misplaced Concerns.Proponents of financial transaction taxes expect that the tax burden would be borne by the financial services industry. In fact, although the industry would collect the tax, the burden would fall primarily on its customers through higher fees and wider spreads between purchase and sale prices. Repeated experience has shown that the ultimate bearers of a tax burden are always those least capable of avoiding it. The financial services industry would hurt a little because the tax would decrease total trading volume, but its customers would hurt much more because they are the ones who ultimately must trade to invest, borrow, and hedge.

Concerns about excessive speculation and excessive volatility also motivate arguments in favor of financial transaction taxes. However, empirical evidence generally does not support this premise. Consider, for example, real estate markets, in which transaction costs and taxes are very high and make trading very expensive. If high transaction costs discouraged speculation, bubbles in real estate markets would be rare. Regrettably, they are not, as events in the real estate markets in the United States, Ireland, Spain, the United Kingdom, and elsewhere showed all too clearly in the years leading up to the global financial crisis of 2008.

Some commentators advocate a financial transaction tax to address perceived problems with high-frequency trading (HFT). Although a transaction tax would reduce HFT along with all other forms of trading, it would not end the practice. Nor would it address the most serious problems with HFT — involving such predatory strategies as front running, which generally depend more on trading speed than on trading frequency. Moreover, dealing and arbitrage strategies are the most common and important HFT strategies, which benefit investors and hedgers by making markets more liquid. Policies that hurt HFT traders who use such strategies would hurt all traders, including long-term investors.

Finally, a financial transaction tax is attractive to some proponents who are concerned about an excessive focus on short-term as opposed to long-term results. Although sympathetic to those concerns, we note that many investment and borrowing problems are, by their very nature, short term. Examples include the following:

• Operating companies, banks, mutual funds, and governments rarely have perfectly synchronized cash inflows and outflows. Instead, they must regularly invest their surpluses and finance their shortfalls. The generally short time horizons of these cash management transactions expose them to even small transaction taxes. The introduction of a transaction tax would greatly complicate attempts to manage cash outside the banking system.
• Many short-term transactions occur because investors, borrowers, and companies rationally delay making long-term commitments until they are sure that they have adequately explored all available options. Taxing short-term transactions would cause some important decisions to be made too early, which would result in wasted resources and opportunities.

Many governments already favor long-term investing over short-term investing by setting lower tax rates on long-term capital gains or on dividends paid to long-term stockholders. Using a transaction tax to further incentivize long-term transactions is unwise given the inefficiencies that it would produce.

Other Problems with Transaction Taxes

Because capital flows are highly sensitive to cost differences, a financial transaction tax would not collect anywhere near the revenues that a simple application of the tax rate to existing volumes might suggest. Instead, volumes — and thus tax revenues — would shrink as trading dropped or moved to other locations or to lower-taxed vehicles.

Financial transaction taxes are inferior taxes because they do not produce reliable revenue streams. Trading volumes vary substantially from year to year, and consequently, so do revenues.

Any transaction tax must tax trades arranged both at exchanges and off exchanges at the same rate; otherwise, trading would migrate to the lower-taxed venue. However, collecting taxes on OTC trades is much more difficult than collecting taxes at exchanges because of the inevitable tax compliance problems associated with trades that often are not reported to a third party. These problems would be particularly acute if the taxable trades occurred outside the jurisdiction attempting to collect the taxes. The ability to avoid taxes by trading OTC — whether legal or not — would lead to an increase in less transparent OTC trading and a reduction in tax revenues.

Finally, we fear that a tax imposed by some countries on transactions in other countries would lead to tax conflicts between countries, which can be politically troublesome and thus costly. But without such extraterritorial provisions, attempts to collect financial transaction taxes would prove both futile and costly to the taxing countries.

Conclusion

Not all taxes are economically sensible, regardless of how desirable they may appear. A transaction tax imposed at any economically meaningful rate by only some countries would cause many transactions to be shifted to other countries, resulting in far less revenue than a simple static analysis might suggest. Furthermore, to the extent that a financial transaction tax would generate substantial revenue, the tax and the associated reduction in liquidity would lower asset prices. Lower asset prices would cause decreased corporate investment, resulting in less capital per worker in the long run and thus lower wages throughout the economy. Therefore, governments should be extremely wary of introducing or increasing financial transaction taxes.

Please note that the content of this site should not be construed as investment advice, nor do the opinions expressed necessarily reflect the views of CFA Institute.