Keynes proposed it. Oxfam is circulating a petition world-wide to garner support. Gordon Brown has urged its implementation.  Bob Herbert has urged it in the New York Times. James Tobin called for it. And Larry Summers (and his wife) wrote an article advocating it. What is “it?” – a tax on financial transactions, an FTT. In its broadest and most effective form an FTT would be levied on all trades in three financial asset markets: equity, debt and currencies as well as the derivatives that are priced off each the underlying markets for each asset class.   

An FTT has multiple appeals. First, the revenue that could be generated is significant in terms of both the absolute dollar amount and when considered in the context of the financial needs of the U.S. government as it responds to the ongoing economic crisis. Second, and equally important, an FTT would impact relatively few individuals and institutions with benefits potentially flowing to many. Finally, the political logic of an  FTT – the dynamics of Wall Street vs Main Street – makes the tax a timely one that should have significant appeal in today’s political climate. In this paper we focus on the revenue raising potential and the economic impact of an FTT and conclude with a few comments on the political appeal of the tax. We also briefly address some of the counter arguments that opponents could be expected to make to the imposition of an FTT.

A.    How Much Revenue Could a Financial Transaction Tax Generate?

An effective FTT must recognize not only the three financial asset classes but also the different venues for trading and the types of instruments that are traded in each of these asset classes. Equity, debt and currencies are traded both on and off exchanges, the latter normally referred to as the OTC market. In addition, in each class there are trades in the underlying asset (“cash market’) and in instruments based on the underlying asset (“derivative market”), particularly futures and options. An effective FTT should cover both venues and all instruments in a manner that does not advantage one venue or instrument over others. Failure to do so may simply shift volume of trading between venues or instruments. 

Much of the earlier work done by advocates of an FTT was based on data from the 1980s or 1990s. [1] Although the revenue raised from calculations using this data is significant, it misses the huge growth in trading of financial assets, both in the underlying cash markets and in derivative markets, that has occurred as financialization has proceeded apace in the US (and other) political economies during the last decade. Thus, as we argue below, the amount of revenue that can be raised has increased, both absolutely and as a percentage of GDP or federal government revenue.

1.    Equity Markets

In 1998, the total value of stock trading on the New York Stock Exchange (NYSE) and the NASDAQ Stock market was $13.1 trillion; in 2008 the traded value on these two markets, that together account for well over 90 percent of stock trading in the US, totaled slightly over $64 trillion, an almost five-fold increase in a ten years. The value of trading in currency and debt markets has increased at similar rates over the period of financialization in the U.S.

In thinking about the social costs and benefits of an FTT, it is imperative to keep in mind that this five fold increase in equity traded value, like the increases in traded value of currencies and debt markets, is not primarily the result of more stocks listed for trading, an increase in market capitalization. At year-end 1998, the total capitalization of the US equity market was $13.4 trillion; at year-end 2008 US market capitalization was actually lower than in 1998: $11.6 trillion. Even taking into account the 2008 market collapse and using year end 2007 data, total US market capitalization was $19.7 trillion.  Thus US equity market capitalization grew 49.2 percent in the 1998 – 2007 decade but traded value in the cash market jumped 245 percent.  As a measure of financialization this means that, on average, shares in the US stock market changed hands about once/year in 1998; by contrast, in 2007 shares changed hands about every five months. It is difficult to ascertain any benefits that have accrued to the US economy or to the average equity investor as a result of this increase in turnover rate (in 2008 the average share change hands every two months). [2]

In 2008, an FTT of only 0.25 percent on cash market equity trading would have generated $160 billion, one side only. Levied on both the buy and sell side it would have generated $320 billion. Of course, as is apparent from the foregoing discussion, 2008 was an unusually active year for stock trading, however, the same FTT on the 2007 trading value would have raised $260 billion (two sides). 

As emphasized above, however, an effective FTT must be levied on all forms of equity trading. Thus, these numbers, as impressive as they may be, far from exhaust the revenue that could generated from an equity market FTT. In addition to the $64 trillion in equity trading that occurred in 2008, another $54 trillion of equity notional value [3] was traded in the S&P 500 Index Futures on the CME ($49 trillion in 2007) and $26 trillion in S&P 500 Index options on the CBOE ($25 trillion in 2007). [4] 

The existence of multiple methods of trading in equities (and other financial asserts) raises an important question: at what level should an FTT be set on the different products so that volume would not simply be shifted from one product to another?  Identical FTT rates on cash and derivatives probably disadvantages the derivative market since these products, futures and options, have expiration dates while actual (cash) equities can be held indefinitely. At the same time, however, because a cash portfolio of S&P 500 stocks can be easily replicated using derivatives and short term bills, exempting or under taxing derivatives would unfairly disadvantage the cash equity markets. Rather than undertaking a detailed analysis of this issue we apply FTT levels in the range considered by previous commentators. In addition to the 0.25 percent/side for cash equity trading (used for example by Baker, 1989), we apply 0.05 percent of the notional value per side for index futures and calculate an index options FTT at both the same rate as that for index futures and at a 0.5 percent of the premium rate. In both cases these levies are per side. These levels of FTT would have generated $54.6 billion from index futures trading in 2008 ($50 billion in 2007) and, using the 0.5 percent of premium rate, an additional $5.3 billion from index option trading in 2008 ($5.0 billion in 2007). 

We believe that levying and FTT on the notional value of index option trading is more appropriate than a tax based on the premium. This follows from the same reasoning as in the case of futures. It is easy to replicate a cash portfolio by simply buying a call and selling a put. The premium paid for these two transactions would be only a small portion of the cost of executing the equivalent trade in the cash equity market. Thus, a tax set as a percentage of the premium would significantly advantage the trading of index options over either the cash equity market or the equity index futures market. Index options, like index futures, are settled in cash with no exercise into the component stocks. Thus there is no additional transaction on which an FTT could be levied. An FTT of 0.1 percent/side, the same as for index futures, on the notional value of index option trading would have raised an additional $25 billion in 2007 and $26 billion in 2008. The premium based approach to an FTT may be more appropriate for options on individual stocks where exercise into the underlying equity is possible. Because the value of such trading is small relative to that for index options and futures we do not include an estimate of the additional revenue that could be raised from this source. 

Taking the above calculations that cover only equity and equity derivative trading, an FTT would raise the amounts listed in the table below.

2. Currency Markets

Equity trading is the most transparent in terms of value and equities are the most widely held financial asset class. The equity and equity derivative markets represent only one of the three for financial assets and are considerably smaller than the currency markets. Currency (and debt) markets are less transparent with the result that it is more difficult to estimate the revenue that could be raised from an FTT. However, the Bank for International Settlements (BIS) compiles a triennial report on the volume and value of spot (cash) currency trading and OTC trading of both currency and debt derivatives. This data allows us to develop estimates for the revenue potentially generated by an FTT. The most recent BIS report is from December 2007 and, as is the case for all of these reports, covers trading during the previous April. All major and most smaller central banks in the world provide data for the BIS report.5 

In April of 2007 the BIS reported world-wide average daily value of currency trading to be $3.2 trillion. The largest amount of trading was USD/EURO but USD/Pound and USD/Yen also accounted for significant trading. 

By location the center of trading was London, accounting for more than 35 percent of the total with NYC second at almost 17 percent and Tokyo a distant third at less than seven percent. The first issue, then in calculating the possible revenue from applying an FTT to currency trading is to determine the scope of what could be taxed by a single government entity such as the US federal government. 

At a minimum, an FTT could be applied to trades executed by entities located in the country levying the FTT. Thus a US FTT could cover the 17 percent of currency trading executed in NYC. In addition, an FTT could be levied on all trades executed by subsidiaries of US based financial institutions, raising the potential revenue sources significantly. The BIS does not break out currency trading by national origin of the executing entities; however, a conservative allocation of 25 percent of London currency trading to US firms would raise the percent of trading subject to an FTT to at least 25 percent, or approximately $800 billion/day. With a levy of only 0.01 percent/side an FTT would generate $80 billion/year, applied to both sides of the trade, assuming 250 trading days/year.

As was the case in equity trading, the cash currency markets account for only a portion of total currency trading. Trading in the OTC currency derivatives markets reached the level of $2.3 trillion/day in the BIS survey while trading in exchange listed currency derivatives totaled $1.7 trillion/day in 2007 ($2.4 trillion/day in 2008). [6] Applying the same 0.02 percent FTT (0.01 percent/side) to these markets would generate another $200 billion in a 250 trading day year. 

Another approach to the scope of an FTT levied on currency trading would suggest applying the tax on the basis of the currencies involved in the transaction. The USD is by far the most frequent currency traded, represented on one leg of almost 90 percent of the transactions in the 2007 BIS survey. In this case the same levy would raise considerably more revenue, generating approximately $360 billion/year. It may be objected that other countries are unlikely to cooperate in collecting an FTT for the US on currency trades where the sales desk executing the trade (the geographical basis use by the BIS) is not located in the U.S. However, the political dynamics of the FTT in the context of the 2008 financial crisis may open possibilities for cross border taxation that did not exist previously. For example, the head financial regulator in the UK, Alistair Darling, has repeatedly called for an FTT. The US would find a willing partner should we decide to pursue such a tax.

Trading in US markets has also grown significantly over the past two decades and, as was the case for currencies, exceeds total trading in the equity markets. Also as in the case for currencies, a large portion of this trading occurs in the OTC market although exchange listed interest rate products are significant in volume and value.

The table below summarizes the potential revenue raised from the FTT applied to both long dated and short dated debt derivatives in 2007 and 2008.

B.    The Economic Logic of an FTT: Who Would Bear the Tax Burden?

Taxes are best levied when they meet two criteria: first, that they are progressive in impact and, second, that the tax discourages unproductive resource use and may even encourage a shift of resources to socially productive uses. [7]  A progressive tax raises more revenue from those individuals or institutions that have more income and or wealth while a tax that discourages unproductive resource may encourage reallocation of resources such as labor and money to uses that are at least as productive as those to which the taxed individuals or institutions would otherwise apply them. An FTT meets both conditions.   

1.    The FTT is a Progressive Tax

A tax on the trading of financial assets will, obviously, be paid by those who trade such assets. Consider the most widely held of the three asset classes, equities and equity derivatives. While almost half of all households in the US own some stocks, the majority of these households own stocks indirectly, that is through a pooled investment scheme of some type. This is commonly a mutual fund or a pension fund (that may itself be invested in a series of mutual funds). Only slightly over 20 percent of US households own stock directly and the bulk of this ownership is concentrated in a much smaller number of households. It is only among the top 10 percent of households by income that more than half report direct ownership of stocks and only among this group that the value of such holdings exceeds $20,000/household. [8] Therefore a tax on trading activity will fall heavily on this affluent 10 percent of all households. Of course, even in this case, if a household follows a buy and hold strategy, they will pay very little tax.

Households that own stock indirectly will pay the tax indirectly to the extent that the portfolio managers who invest their savings engage in trading activity. These households can also exercise considerable control over the extent to which they pay the FTT by choosing funds that trade infrequently. Such funds also tend to be those that charge lower management fees, for example index funds that simply track a particular measure of the stock market such as the S&P 500. Since there is no evidence that increased trading by active managers out performs index funds (although it does generate increased revenue to the active managers), any shift by households into such funds will certainly not depress and will probably improve their long run returns. A significant shift of savings out of the hands of active managers and into the hands of index fund managers would likely reduce the compensation of the former and diminish the flow of individuals into these highly paid jobs. Any concerns about a possible resulting diminution of investment choice are unfounded. There are more mutual funds than individual stocks listed on the NYSE. We could experience a considerable decline in the number of the former without imperiling individual choice of investment vehicles. 

Further, there is no evidence that a decrease in the number of highly educated individuals entering the financial sector would have a negative impact on either the markets or the larger economy. In fact, it can be plausibly argued that fewer MBAs going into finance and more bright young people into education, medicine, or research careers would be a net positive impact of an FTT. The cost of carrying the financial sector suggests the desirability of reducing the portion of our economy accounted for and the share of our economic resources devoted to these activities. 

Aside from the relatively small percentage of households who own a significant amount of equities and are also active traders, who else would pay an FTT levied on equity and equity derivatives? There are two other categories of individuals and institutional traders that would be expected to bear most of the taxation: day traders and institutions such as hedge funds, proprietary trading desks at broker-dealers, and a large number of non-financial corporations who have borrowed to engage in financial activities that were frequently unrelated to their core business. [9]

The percentage of trading accounted for by day traders is undoubtedly down from the glory years of the late 1990s but, at least for NASDAQ listed stocks, it is still quite significant. We wish no ill to day traders but we also do not see any compelling reason to believe that a decline in their number would be detrimental to our economic well being. Stock markets survived and thrived for centuries without relying on people investing for returns of less than 0.5 percent (the round turn impact of our proposed FTT) and we are confident they will continue to do so. Day traders who seek larger returns will undoubtedly continue their activities although at somewhat reduced levels of profitability.

2. The FTT is a Socially Positive Tax

The question of institutions such as hedge funds and proprietary trading desks is more interesting and goes directly to the question of productive and unproductive use of resources.  The unprecedented increase in equity trading over the past 30 years is not primarily the result of increased activity by that small percentage of households with significant equity holdings or the emergence of day trading. Instead it is the result of (i) a shift in focus and activity of broker-dealers away from their brokerage function and towards their dealer function and (ii) the rapid growth in proprietary trading activity by components of the shadow banking sector, especially investment bank trading desks and hedge funds. 

In 1975 proprietary trading and associated revenue accounted for less than 25 percent of NYSE member firm broker-dealer revenues while commission revenues were 46 percent of revenues (although a minority of all broker-dealers, the NYSE members are the largest broker-dealers and account for the bulk of all broker-dealer activity and revenues).  By 2000 the proportion of NYSE broker-dealer revenues accounted for by trading and associated revenue was over 56 percent and commission revenues represented only slightly over 40 percent of NYSE broker-dealer revenues. [10] What was happening to NYSE member broker-dealers, and at other financial services firms, is a shift into trading and fee collection and a move away from the brokerage function. The increase in proprietary trading drove growth in stock trading, accounting for a significant portion daily stock market activity. Within the broker-dealer category, the proprietary trading desks of major firms such as Goldman-Sachs and Morgan Stanley (and others such as Bear Sterns and Lehman Bros before their collapse) have dramatically increased their market activity and share of total equity trading.

Broker-dealers, then, would be among the entities that would pay the FTT. To the extent these firms are simply skimming a few cents/share, e.g. the flash trading that has been in the news lately, any loss of activity would be of no concern to other market participants since it would reduce trading that is of very questionable legality anyway and, again, activity without which stock markets survived and thrived for centuries. Other proprietary trading that seeks returns in excess of the 0.5 percent FTT should be only slightly affected.  Further, as argued in our discussion of active and index portfolio managers, any diminution in the salaries of proprietary traders that reduces the flow of bright young people into these jobs is likely a plus for the economy as a whole. 

The arguments that apply to broker-dealers also apply to hedge funds and other participants in the shadow banking sector that contribute to the growth in equity and equity derivative trading as well as to non-financial corporations that have sharply increased their participation in financial markets over the past two decades. Any loss of small profit margin (less than the FTT amount) trading is not a loss to the economic well being of the US economy or economic well being of the vast majority of the US population. 

The growth in proprietary trading activity among equity market participants is paralleled in the currency and debt markets.  Currency trading has grown rapidly particularly over the past several years. At the same time, however, as the market has gotten bigger it has also gotten narrower. Thus the number of banks that are major players has actually declined. For example, in the 1998 BIS survey, there were 20 US banks that accounted for 75 percent of currency trading in the US; in contrast, in 2007 there were only 10 banks that accounted for 75 percent of US currency trading. [11] The impact of an FTT would therefore be felt primarily by a few very large banks. Again, this suggests the potential net beneficial impact of an FTT since it may be expected to diminish the dominance of these few large – too big to fail? – banks.

As in the case of equity trading, very little of the activity in the currency and currency derivative markets is related to commercial enterprise. The BIS reports that over 75 percent of currency trades executed in April 2007 had a time duration of less than one week. Further, the entities that represented the bulk of trading activity were not commercial enterprises engaged in the import and export of goods or services. Instead it is again hedge funds, mutual funds, pension funds and insurance companies that have been the drivers of growth in the currency and currency derivatives markets. “Technical” trading now dominates these markets, much as it does in equity markets. [12] As a result, turnover in the currency markets is almost 25 times the actual value of international trade. [13]

3. Two Arguments Against an FTT

Although there will be very loud opposition to an FTT, it is important to realize that very little of this trading activity that would be impacted has anything to do with generating new economic activity with long term growth potential or even with growing the GDP. Most of the trading that occurs in financial asset markets today is unconnected to actual commercial transactions between companies or to long term investment strategies. As James Tobin observed 25 years ago, “very little of the work of the securities industry, as gauged by the volume of market activity, has to do with the financing of real investment in any very direct way." [14] If some of this trading activity was to disappear because of a very small FTT and there were therefore fewer jobs for equity, currency, or bond traders, it is not at all clear that this would represent a net loss to society. 

Of course, those engaged in trading financial assets as well as many economists who study financial markets will not agree with this conclusion.  We cannot consider in depth all the arguments against an FTT that may be raised but will touch on only two here. [15] These arguments fall largely into two categories. First, many argue that an FTT, by increasing the costs of trading will reduce the level of market activity, lowering the level of liquidity thus increasing transaction costs and market volatility. According to this line of argument, everyone who invests in these markets, whether directly or indirectly, will suffer a loss of economic well being. Second, it is also argued that imposing and FTT will drive business to other venues that lack a tax. [16] 

We believe that neither of these arguments is born out by the weight of the evidence. More importantly, we also believe that any impact of an FTT that reduced the role of finance in the U.S. political economy, shifting resources into other uses would be positive rather than negative. Following is a brief summary of the basis for our conclusion.

First, would an FTT reduce liquidity in the financial markets? As we have illustrated, trading in financial assets has grown dramatically over the past few decades. Opponents of FTTs take the position that more trading is always better; however, if increased trading produces increased liquidity and lower volatility, it should surely have done so over the past two decades.  There is no evidence that volatility has declined; in fact it may actually have increased. In this respect a study by French and Roll 17 is quite suggestive. They compared equity market inter-day volatility in two different circumstances: first, from Tuesday – Thursday when the NYSE was open on the intervening Wednesday and from Tuesday – Thursday during the period when the NYSE was closed on the intervening Wednesday in order for the trade clearing process to catch up with volume. They found that volatility in the second instance was only half that of the former and concluded that trading itself may be a source of volatility. This conclusion is consistent with the notion that herd behavior dominates markets for financial assets, at least in the short run, and generates considerable volatility and resulting higher costs for investors. 

More generally, financial economists judged the U.S. equity market and other markets for financial assets to be efficient when trading levels were less than a fourth of those that prevail today. It is, therefore, hard to imagine a convincing argument that a decrease in trading activity of even 50 percent would result in inefficient markets. Transaction costs would increase slightly with the FTT we propose. However, the actual costs would be less than those prevailing in the 1970s and 1980s, again a time when the U.S. equity market was judged to be efficient. 

On the question of an FTT shifting business to other locations it is useful to consider the London Stock Market (LSE). The LSE has functioned with stamp tax that is much like the FTT we have been discussing for many years. The stamp tax has some flaws – for example it is not levied on derivatives but only on cash equities – but this makes the case even more interesting. The LSE has been one of the largest equity markets in the world for decades and has not grown more slowly than other major markets. In fact, during the 1999 – 2009 period the LSE moved from 3rd place to 2nd place in the world equity tables. Clearly the stamp tax has not moved business to other trading centers. Further, an FTT on US stocks would be applied to the stock without regard to where the trade occurred so long as the issue was listed on a US market. It is also interesting and suggestive that the imposition of stamp tax on the cash equity market has not resulted in any shift of business to the derivative markets in London. Nonetheless, we believe that the logic of an FTT across all related instruments is compelling.

C. Conclusion

We have argued that a financial transaction tax has the potential for raising a significant amount of revenue. An FTT should be designed to apply to all classes or financial assets and to apply across the range of products traded in financial asset markets. At current trading levels an FTT could generate well over $1 trillion in revenue. Spot currency and currency derivative markets would be the largest source of tax revenues with equity the second largest source. Of course if trading were reduced because of such a tax, the revenue raising potential would also be reduced but would still appear to be significant.

An FTT is progressive in impact and would also have the potential to assist in restructuring the U.S. political economy away from an overdependence on finance and financial services. It is widely urged that such shift out of financial activities is a goal but, to date, very little has been done along these lines.

Perhaps more important, however, is the political significance of an FTT. This is cast in terms of Wall Street vs Main Street or it can be articulated as a matter of economic justice. In either case the financial sector would be called upon to return some of the assistance that was rendered to it during the melt down of 2008. This formulation would have immense political appeal and, we think, help the U.S. break out of the politics as usual gridlock into which we have drifted after the initial enthusiasm of the Obama election and the sense that the times were right for change. As our earlier reference to Gordon Brown and Alistair Darling of the UK suggests, we believe that the political appeal of a well structured FTT would also transcend national borders. 

In urging an FTT progressives should, we believe, tie the revenues to programs and policies that are widely beneficial and would be seen to be such. For example, a significant portion of FTT revenues could be earmarked for a jobs program that no simply designed to recover the over 8 million jobs lost since December 2007 but to expand access to good jobs across the U.S. labor force. Revenues from an FTT tied to a  jobs program could also be the basis for an industrial policy that restructured the U.S. economy along the lines of sustained and sustainable growth that increases equality rather than inequality. [17] The possibilities are visionary in scope – what we need now is the political will to move forward.

Notes:

See for example, Lawrence and Victoria Summers, “When Financial Markets work too Well: A Cautious Case for a Securities Transaction Tax,” Journal of Financial Services Research, 3:261-286 (1989) or Dean Baker, “The Benefits of a Financial Transactions Tax,” Center for Economic and Policy Research, Dec 2008. Both of these articles raise excellent points in their discussions of an FTT but both are based on data from a period in which financial markets were significantly smaller than today. In late 2009, Dean Baker, Robert Pollin, Travis McArthur & Matt Sherman have published an assessment of potential FTT revenue using recent market data: “The Potential Revenue from Financial Transaction Taxes,” PERI Working Paper #212.

1.    Data in these two paragraphs are taken from the 1999 Securities Industry Fact Book and the 2008 Annual Report of the World Federation of Exchanges (WFE).
2.    In derivative trading, e.g. options or futures, notional value refers to the value of the underlying instrument represented by the derivatives contract.  For example, if the S&P index is at 1000 and the futures contract is value at $100 x the index, the notional value of each contract is $100,000.
3.    Data on notional value are taken from the World Federation of Exchanges 2009 Annual Report.  See www.world-exchanges.org. 
4.    See BIS: Triennial Central Bank Survey of Foreign Exchange and Derivative Market Activity in 2007 – Final Results.
5.    See WFE, op cit.
6.    William Funk, “On and Over the Horizon: Emerging Issues in US. Taxation of Investments”, Houston Business and Tax Journal.  Funk discusses FTTs on pp 40 – 44.
7.    “Changes in U.S. Family Finances from 2004 to 2007: Evidence from the Survey of Consumer Finances,” Federal Reserve Bulletin, February 2009.
8.    On non-financial corporations move into financial activities, see Robert Pollin and Dean Baker, “Public Investment Policy and U.S. Economic Renewal,” PERI Working Paper #211, esp. pp10 - 12.
9.    Calculated from Securities Industry Association Fact Books, relevant years tables on the NYSE member firm revenues and expenses. 
10.    See BIS, op cit, p. 9.
11.    See BIS, op cit, p. 6.
12.    See BIS, op cit, p. 5.
13.    James Tobin, “On the Efficiency of the Financial System,” Lloyds Bank review, July 1984
14.    See Summers and Summers for a good treatment of many of the arguments against an FTT.  We draw heavily on their discussion. 
15.    See for example, Habermeier and Kirilenko, “Securities Transaction Taxes and Financial Markets” in Debating the Tobin Tax, New Rules for Global Finance Coalition, Washington DC, 2003. 
16.    Kenneth French and Richard Roll, “Stock Return Variances: The Arrival of Information and the Reaction of Traders,” Journal of Financial Economics, 1987, pp.5 – 26, cited in Summers & Summers, op. cit.
17.    For one example of such a jobs program, see “A Permanent Jobs Program form the U.S: Economic Restructuring to meet Human Needs,” at www.cpegonline.org

Photo by Eneas de Troya/cc by 2.0/Flickr