transaction cost important

7/30/2019 Transaction Cost Important

1/24

Executive Summary

In the wake of the recent financial crisis,several commentators have suggested a trans-action tax on financial markets. The potentialconsequences of such a tax could be hazardousto the financial markets affected as well as to theeconomy. In this paper, we review the relevanttheoretical and empirical literature and applyour findings to estimate the possible impact ofa transaction tax on U.S. futures market activity

as well as its utility as potential tax revenue.We find that the impact of a transaction tax

on market activity (trading volume, bid-askspread, and price volatility) will determine thepotential of such a tax as a source of governmentrevenue. We also find that the current estimat-ed elasticity of trading volume with respect toa transaction tax in the U.S. futures markets ismuch higher than those reported in the extantliterature and those used by the government in

such computation. We show that a transactiontax on futures trading will not only fail to gener-ate the expected tax revenue, it will likely drivebusiness away from U.S. exchanges and towarduntaxed foreign markets.

A review of the literature and estimates con-tained here indicates that there is an inverserelationship between transaction cost (bid-askspread) and trading volume; to the extent that

a transaction tax increases costs, trading vol-umes will likely fall. There is also a positive re-lationship between transaction cost and pricevolatility, suggesting that the imposition of atransaction tax could actually increase financialmarket fragility, increasing the likelihood of afinancial crisis rather than reducing it. Perverse-ly, the imposition of a financial transaction taxcould have results that are exactly the oppositeof those hoped for by its proponents.

Would a Financial Transaction TaxAffect Financial Market Activity?

Insights from Futures Markets

by George H. K. Wang and Jot Yau

No. 702 July 9, 2012

George H. K. Wang is research professor of finance at the School of Management, George Mason University, andformer deputy chief economist at the U.S. Commodity Futures Trading Commission. Jot Yau is Dr. Khalil DibeeEndowed Chair in Finance at the Albers School of Business and Economics, Seattle University.


2/24

2

Financialtransaction

and securitiestransaction

tax proposalshave been

brought up forconsideration byseveral American

administrationsand Congresses.

Introduction

The financial crisis of 20072008 hasraised many concerns and questions aboutfinancial regulation and policymaking. One

of the more popular proposals in the wake ofthe crisis has been to impose financial trans-action taxes (FTTs).1 For example, the Re-public of Korea pushed for an internationallevy on bank transactions at the G-20 meet-ing in 2010, while the International Mon-etary Fund had presented its own bank-taxproposal at the same meeting.2 The Euro-pean Union (EU) has also considered variousFTTs, though a proposal for a bank transac-tion tax was rejected by the EU in 2010.3 InSeptember 2011 the European Commission

proposed a new plan for a pan-EU Tobin taxtaking effect in 2014. It had the backing ofFrance and Germany but outright opposi-tion from Britain.4 Proponents of the finan-cial transaction tax suggest that it can beused effectively as a means to curb excessivefinancial market volatility, stabilize the mar-kets, and raise revenues for various purposes.

Financial transaction and securitiestransaction tax (STT) proposals have beenbrought up for consideration by severalAmerican administrations and Congresses.5

During the fiscal year 1990 budget negotia-tions, the Bush administration proposed abroad-based 0.5 percent tax on transactionsin stocks, bonds, and exchange-traded deriv-atives.6 In 1993 the Clinton administrationproposed a fixed 14-cent tax on transactionsin futures and options on futures.7 TheObama administration has proposed a userfee in the 2012 federal budget on all futurestrading to fund the U.S. Commodities Fu-tures Trading Commission, while 28 mem-bers of the House of Representatives have

co-sponsored legislation that would im-pose a transaction tax on regulated futurestransactions. The proposed tax is 0.02 per-cent of the notional amount of each futurestransaction, to be charged to each party ofthe transaction, with a projected revenue ofhundreds of billions of dollars per year.8

FTTs have a long history in various

forms, dating from Great Britains 1694stamp duty9 to recent Tobin taxes on cur-rency transactions and the latest rejectionof a proposed bank tax in Europe.10 In 1978James Tobin first proposed a tax on foreign

exchange transactions. It aimed to curb ex-cessive speculative volatility as exchangerates freely fluctuated after the collapse ofthe Bretton Woods Agreement, which hadestablished a fixed exchange rate system fea-turing many other countries pegging theircurrencies to the U.S. dollar.11

FTTs have come under severe criticism bylegislators, regulators, and the publicespe-cially the financial services industry and in-vestorsbecause the effects of the taxes areso wide-ranging. Those who oppose FTTs

argue that a tax on securities transactionswould increase price volatility, reduce mar-ket liquidity (trading volume), and decreaseprice efficiency, thus increasing the cost ofcapital and lowering security values.12 In ad-dition, an STT could drive trading in somesecurities to overseas markets not burdenedby tax. Thus, the tax may affect the relativecompetitiveness of taxing countries in globalfinancial markets and would naturally beof particular concern for the U.S. futuresindustry.13 Some pundits warn that FTTs

are easily avoidable and likely to drive finan-cial activity underground, beyond regula-tory oversight.14 FTT proponents argue thatFTTs would increase government revenuesthat could be used for various purposes, in-cluding funding regulatory agencies (e.g., theU.S. Commodity Futures Trading Commis-sion or Securities Exchange Commission)pay back the bailout money to the govern-ment, fund a countrys future budget, or ex-tract a larger contribution from the financialsectors toward funding public goods.15

Transaction tax rates vary with the typeof financial instruments in question (e.g.equities are typically taxed at higher ratesthan debt instruments or derivatives), the lo-cation of trade (i.e., on- or off-exchange, ondomestic or foreign markets), and the statusof the buyer or seller (domestic or foreignresident, market maker, or general trader). As


3/24

3

Extant literaturepresents myriadtheoreticalargumentsin support ofand against atransaction tax.

financial markets have become more global-ized with advances in information and trad-ing technologies, exchanges can now attractmore business by lowering trading costs. Forexample, countries such as Sweden and Fin-

land removed all STTs, while others such asAustralia, Japan, the United Kingdom, andTaiwan lowered their tax rates. John Camp-bell and Kenneth Froot in 1994 referred tothis shift in taxation as an important func-tion of STTs, that they reveal the natureand scope of powerful underlying changesin international capital markets, and offer aglimpse into a future in which governmentpolicy not so much disciplines, but is insteaddisciplined by, competition in modern capi-tal markets.16 That STTs are not common-

place in most countries lends credence to thehypothesis that implementing an STT willhave a negative impact on a given marketsrelative competitiveness.

Before one can properly evaluate the prosand cons of a transaction tax on financialmarkets, one needs to know what potentialimpact an increase in transaction cost wouldhave on trading volume, market liquidity,price volatility, potential revenue, and thegeneral welfare of market participants.

The objective of this paper is to review the

relevant literature on the theoretical ratio-nale for a financial transaction tax as well asempirical evidence that measures outcomesfrom the imposition of such a tax. In this re-view, we concentrate on STTs because (1) thistype of transaction tax has been proposed inEurope, and a proposal is still under consid-eration in the United States, and (2) STTsare the most common FTT that has been ap-plied in other parts of the world, allowing usto analyze the impact of the tax on tradingvolume, volatility, and price efficiency.17

The rest of the paper is organized as fol-lows. In the following section, we review thetheoretical arguments for and against a FTT.Specifically, we review the literature on theanalysis of the imposition of a FTT on trad-ing activities (measured in terms of tradingvolume) and price volatility. Next, we reviewthe empirical evidence on the imposition of

an FTT/STT with regard to trading volume,price volatility, pricing efficiency, and esti-mated revenue. Following that, we discuss amethodology for estimating potential trans-action tax revenue with an example illustrat-

ing the application for estimating potentialtax revenues that can be raised from U.S.futures markets. Finally, we present our con-clusion.

What Does TheorySay about the FTT?

The theoretical arguments for an FTT arebased on rational economic theories, whichassume participants in the markets are all ra-

tional, having complete information aboutfuture prospects. Participants make deci-sions based on the maximization of theirutility functions given the assumed (cost-less) complete information they have. Trans-action costs in financial transactions, in-cluding all kinds of taxes and levies chargedby the authorities, are considered by tradersas they optimize their welfare in these mod-els. However, with different assumptions onthe rationality and composition of marketparticipants and the degree of market effi-

ciency, arguments for and against the FTTboth have reasonable theoretical appeal.

Extant literature presents myriad theo-retical arguments in support of and againsta transaction tax. The arguments all addressthe following basic questions: (1) Does theFTT reduce price volatility? (2) Does the FTTreduce trading volume and market liquidity?(3) Does the FTT affect cost of capital andstock prices? Also, does the FTT cause cor-porate management to emphasize long-termor short-term results relatively more? (4) Will

the FTT cause trading to shift to overseasuntaxed markets and make domestic mar-kets less competitive? (5) Will the FTT raisesubstantial tax revenue?

Reduced Excess Speculation and PriceVolatility

Proponents of an FTT have suggested


4/24

4

Keynes proposeda securities

transaction taxas a means of

mitigating thepredominance

of speculation inthe stock marketduring the Great

Depression.

that a transaction tax may reduce specula-tive trading and excess market volatility.They believe short-term speculative tradingis the source of excess volatility, so by impos-ing a transaction tax, short-term speculative

trading and price volatility will be reduced.The argument rests on the assumptionthat there is a positive relationship betweenshort-term speculation and excess price vol-atility. To FTT proponents, there appear tobe two types of traders in a financial mar-ket: value investors and noise traders. Valueinvestors, also known as fundamentalists,purchase stocks on the basis of comparisonof security price with estimates of funda-mental values. That is, they buy stocks whenmarket price is below the fundamental value

and sell stocks when market price is abovethe fundamental value. This value-basedtrading is assumed to reduce stock pricevolatility by pushing stock prices back to-ward estimates of the worth of the compa-ny. Conversely, short-term noise traders acton the basis of past price movements or theresults of technical analysis of stock marketdata itself. They purchase stocks when mar-ket prices rise and sell the stocks when theyfall. This type of trading, based on positivefeedbacks from market prices, is assumed to

destabilize markets because it often drivesmarket price away from estimates of funda-mental values and thus creates excess pricevolatility. Value investors who trade on thebasis of market price deviations from thefundamental values are long-term inves-tors and have no need to trade frequently.On the other hand, short-term speculativetraders do need to trade frequently becausetheir strategy is to follow recent past pricebehavior. Because the trading frequencyof short term traders is much greater than

that of long-term investors, the impositionof a transaction tax will increase the trad-ing cost for short-term speculative tradersbut will have less impact on the trading costof long-term value investors. As a result, atransaction tax will curb the frequency ofshort-term speculative trading and thus,theoretically, curb excess volatility.

This line of argument for a transac-tion tax can be attributed to John MaynardKeynes. In light of excessive speculationand volatility during the Great Depression,Keynes proposed a securities transaction tax

as a means of mitigating the predominanceof speculation in the stock market duringthe Great Depression.18 Witnessing excessvolatility in the foreign exchange marketsafter the dissolution of the Bretton WoodsAgreement, James Tobin proposed an international transfer tax on currencies in 1978Keyness and Tobins proposals, althoughmade decades apart, were based on thesame assumption that short-term trades arelikely to be more destabilizing to financialmarkets than longer-term trades.19 In sub-

sequent work, Tobin stated that a financialtransaction tax is fundamental valuationefficient since it lowers excess volatility. 20

In 1953, however, Milton Friedman ar-gued that speculation cannot be destabiliz-ing in general; if it were, the participantswould lose money.21 Other advocates of theEfficient Market Hypothesis argue that spec-ulatorsby rationally arbitraging the unex-ploited profit opportunities when a marketbecomes inefficienthelp to clear marketsstabilize prices, and bring the assets and se-

curities back to their fundamental values.22This suggests that the impact of a transactiontax on price volatility may hinge on whethermarkets are dominated by speculators, ar-bitrageurs, or long-term investors. In otherwords, any FTTs success is dependent uponthe composition of traders in the market.

Joseph Stiglitz suggests that a desirableFTT should not impede the functioning ofthe capital market as an allocator of scarceresources. As such, an FTT based on thevalue of the transaction (e.g., a turnover tax

on trades) should be broad-based in order toavoid the frequent introduction of unneces-sary distortions, set at a low rate, and be eq-uitable. Stiglitz further suggests that a suffi-ciently small transaction tax (~ 0.5 percent to1 percent) is negligible and would not affectexchange efficiency, but would have differentimpacts on the welfare of different groups of


5/24

5

Proponents oftransaction taxeargue that theywill discourage

short-termtrading.

traders. He asserts that the uninformed trad-ers are hardly likely to be affected by a tax ata moderate rate of less than 1 percent. Like-wise, the informed traders (insiders) wouldnot be discouraged from trading with that

amount of transaction tax. This is becausethey operate more often on long-term basesfor valuation and thus are unlikely to havetheir behavior greatly affected.23

Stiglitz posits that the turnover tax pri-marily affects short-term market partici-pantsnoise traders and speculatorswhobuy and sell within the trading day and withindays or weeks. As such, a transaction tax mayrepresent a significant fraction of the returnsthey hope to achieve on each transaction. Heargues that the large number of noise trad-

ers and liquidity providers (those who tradewith the noise traders) bear the lions shareof the transaction tax and may actually ex-perience a welfare gain from impeding theseexchanges. There may be greater volatility ifthe FTT is too big (barring arbitrage trades),but this is unlikely if the tax is small. Basedon this logic, Stiglitz establishes that theupper bound on the volatility increase willnot be greater than that without the trans-action tax. He therefore believes that if thetax is small, there will be significant reduc-

tion in volatility as noise traders drop outof the market. Thus, he argues that such atax may actually be beneficial because it dis-courages short-term speculative trading. Thetax wont affect the long-term investors toomuch because a transaction tax, on average,has the property of automatically phasingitself out for long-term investments; that is,as a proportion of returns, it becomes negli-gible as the holding period increases. Thus,the tax will not have a significant effect onlong-term investors. He suggests this feature

makes a turnover tax more desirable than acapital gains tax because a capital gains taxsubsidizes noise traders and penalizes arbi-trageurs, leading to increased price volatility.Lawrence Summers and Victoria Summersalso agree that a securities transactions taximproves the efficiency of financial marketsby crowding out market participants that be-

have irrationally or that waste too many re-sources for this speculative zero-sum game.24

Short-term technical traders are not neces-sarily amateurs or low-volume traders. Port-folio managers, for example, are often evalu-

ated on the basis of quarterly performance.Thus, portfolio managers are incentivizedto maximize performance in the near term.This leads them to give short-term prospectsa disproportionate weight in determiningstock purchases. As a consequence, corpo-rate managers are forced to slight long-terminvestment in favor of delivering short-termearnings.

Proponents of transaction taxes arguethat they will discourage short-term tradingand reduce the number of speculative short-

term traders due to higher trading costsin the markets. More market participantswould, theoretically, look beyond quarterlyearnings reports and short-run prospects,resulting in more stable prices. In this mod-el, corporate managers would pursue morelong-term investment projects.

Reduced Cost of CapitalStock markets allow firms to raise new

capital from shareholders by way of exchange.Thus, a transaction tax that impedes the ex-

change function of the stock market mightinterfere with the capital raising function ofthe market, ironically forcing managementand investors to focus on short-term returnsrather than long-term concerns.25 Stiglitz ar-gues that this potential impact is negligiblefor a small transfer tax and, to the contrary,would enhance the capital-raising functionof the stock market if the tax reduces stockmarket volatility.26 Reducing market volatil-ity will make it easier for firms to raise eq-uity capital at a lower cost, thus increasing

efficiency. If true, management would focustheir orientation toward a longer-term strat-egy.27

Increased Tax RevenueTransaction tax proponents suggest that

the revenue potential of a transaction taxis formidable.28 The Congressional Budget


6/24

6

A small fixedtransaction cost

significantlyreduces trading

volume.

Office (CBO), in its publication Reducingthe Deficit: Spending and Revenue Options, esti-mated the revenue from a broadly based 0.5percent securities transaction tax to be about$12 billion per year based on a five-year aver-

age. Based on the same tax rate used by theCBO, Summers and Summers suggest a sim-ilar figure, estimating government revenuesof at least $10 billion a year.29 Another esti-mate indicates that revenue from a securitiestransaction tax could be as large as $70$100billion per year.30 Outside the United States,it was estimated that a securities transac-tion tax in Japan would bring in $12 billiona year.31 The European Commission in aJune 2011 budget proposal calculated thata financial transaction tax would contribute

50 billion per year to the European budget,or 350 billion over a seven-year period.32

Effects on Trading Volume, MarketLiquidity, and Information Efficiency

Some literature suggests that there is anegative relationship between trading vol-ume and trading costs. Increases in tradingcosts lower the profitability of trading, lead-ing traders to trade less frequently or extendtheir hold period in order to minimize theirtrading costs over time. As trading volume is

reduced, traders will take more time to off-set their trades and face a larger price impactof a given trade, thus diminishing market li-quidity as well. When the market is illiquid,information will be more slowly incorporat-ed into equity or futures prices, impairingoverall market efficiency.

Andrew Lo, Harry Mamaysky, and Ji-ang Wang proposed a dynamic equilibriummodel of asset prices and trading volume.It shows that a small fixed transaction costsignificantly reduces trading volume.33 Even

Stiglitz, a supporter of the transaction tax,agrees that a sizable transaction cost can re-duce trading, thinning market liquidity if thebuy and sell sides are symmetric, althoughfor widely traded stocks, on both theoreticaland empirical grounds, it is hard to believethat this effect [larger bid-ask spread due to atransaction tax] would be significant.34

Franklin Edwards argues that transac-tion taxes increase trading costs, makingU.S. futures markets less competitive be-cause of the impact on price efficiency andon the cost of hedging. He argues that a tax-

induced reduction in trading may decreaseinformational efficiency by discouraginginformation trades by informed specula-tors and hedgers. He admits that it is noteasy to determine the impact on price effi-ciency because the tax also discourages noisetrading.35

Evidence from a pair of studies suggeststhat reducing the transaction tax in the Tai-wan futures market greatly improved theefficiencies of price execution36 and pricediscovery.37 Likewise, a study by Shinhua

Liu examined the impact of a 1989 changein tax rates on securities in Japan. Liu foundsignificant decreases in estimates of the firstauto-correlations in returns for Japanesestocks listed in Japan, but no changes forJapanese stocks dually listed in the UnitedStates as American Depository Receipts(ADRs), which were not subject to the taxlaw change. Liu also found a lower price ba-sis between the ADRs and their underlyingJapanese stocks, concluding that these re-sults are consistent with the hypothesis that

a reduction in transaction costs (transactiontax) improves the efficiency of the price dis-covery process.38

FTTs Do Not Necessarily Reduce PriceVolatility

Donald Kiefer argued that a transactiontax can theoretically increase or decreasevolatility.39 Paul Kupiec demonstrated thata transaction tax has ambiguous effects onprice volatility in a general equilibrium mod-el framework. In the context of his model, he

shows that a transaction tax can reduce theprice volatility of risky assets.40 However, thereduction in price volatility is accompaniedby a fall of the taxed assets price, while con-versely the volatility of risky asset returns wilincrease with the transaction tax. Thus, thenet effect of a transaction tax on price volatil-ity could be to increase it, decrease it, or leave


7/24

7

The net effectof a transactiontax on volatilitydepends on thechange of tradercomposition.

it unchanged, depending on other factors inthe scenario.

In a general equilibrium model, FrankSong and Junxi Zhang examined the effectsof a transaction tax on a set of noise trad-

ers and the resulting market volatility. Theyshowed that a transaction tax may not onlydiscourage the trading activity of noise trad-ers but also discourage rational and stabi-lizing value investors from trading.41 Thenet effect of a transaction tax on volatilitydepends on the change of trader composi-tion that results from the implementationof the tax. They referred to this as the tradercomposition effect.42 Furthermore, a trans-action tax may decrease trading volume andincrease the bid-ask spread. This potential

effect of a transaction tax on liquidity is la-beled as the liquidity effect. The net im-pact of a transaction tax could decrease orincrease market price volatility. The final re-sults depend on the relative magnitude andinteraction of the trader composition andliquidity effects.

Paolo Pellizzari and Frank Westerhoff an-alyzed the effect of a transaction tax on mar-ket price volatility in a number of computa-tional experiments. They showed that theeffectiveness of transaction taxes depends

on the types of trading markets: specifi-cally, they compared a continuous double-auction market versus a dealership market.In a continuous double auction market, theimposition of a transaction tax is not likelyto reduce market volatility since a reductionin market liquidity amplifies the averageprice impact of a given trade order. Liquid-ity is endogenously generated in this type ofmarket. Their model predicts that in a deal-ership market, a transaction tax may reducemarket volatility because abundant liquidity

is exogenously provided, prompting special-ists and some traders to retreat from themarket, causing volume to decline.43

Kang Shi and Juanyi Xu examined the im-pact of a transaction tax on foreign currencytransactions, which was designed to limit theimpact of noise traders in order to reducevolatility.44 They also believe exchange rate

volatility is caused by changes in the relativeshare of noise traders and fundamentalists.In their general equilibrium model, Shi andXu analyzed entry costs for both informedand noise traders after an introduction of

a transaction tax. The model assumed in-formed traders unconditional expectationof excess return depends on the ratio of noiseentrants to informed entrants, but this doesnot influence noise traders expectations. Anincrease in the noise component increasesmarket volatility. In analyzing three equilib-ria with different entry costs to the market,their major finding was that the impositionof a Tobin tax did not reduce volatility andmay, in fact, increase it, depending on the ra-tio of noise to informed entrants. An increase

in market volatility was also an important as-sumption for the impact of a transaction tax,demonstrated in models assuming stochas-tic interaction between agents, who are as-sumed not to be able to influence aggregatevariables.45 In these models, exchange ratevolatility will be low if the market is domi-nated by fundamentalists and will be highif the market is dominated by noise trad-ers. These models reflect the stylized factsof financial markets, most notably, volatil-ity clusteringin which the exchange rate

switches irregularly between phases of highand low volatility.

Overall, the implications of theoreticalmodels on the price volatility effects of anFTT are mixed. Conclusions about the im-pact of a transaction tax on price volatilitydepend on the assumptions of the theoreti-cal models and assumed mechanisms of in-formation transmission. We will examinesome of these in section III.

Increased Costs of Capital and of

HedgingTrading costs affect stock prices because

trading costs reduce the expected return ofstocks. Investors demand higher expectedreturn when paying increased costs.

Yakov Amihud and Haim Mendelsonfound that the expected rate of return onequities (i.e., the cost of equity) is an increas-


8/24

8

The impositionof transaction

taxes willincrease the

trading costs onstocks, and thus

investors willdemand a higher

expected returncommensuratewith the added

cost.

ing concave function of the bid-ask spread,a proxy measure of liquidity.46 Since othertransaction costs of equity trading (e.g., thebrokerage fee) are positively related to thebid-ask spread, the estimates obtained by

Amihud and Mendelson imply that, for agiven increase in the bid-ask spread, expect-ed returns increase at a larger amount as theequity issue becomes more liquid. Likewise,a securities tax that is analogous to a biggerbid-ask spread will raise the expected returnof equity (i.e., the cost of capital). Later stud-ies have also documented two similar re-sults: (1) The greater the liquidity of a givenstock, the lower its expected return,47 and(2) lower trading costs are associated withlower expected return of the stock.48

In 2002 Amihud investigated the effectsof changing overall market liquidity on stockprices over the period from 1963 to 1996. Heobserved that a decline in market liquiditywas accompanied by a significant decline instock prices and subsequent increase in ex-pected return (i.e., the cost of capital).49 Pre-vious studies clearly indicate that the impo-sition of transaction taxes will increase thetrading costs on stocks, and thus investorswill demand a higher expected return com-mensurate with the added cost. As a conse-

quence, firms cost of equity would rise andtheir stock prices will decrease.

Franklin Edwards argued that a declin-ing trading volume due to a transaction taxwould likely increase the risk premiums thathedgers would have to pay to speculatorswho provide liquidity. This makes futuresless efficient risk management instrumentsand thus the FTT undermines one of theprimary economic functions of futures mar-kets.50

Migration of Trading and Relative Com-petitiveness

Previous literature on transaction taxesshed light on the potential adverse effects ofFTTs on the international competitivenessof the U.S. financial services industry. WhileSummers and Summers did not believe atransaction tax would cripple U.S. equities

trading in 1989, they admitted that a trans-action tax could have damaging impacts onthe industry as evidenced in the demise ofthe Sweden Options and Futures Exchangefollowing implementation of a tax on op-

tions.

51

Joseph Grundfest and John Shoven sug-gest that an STT would cause distortions inthe financial markets and could cause manyinvestorsparticularly institutionsto shifttheir equity trading away from organizeddomestic exchanges toward foreign coun-tries. They believe even a small STT can havemajor adverse consequences for the value ofinstruments subject to the tax and for thecost of capital in the U.S. economy. Theyalso criticize the CBOs static model because

it does not consider the STTs effect on trad-ing volume or market prices, nor does itconsider the possibility of substitution awayfrom taxable instruments and transactionsThus, they contend, the CBO overstates theactual tax revenue that can be collected.52

Franklin Edwards has argued that evena very small transaction tax would be suf-ficient to drive all U.S. futures trading tountaxed overseas markets.53 He considereda tax of 0.5 percent on the value of the con-tract to be prohibitively high as a percentage

of total transaction cost on trading as com-pared to stocks.54 Should a lower rate be seton futures trading vis--vis stocks, the differ-ence in the transaction tax may cause tradersto pursue transactions that bear a lower taxSubstitution may take place domestically oracross international borders. This is why Sti-glitz suggested in 1989 that transaction taxrates should be uniform within the UnitedStates on substitutable assets.55

Edwards also pointed out that a criticalfeature of futures markets across the globe is

low transaction costs. If U.S. markets wereto have higher trading costs . . . it wouldbe a relatively simple matter for trading toshift to foreign markets.56 The shift willtake place because: (1) there are restrictionsthat require foreign exchanges to trade thesame contract in order to compete for thesame business (e.g., TAIFEX and Nikkei 225


9/24

9

Investors inSweden movedequity tradingoffshore andfixed incometrading tountaxed localsubstitutes.

indexes were traded on the Singapore Ex-change) and commodity futures are tradedall over the world; and (2) there are no re-strictions on Americans shifting their trad-ing to foreign exchanges. This has been the

main reason why the futures industry op-poses a transaction tax on futures trading.John Campbell and Kenneth Froot ex-

plained that investors in Sweden moved eq-uity trading offshore and fixed income trad-ing to untaxed local substitutes in responseto the countrys imposition of an STT. In theUnited Kingdom, the stamp duty STT stim-ulated trading in untaxed substitute assetsand seemed to reduce total trading volumeto some degree. They concluded that if anSTT is aimed at reducing overall trading vol-

ume or raising revenue, most likely it wontachieve its goal because investors wouldmove trading to offshore markets or un-taxed assets. They believe the British type ofSTT would be more workable for the UnitedStates than the Swedish type.57

What Does theEmpirical EvidenceSay about the FTT?

There are ample empirical studies on theimpact of FTTs on various aspects of finan-cial market quality, including trading vol-ume, volatility, liquidity, and price discov-ery. In general, the literature can be placedinto two groups. The first group of studieshas used direct ex post tests on the impactof transaction taxes on market quality incountries where actual direct taxes, whetherSTTs or Tobin taxes, had been charged onfinancial transactions.

The second group of studies has usedex ante analysis of the impact of a transac-tion tax on the quality of financial marketswhere there have been no actual STT orTobin taxes. The studies use different mea-sures and proxies of transaction costs (e.g.,changes in bid-ask spreads, brokerage com-mission, and tick size58) but not necessarily

the actual transaction tax. Some caveats arein order here. First, although test results onthe hypothesized impact of a transaction taxare useful in explaining its possible impact,the actual impact may differ in practice. An

explicit tax charged on a transaction maynot be perceived as an implicit transactioncost embedded in the bid-ask spread or areduction in brokerage fee. Second, despitecontrolling for variables that could affect theempirical results, the results obtained fromforeign countries where market environ-ments are different may vary considerablywhen and if an FTT is applied to U.S. mar-kets. Third, because most studies are done insecurities and foreign exchange markets andvery few in futures markets, observers should

recognize the limits of similarities betweenthe results and implications obtained fromthose markets and what would happen ifthey were applied to a futures market.

Effects on Trading Volume and MarketLiquidity

One major argument against transac-tion taxes is that a transaction tax wouldincrease trading cost, which would reducetrading volume and market liquidity. A nar-rowly based transaction tax would provide a

strong incentive for traders to migrate to analternative domestic instrument or to un-taxed foreign markets that have lower costs.Furthermore, a reduction in trading volumewould increase trading costs (e.g., a widerbid-ask spread) and decrease market liquid-ity. Market and price efficiency would be im-paired when market liquidity deteriorated.

Several studies provide estimates of theelasticity of trading volume in equity mar-kets. In 1976 Thomas Epps estimated theshare transaction cost turnover elasticity

to be about -0.26 in U.S. equity markets,59whereas Patricia Jackson and Gus ODonnellestimated the transaction cost turnoverelasticity of equities traded in London to be-0.70 nine years later.60 Put simply, previ-ous studies find that transaction costs andtrading volume have a negative relationship.Likewise, empirical studies find that the


10/24

10

Sixty percentof the trading

volume of the11 most activelytraded Swedish

share classes

shifted to theLondon stock

exchange whenthe Swedish

transaction tax onequity increased.

SST had a negative effect on local trading.For example, Steven Umlauf documented in1993 that 60 percent of the trading volumeof the 11 most actively traded Swedish shareclasses, amounting to 30 percent of the total

trading volume, shifted to the London stockexchange when the Swedish transactiontax on equity increased from 1 percent to 2percent in 1986.61 Two econometric studieson equity turnover, one in the United King-dom62 and one in Sweden,63 found that thelong-run elasticity of turnover with respectto overall transactions costs is in the rangeof -1 and -1.7. Their best estimate is -1 (i.e.,for each reduction or removal of 1 percentround trip transaction tax (or 0.5 percent onbuy and 0.5 on sell), trading volume would

increase by 100 percent.64

Taking into consideration the margins

of substitution, Campbell and Froot esti-mated the elasticity of trading volume af-ter changes in transaction taxes in Swedenand found evidence that foreign investorstended to move toward more trading abroadand domestic investors became less likely toengage in any trading at all. They reportedthat when the SST was in place from 198891, the fraction of trading taking place inStockholm was much lower for unrestricted

shares. This is corroborated by further evi-dence that commissions paid by large U.S.institutional investors when trading Swed-ish equities remained constant but the shareof their taxes paid fell from 68 percent in1987 to 13 percent by 1990. That is, foreigninvestors such as U.S. institutions (and theirbrokers) were increasingly able to evade thetax by eliminating the use of Swedish bro-kers when trading in Sweden or by exchang-ing Swedish securities in London and NewYork. They reported that by 1990, 50 percent

of trading volume was shifted to the Londonequity exchange. The authors also foundthat the Swedish tax shifted fixed-incometrading activity from income-securities andfutures markets to untaxed markets suchas variable-rate notes, corporate loans, andforward rate agreements. They contendedthat the Swedish tax had only a marginal ef-

fect on the volume of trade in Swedish eq-uities by foreign institutions. There is littleevidence that total trading volume in Swed-ish stocks responded strongly to changes intaxation of trades in Stockholm. This lends

additional support to the view that interna-tional investors easily evaded Swedish turn-over taxes. They find that the transaction taxhad a larger impact on local fixed-incometrading volume than on stocks. Campbelland Froot offer several observations: (1) Theeffect of the tax seems to be quite large; (2)much of the volume decline in futures oc-curred in anticipation of the tax; (3) these ef-fects ran in reverse once the tax was removedin April 1990. In short, the turnover tax infixed income securities raised little revenue

since substitution toward other Swedish do-mestic securities was easy, with little needfor migration abroad given the existence ofless costly domestic substitutes.

Campbell and Froot proposed two prin-ciples that might be used to rationalizetransaction tax rates across securities. Thefirst principle is that the transactions thatgive rise to the same pattern of payoffsshould pay the same tax, though they admitthat it is conceptually impossible to applythis principle consistently. The second prin-

ciple is that transactions that use the sameresources should pay the same tax. For ex-ample, they point out that Sweden used totax domestic brokerage services, whereas theUnited Kingdom taxes registration (i.e., thestamp duty).

Previous evidence of STT impacts in eq-uity markets shows that a transaction tax re-duces trading volume and market liquidityFor highly elastic instruments, substitutionwill take place, driving some or all tradingto overseas markets where the tax rates are

lower, or out of the market entirely.Likewise, previous studies in futures mar-

kets demonstrated a statistically significantnegative relationship between trading vol-ume and trading costs in U.S., Taiwanese,and Indian futures markets.65 For exampleJohan Bjursell, George Wang, and Jot Yau ex-amined the relations among trading volume,


11/24

11

Transactiontaxes do notnecessarily causevolatility to

decrease.

bid-ask spread, and price volatility in 11 U.S.futures markets from 2005 to 2010. Theyfound that a transaction tax would have thesame effect as a wider bid-ask spread, reduc-ing trading volume, increasing price volatil-

ity, and generating a modest amount of taxrevenue.66

Robert Aliber, Bhagwan Chowdhry, andShu Yan studied the impact of a small trans-action cost (averaged around 0.05 percent)on the trading volume and price volatility offour currency futures traded on the ChicagoMercantile Exchange (CME) over the periodof 19771999. They found that an increaseof 0.02 percent in transaction costs leads to areduction in trading volume as well as an in-crease of volatility of 0.5 percentage points.67

Robin Chou and George Wang foundthat before Taiwan cut the tax on the TaiwanIndex (TIX) futures trading by 50 percent in2000, futures trading volume was smallerthan that in the Singapore futures exchange.However, since July 2002, the trading volumefor TIX exceeded that of the same contracton the Singapore futures exchange.68 Thisevidence indicates that lower transactioncosts change the relative competitivenessof exchanges and significant migration oftrade may take place because of lower trad-

ing costs.In summary, evidence in extant futures

literature is consistent with the anti-taxhypothesis that increasing trading coststhrough a transaction tax would reducetrading volume and market liquidity, andincrease price inefficiency of taxed financialinstruments.

Effects on Price VolatilityThe empirical studies of transaction

taxes impact on price volatility can be clas-

sified into two groups. The first group of pa-pers that examine the relationship betweenprice volatility and trading costs does notfind any definitive pattern or relationship,whereas the second group of papers findsevidence of either an increase or a decreasein volatility. Putting them together, empiri-cal results are inconclusive.

The first group of literature includes thework of Harold Mulherin, who examinedthe relationship between trading costs inthe New York Stock Exchange and the dailyvolatility of the Dow Jones Industrial Aver-

age returns from 1897 to 1987. He conclud-ed that the imposition of a transaction taxmay not necessarily reduce volatility.69 Like-wise, Charles Jones and Paul Seguin used theelimination of the minimal brokerage com-missions in the U.S. stock market in 1975 asa proxy for a one-time reduction in the trans-action tax. They found that volatility fell theyear after the abolishment of the minimumcommission rates in NYSE and AMEX mar-kets, but the decline in volatility was also ob-served in the NASDAQ market.70

Using the cross-sectional data of 23 coun-tries for the period up to, during, and afterthe October market crash (19871989),Richard Roll found no significant evidencethat volatility is negatively related to trans-action taxes.71 In other words, transactiontaxes do not necessarily cause volatility todecrease across countries, and it is question-able whether transaction taxes should beused with confidence as an effective policyinstrument. Shing-yang Hu analyzed nu-merous changes in STT rates in Hong Kong,

Japan, Korea, and Taiwan during the period19751994 but concluded that, on average, achange in STT rates had no effect on volatil-ity and market turnover.72

In one study, Ragnar Lindgren and An-ders Westlund found no significant effect ofan STT on price volatility of Swedish stocks,but in later work they found a weak positiverelationship between STT and price volatil-ity.73 Steven Umlauf examined the impactof increasing the transaction tax on marketprice volatility. He reported several inter-

esting results: (1) there was no significantdifference in volatility across the three taxregimes (i.e., before 1984 [0 percent], 19841986 [1 percent ], after 1986 [2 percent]);74(2) there was a statistically significant in-crease in daily volatility of returns, whichwas higher during the 2 percent tax regimethan in other regimes, but there was no sys-


12/24

12

Most of theprevious

empiricalevidence does not

support the useof a transaction

tax as an effectiveregulatory policy

tool to reduce

market pricevolatility.

tematic relationship between transaction taxregimes and volatility;75 and (3) the volatil-ity of London-traded shares of 11 companieswas lower than the volatility of these com-panies Stockholm-traded classes.76 Robin

Chou and George Wang found that therewere no significant changes in the daily pricevolatility of Taiwan index futures after thetax reduction.77

Among studies that belong to the secondgroup, Shinhua Liu and Zhen Zhu studiedthe deregulation of fixed brokerage commis-sions and the removal of an STT in Japan inOctober 1999, and found results contrary tothose of Jones and Seguin. They found thatthe reduction in transaction costs (in whichan STT was included) increased volatility in

the Tokyo Stock Exchange.78

Liu and Zhuoffered a possible explanation for contradic-tory results: commission rates were drasti-cally reduced in Japan, whereas they werenot in the United States. Hendrik Bessem-binder found that larger tick sizes were as-sociated with higher transaction costs andalso with higher volatility.79 Bessembinderand Subhrendu Rath found that stocksthat had moved from NASDAQ to NYSE,where trading costs were lower, saw a reduc-tion in volatility.80 Harald Hau also found

a positive relationship between transactioncosts and price volatility in the French stockmarket, where significant volatility increaseswere observed when there was an increase inthe cost of trading stocks due to an increasein the tick size.81

Franklin Edwards examined the relationbetween trading volume and volatility for16 U.S. commodity markets during 1989and found no significant relationship be-tween the two.82 He concluded that even if atransaction tax were to succeed in reducing

speculative or short-term trading in futuresmarkets, there was no evidence that it wouldreduce price volatility in either futures orthe underlying spot markets. Based on thethree-equation structural model used byGeorge Wang and Jot Yau in a 2000 study,Bjursell, Wang, and Yau examined the rela-tions among volatility, bid-ask spread, and

trading volume for selected U.S. futurescontracts. Pravakar Sahoo and Rajiv Kumaranalyzed the relations for five most tradedcommodity futures contracts in India. Thesestudies showed that there is a significantly

positive relation between price volatility andbid-ask spread (transaction costs) for eachfutures contract examined.83

Robert Aliber, Bhagwan Chowdhry, andShu Yan studied the impact of a small trans-action cost (averaged around 0.05 percent)on four currency futures traded on CME inthe period of 19771999 on their trading vol-ume and price volatility. They found that anincrease of 0.02 percent in transaction costson the four currency futures traded on CMEleads to an increase of volatility of 0.5 per-

cent points, coupled with a decline in assetprices due to the decline in the demand be-cause of higher transaction costs.84 MarkkuLanne and Timo Vesala found larger trans-action costs impact on foreign exchange ratevolatility between 19921993.85 Badi BaltagiDong Li, and Qi Li investigated the effect ofan increase in the stamp tax on price volatil-ity in the two Chinese stock exchanges usingan event study methodology. They foundmarket price volatility significantly increasedafter the increase in the stamp tax rate.86

Overall, most of the previous empiricalevidence does not support the use of a trans-action tax as an effective regulatory policytool to reduce market price volatility.

Effects on Information Efficiency andPrice Discovery

Robin Chou and Jie-Haun Lee providedinteresting empirical evidence of the effect ofa transaction tax cut on the price efficiencyof the Taiwan Futures Exchange (TAIFEX)and the Singapore Stock Exchange (SGX).87

They demonstrated that after the tax reduc-tion in 1986, the TAIFEX assumed a lead-ing role over the SGX in the price discoveryprocess for index futures contracts. Theyshowed that a reduction in the transactiontax greatly improves the efficiency of priceexecution. Wen-Liang Hsieh also noted thatthe information advantage of the SGX di-


13/24

13

A sizable

transactiontax could havesignificantadverse impactson marketquality.

minished as the TAIFEX lowered its transac-tion tax.88 Badi Baltagi, Dong Li, and Qi Lifound that volatility shocks were not quicklyincorporated into stock prices in the Shang-hai and Shenzhen exchanges once China

had increased its STT from 0.3 percent to 0.5percent in July 1997, suggesting that the in-crease adversely affected the price discoveryfunction of the stock market.89 In contrast,Shinhua Liu showed a reduction in the first-order autocorrelation of Japanese stock pric-es after the STT reduction in 1989, implyingan improvement in the efficiency of the pricediscovery process.

Effects on Potential Tax RevenueTransaction tax proponents argue that

the revenue potential of a transaction tax isformidable.90 The U.S. Congressional BudgetOffice estimates the revenue from a broad-based 0.5 percent securities transaction tax tobe about $12 billion per year over five years.91Based on the same tax rate used by the CBO,Lawrence Summers and Victoria Summersprovided a similar estimate of at least $10 bil-lion a year.92 Robert Pollin, Dean Baker, andMarc Schaberg suggested that revenue froman STT could be as large as $70$100 billionper year.93 Outside of the United States, a

Japanese STT was estimated to bring in $12billion a year in the late 1990s,94 and theEuropean Commission expected an FTT toprovide 50 billion per year over a seven-yearperiod as recently as 2011.95

Edwards doubts that a tax on futurestransactions would potentially generate sig-nificant tax revenue. He argues that the elas-ticity of trading volume in futures marketsis much more than that of equities becauseclose substitutes are easily available in inter-national futures markets, due to the growth

in trading and information technologies.96Thus, he contends that the CBO overesti-mated the revenue from the STT becausethe elasticity of demand (-0.26) used in theestimate was based on an assumption of nosuitable international substitutes. Given amore elastic trading volume in futures, Ed-wards argues that a transaction tax of the

magnitude of 0.5 percent would probablyeliminate all futures trading in the UnitedStates, driving all of those transactionsoverseas. Of course, no revenue would becollected in that case. According to his con-

servative estimate, only negligible revenue (~

$287 million) could be raised from futurestrading even if the lowest tax rate (0.0001percent) and a low demand elasticity (-0.26)were assumed.97 He also computed esti-mated potential tax revenue with a rangeof assumed elasticities (from -1 to -20) andconcluded that a transaction tax on futurestrading would not generate substantial rev-enue.98

George Wang, Jot Yau, and Tony Baptisteprovided the first empirical estimates of the

elasticity of trading volume for several U.S.futures contracts. They documented thatestimates of trading volume elasticity withrespect to trading costs were in the rangeof -0.116 to -2.72, which were less than theelasticities (-5 to -20) used by Edwards, buthigher than the elasticity of -0.26 used inthe CBO report.99 Bjursell, Wang, and Yauprovided empirical estimates of the elastic-ity volume for 11 U.S. futures for the years20052010 in a three-equation structuralmodel.100 They estimated the potential post-

tax trading volume and tax revenue with up-dated cost estimates of trading volume elas-ticity under alternative tax rates. For a 0.02percent tax rate, they found that the tradingvolume for six futures (S&P 500, E-mini S&P500, 10-year T-Note, British pound, soy-beans, and gold) would be totally eliminat-ed, resulting in zero post-tax revenue fromthese six futures.101 They concluded that asizable transaction tax could have signifi-cant adverse impacts on market quality andwould not raise substantial revenue for the

government as suggested in other studies.Moreover, the tax could potentially hurt theinternational competitiveness of U.S. futuresmarkets. A study on Indian futures tradingalso concluded that implementation of atransaction tax would not bring in substan-tial tax revenue.102

Chou and Wang found that the 50 per-


14/24

14

Tradingvolume mayprecipitously

decline in

response toincreased

tax-inducedtrading costs.

cent reduction in the TAIFEX transactiontax rate reduced tax revenue, but the pro-portional decrease in the tax revenue (30percent) was less than the 50 percent reduc-tion in the tax rate. Interestingly, tax revenue

increased in the second and third year afterthe tax reduction when compared to theyear before the tax reduction.103 This sug-gests that tax reduction has no permanentnegative impact on tax revenue.

Finally, some of the literature suggeststhat the burden of transaction taxes on mar-ket participants depends on the availabilityof no-tax substitutes for their instruments.For instance, the elasticity of financial andmetals futures are much higher than thoseof agriculture futures. Thus, the traders of

agriculture futures would have a larger taxburden relative to traders of financial fu-tures.104 A transaction tax would raise therelative cost of hedgers using agriculture fu-tures to hedge against their underlying assetprice risk.

In sum, the review presented above sug-gests the following:

1. There is an inverse relationship be-tween transaction cost (bid-ask spread)and trading volume;

2. There is a positive relationship thatmay or may not be statistically signifi-cant between transaction cost and pricevolatility;

3. There is a positive relationship betweentrading volume and price volatility;

4. Relationships among trading volume,bid-ask spread, and price volatility arejointly determined;

5. Demand for U.S. futures trading is verysensitive (i.e., very elastic) with a strongsubstitution effect between domestic

and untaxed overseas markets; and6. Although estimated potential tax rev-

enue is formidable in the equities mar-kets, tax revenue raised by a transactiontax in the U.S. futures markets wouldnot be as much as many would believebecause the demand for U.S. futures isfound to be very elastic.

Estimation of PotentialTransaction Tax Revenue

One major argument for implementinga transaction tax in security and futures

markets is to raise substantial tax revenuefor the government. However, proponentsof the transaction tax often employ a navemethod to calculate transaction tax revenuemultiplying the tax fee by current aggregatetrading volume in the given markets assum-ing a static model.105 In other words, theirmodels assume the imposition of a tax willnot affect the trading volume in the marketThis assumption can vastly overestimatethe potential tax revenue because it doesnot take into account the relation between

transaction costs and trading volume (seeno. 1, above). Trading volume may precipi-tously decline in response to increased tax-induced trading costs.

William Schwert and Paul Seguin pre-sented a model to estimate tax revenues thataccounts for the impact of the transactiontax on trading volume and price volatility.106

They assumed a flat tax rate, , for all finan-cial transactions. Revenues can be estimatedas follows:

R = (P +P) (Q + Q) +OR, where R is

the revenue, P the volume-weighted averageprice level, Q the quantity of transactions,Pand Q the change in P and Q, respectively,and OR the change in other governmentrevenues associated with the tax. The magni-tude of the decline in trading volume (Q)depends on the elasticity of trading volume,which requires estimation with respect to thepercentage increase in trading costs due tothe transaction tax for each financial instru-ment (see no. 5, above). Likewise, the impactof the transaction tax on prices (P) needs

to be estimated. More importantly, previousstudies have shown that the elasticity of de-mand in the futures market is not likely to bethe same as that in the equities market.107 Assuch, the potential revenue that can be raisedfrom a transaction tax in these two marketscan be substantially different depending onthe differing elasticities (see no. 6, above).


15/24

15

The empiricalrelation betweenprice volatilityand transactioncosts depends onhow transactioncosts affecttrading volume.

It can be inferred that the estimation ofthe elasticity of trading volume is critical tothe accurate estimation of the potential taxrevenue. Previous studies found a strongand significant positive relation between

trading volume/liquidity and price volatil-ity (see no. 2, above), and an inverse rela-tion between transaction costs and tradingvolume/liquidity (see no. 1, above). The em-pirical relation between price volatility andtransaction costs depends on how transac-tion costs affect trading volume, which, inturn, affects the price volatility as theorysuggested. Model specifications of theseprevious studies, however, are incompletebecause trading volume is assumed to be afunction of transaction costs and/or price

per share only.108

Likewise, one study foundthat trading volume was not only a functionof volatility, but also one of open interest,interest rates, exchange rates, and other vari-ables in futures markets.109 Unfortunately, itdoes not include any measure of transactioncosts (such as the bid-ask spread) or transac-tion taxes/fees as an explanatory variable inthe model. In other words, the estimationof the relationship between trading volumeand other explanatory variables is done sepa-rately instead of jointly in a structural model

(i.e., ignoring no. 4, above).In light of the deficiencies in the empiri-

cal estimation of the relation of trading vol-ume and price volatility, and the relation ofbid-ask spread and price volatility, Wang,Yau, and Baptiste proposed a two-equationstructural model to examine the relations be-tween trading volume and transaction costsin seven financial, agricultural, and metalsfutures.110 By estimating the elasticities in asimultaneous-equation system, they explicit-ly formalize the jointly determined relation-

ship between trading volume and bid-askspread. Their study confirms that tradingvolume and bid-ask spread are jointly de-termined in the U.S. futures markets. It alsoshows that the differences in the estimatesfor four U.S. futures markets underestimatethe elasticities and overestimate the poten-tial tax revenues in these markets. Thus, in

estimating the elasticity of trading volumefor the purpose of estimating the potentialtax revenue, one needs to have a model thatallows relevant variables to be jointly deter-mined in estimating the parameters of the

model. However, in Wang, Yau, and Bap-tistes model, price volatility was omitted.It is imperative to estimate the elasticity oftrading volume with respect to the transac-tion tax (bid-ask spread) in a structural mod-el that jointly determines the relationshipsbetween price volatility, bid-ask spread, andtrading volume.111 In 2000 George Wangand Jot Yau proposed a three-equation struc-tural model that allows trading volume andprice volatility to be jointly determined to-gether with transaction costs in the estima-

tion of tax revenues, which has been used inother studies discussed above.112

In 2011, Bjursell, Wang, and Yau provid-ed updated estimates of the trading volumeelasticity to bid-ask spread on the 11 select-ed U.S. futures based on Wang and Yausmethodology.113

In Table 1, the estimates of the elasticityof trading volume with respect to transac-tion costs (proxied by the bid-ask spread) for11 U.S. futures are presented. The estimatesrange from -2.6 (E-mini S&P 500 index fu-

tures) to -0.81 (heating oil), suggesting thatthe trading volume of a futures contract willdecline if transaction costs increase, as by theimposition of an FTT. For example, the elas-ticity of -0.81 for the S&P 500 index futuresindicates that the trading volume for thesefutures will decrease 0.81 percent for eachone percent increase in the bid-ask spreador financial transaction tax. The lower-endof the corresponding interval estimates witha 95 percent confidence level are all greaterthan one, except for 30-year T-bond (-0.972)

and heating oil (-0.923) futures. These resultssuggest that the elasticity of trading volumewith respect to transaction costs had beenvery high during the period 20052010 formost futures examined. Bjursell, Wang, andYau pointed out the important implicationthat an increase in the bid-ask spread due toa new transaction tax would substantially re-


16/24

16

The elasticityused in the

CBOs 1990report seriously

understatescurrent elasticity

in the futuresmarkets.

duce trading volume and decrease liquidityfor the U.S. futures exchanges. The elasticityused in the CBOs 1990 report, -0.26, esti-mated by Thomas Epps based on U.S. stockdata, seriously understates current elasticityin the futures markets. Hence, the CBO over-estimated the potential revenue of a transac-tion tax in futures markets.

We use the following example to illustrate

how Bjursell, Wang, and Yau computed theestimated tax revenue for S&P 500 indexfutures, using a transaction tax of 0.02 per-cent.114

First, we calculate the 0.02 percent trans-action tax revenue on the S&P 500 futurestransactions based on the notional valueof the futures contract, or $283,981 as ap-

proximated by the average yearly price in2010. The transaction tax revenue is thenexpressed as a percentage of the total fixedtransaction costs (TFC), which is $14.8Thus, for S&P 500 futures, the transactiontax revenue as a percentage of the total fixedtransaction costs (TR%TC) is

$283,981 * 0.0002

= 3.837581 or 383.7581%$14.8

Second, the post-tax volume (PTV), i.e.the estimated trading volume after a trans-action tax is imposed, is calculated basedon the current elasticity of trading volume(TV) with respect to transaction costs/taxes(-0.81 for the S&P 500 futures, Table 1) and

Table 1Elasticity of Trading Volume with Respect to Transaction Costs in Selected U.S.Futures Markets

Contract

January 2005December 2010

Elasticity EstimatesaJanuary 1990April 1994

Elasticity Estimatesb

S&P500 -0.81 -0.78

E-mini S&P500 -2.60

30-Year T-Bond -0.87

10-Year T-Note -1.36

British Pound -0.97

Wheat -0.98

Soybean -1.66

Copper -1.44

Gold -2.02 -1.31

Crude Oil -1.00

Heating Oil -0.80

Deutschemark -1.30

Silver -0.90

Source: aC. Johan Bjursell, George H.K. Wang, and Jot Yau, Transaction tax and market quality of U.S. futuremarket: An ex ante analysis, Review of Futures Markets (2012), forthcoming. bGeorge H.K. Wang, and Jot YauTrading volume, bid-ask spread, and price volatility in futures markets,Journal of Futures Markets 20, no. 10(2000): 94370.

Note: Numbers in parentheses denote standard errors for the corresponding point estimates.


17/24

17

Even a smalltransactiontax (e.g., 0.02percent) is big

enough to wipeout all S&P 500index futurestransactions,leaving no taxrevenues to becollected by thegovernment.

the total trading volume (the number ofcontracts traded in 2010). Thus,

PTV = Total TV*[1 + (current elasticity of

TV*(TR%TC/TFC))]

PTV =7,689,961*[1 + (-0.81*3.837581)]

= -16,213,825

Third, the change in trading volume(TV) is computed to be equal to

TV = PTV TVTV = -16,213,825 7,689,961

= -23,903,786

This result shows that the trading volume

of S&P 500 index futures is very sensitive tochanges in transaction costs. Even a small

transaction tax (e.g., 0.02 percent) is bigenough to wipe out all S&P 500 index fu-tures transactions, leaving no tax revenues tobe collected by the government. This resultsuggests that the impact of a transaction tax

on trading costs and trading volume can varysignificantly with different types of futuressince they have different degrees of tradingvolume elasticity.

Finally, the estimated potential tax rev-enues to be collected on various futures con-tracts is calculated based on the post tax vol-ume (column 3, Table 2) estimated with therecent estimates of trading volume elasticity(from Table 1). Column 5 in Table 2 presentsthe estimates of the potential post-tax reve-nue for the eleven futures computed by Bjur-

sell, Wang, and Yau with recent estimates oftrading volume elasticity. The potential tax

Table 2Estimates of Post-Tax Revenue in Selected U.S. Futures Markets

(1) Contract

(2) Average Yearly

Price (2010) ($)

(3) Post Tax

Trading Volume

(4) Post Tax

Revenue Nave

Method ($)a

(5) Post Tax

Revenue Elasticity

Adjusted ($)b

S&P 500 283,981 0 436,760,532 0

E-mini S&P 500 56,776 0 6,305,896,991 0

30-Year T-Bond 124,069 29,208,455 2,072,187,486 724,770,376

10-Year T-Note 121,174 0 7,118,223,079 0

British Pound 96,522 0 583,383,582 0

Wheat 29,512 14,747,726 136,289,676 87,048,101

Soybean 52,434 0 387,315,432 0

Copper 8,572 8,340,933 17,668,989 14,300,463

Gold 122,616 0 1,096,935,043 0

Crude Oil 79,621 0 2,685,649,749 0

Heating Oil 9,033 21,518,357 48,725,003 38,875,710

Total 20,889,035,562 864,994,651

Source: C. Johan Bjursell, George H.K. Wang, and Jot Yau, Transaction Tax and Market Quality of U.S. FuturesMarket: An ex ante Analysis,Review of Futures Markets (2012), forthcoming.Notes: aEstimated potential revenue under this method is computed as Trading volume 2010 x Average YearlyPrice (2010) x Tax rate (0.02%). bEstimated potential revenue under this method is computed as: Post-tax trading

volume x Average Yearly Price (2010) x Tax rate (0.02%).


18/24

18

The transactiontax revenue

estimated by thepre-tax trading

volume or withan unrealistically

low elasticitycan seriouslyoverestimate

the potential taxrevenue.

revenues (PTR) collected from each futurescontract is computed as follows:

PTR = PTV*Average Yearly Price (2010)*Tax rate (0.02%)

For comparison purposes, presented in col-umn 4 in Table 2 are the post-tax revenuescalculated by the nave method, that is, as-suming trading volume will stay the sameand not be affected by the transaction tax.The tax revenue generated by the navemethod (column 4) is often used by propo-nents of transaction tax as the basis for ar-guing that transaction taxes would generatesubstantial revenue.115 For the 0.02 percenttax rate, six futures (S&P 500, E-mini S&P

500, 10-year T-Note, British pound, soy-bean, and gold) would cease to be traded atall in U.S. markets, and would therefore gen-erate zero tax revenue (column 5). The otherfive futures (30-year T-Bond, wheat, copper,crude oil, and heating oil), given their trad-ing volume elasticity, generate tax revenuesthat are less than the corresponding esti-mated tax revenues from the nave method.

There are three noteworthy findings fromthe study by Bjursell, Wang and Yau. First,the magnitude of the decline in the post-tax

volume depends on the relative importanceof the transaction tax to the total fixed costand/or the elasticity of trading volume withrespect to transaction costs on each future.For example, the post-tax trading volume ofthe S&P 500 index futures is reduced to zerowhen the transaction tax is 383.76 percentof the total fixed transaction cost with anelasticity of -0.81. In the soybean case, theelasticity is high (i.e., -1.66) but the post-taxtrading volume still drops to zero even if thetransaction tax is only 65.75 percent of the

total fixed transaction cost.Second, the impact of a transaction tax

on transaction costs and trading volumesvaries significantly with different types offutures. Third, the transaction tax revenueestimated by the pre-tax trading volume orwith an unrealistically low elasticity can seri-ously overestimate the potential tax revenue.

Conclusion

In this paper, we reviewed the theoreti-cal and empirical studies on the impact of atransaction tax.116 Specifically, we reviewed

the empirical evidence on the impositionof a FTT in futures markets with regard totrading volume, price volatility, pricing effi-ciency, and estimated revenue. We discusseda methodology for estimating the potentialtransaction tax revenue that can be raisedfrom U.S. futures markets. The empiricalmodel proposed by the authors and used inseveral previous studies accounts for the en-dogenous relationships among trading vol-ume, bid-ask spread (transaction cost), andvolatility.117 We explained the estimation of

the empirical elasticity of trading volumeand post-tax adjusted trading volume us-ing Bjursell, Wang, and Yaus estimates on11 futures traded in the United States. Weshowed that current estimates of the elastic-ity of trading volume with respect to a trans-action tax in U.S. futures markets are muchhigher than those reported in the extant lit-erature and those used by the government intransaction tax revenue estimation. As such,a transaction tax would reduce trading vol-ume significantly, may not reduce price vola-

tility, and might only raise a modest amountof tax revenue, much smaller than expected.More importantly, results indicate thatwith such high estimates of trading volumeelasticity, it is very likely that futures trad-ing activities would be shifted to untaxedforeign markets should a transaction tax beimposed. We conclude that a transaction taxon futures trading will not only fail to gen-erate the expected tax revenue, it will likelydrive business away from U.S. exchangesand toward untaxed foreign markets.

NotesWe would like to thank Mark A. Calabria, direc-tor of financial regulation studies at the CatoInstitute, for encouraging us to write this paperThe views expressed here do not necessarily re-flect those of our current or former employers.


19/24

19

1. Financial transaction taxes (FTTs) can beclassified into (1) securities transaction taxes(STT); (2) currency transaction taxes (CTT or To-bin tax); (3) capital levy or registration taxes; (4)bank transaction taxes (BTT); and (5) real estatetransaction taxes. Thornton Matheson, TaxingFinancial Transactions: Issues and Evidence,

International Monetary Fund working paper,WP11/54, 2011. The term Tobin tax originallyreferred to the tax on currency transactions (i.e.,CTT). It is now used interchangeably with finan-cial transaction taxes, as in this paper.

2. Evan Ramstad, World News: Seoul WillSeek Support for Levy, The Wall Street Journal,

June 4, 2010, p. A14.

3. Jason Zweig, Flash Tax: Why Levies onHigh-Speed Trading Wont Work, The Wall Street

Journal, September 34, 2011, p. B11.

4. Alex Barker, George Parker, and Brooke Mas-

ters, Fresh Clashes Brew Over Tobin Tax,Finan-cial Times, January 5, 2012, p. 4.

5. Joseph A. Grundfest and John B. Shoven,Adverse Implications of a Security TransactionExcise Tax, Journal of Accounting, Auditing, and Fi-nance 6 (1991): 40942. Grundfest and Shovenbelieve the STT is a resilient proposal, p. 410. G.William Schwert and Paul J. Seguin, SecuritiesTransaction Taxes: An Overview of Costs, Bene-fits and Unresolved Questions,Financial Analysts

Journal49 (SeptemberOctober 1993): p. 28.

6. Donald W. Kiefer, The Securities Transac-

tion Tax: An Overview of the Issues, CRS Reportfor Congress (1990).

7. General Accounting Office, GGd-93-108-Futures Transaction Fee, p. 1.

8. Kathleen Cronin, A Transaction Taxs Unin-tended Consequences, 2010, http://archive.opnmkts.com/regulatory/a-transaction-taxs-unintended-consequences/; E. Noll, The Perils of aFinancial Services Transaction Tax, 2010, http://www.rollcall.com/news/-42770-1.html.

9. The stamp duty in the U.K. is a levy on thetransfer of assets and securities. Richard Roll,

Price Volatility, International Market Links, andTheir Implications for Regulatory Policies, Jour-nal of Financial Services Research 3 (1989): 21146;Lawrence H. Summers and Victoria P. Summers,When Financial Markets Work Too Well: A Cau-tious Case for a Security Transaction Tax, Jour-nal of Financial Services Research 3 (1989): 26186;

John Y. Campbell and Kenneth A. Froot, Inter-national Experiences with Securities TransactionTaxes, in Jeffrey A. Frankel, ed. The Internation-alization of Equity Markets (Chicago: University of

Chicago Press, 1994), pp. 277308.

10. The European Commission rejected a pro-posal for a bank tax to pay for the bailout offailed European banks due to the 20072008 fi-nancial crisis.

11. James E. Tobin, A Proposal for Internation-al Monetary Reform, Eastern Economic Journal4,no. 3 (1978): 1539.

12. For a review of arguments, see Schwert andSeguin, pp. 2735; Paul H. Kupiec, Patricia A.White, and Gregory Duffee, A Securities Trans-action Tax: Beyond the Rhetoric, Research in Fi-nancial Services Private and Public Policy 5 (1993):5576; R. Glenn Hubbard, Securities Transac-tion Taxes: Can They Raise Revenue? MidAm-erica Institute Research Project on TransactionTax, July 1993; Summers and Summers; StevenR. Umlauf, Transaction Taxes and the Behaviorof the Swedish Stock Market,Journal of Financial

Economics 33 (1993): 227240; and Neil McCull-och and Grazia Pacillo, The Tobin TaxA Re-

view of the Evidence, working paper, Instituteof Development Studies, University of Sussex,2010.

13. See Franklin R. Edwards, Taxing Transac-tions in Futures Markets: Objectives and Effects,

Journal of Financial Services Research 7 (1992): 7591;and George H. K. Wang, Jot Yau, and Tony Bap-tiste, Trading Volume and Transaction Costs inFutures Markets,Journal of Futures Markets 17, no.7 (1997): 75780.

14. Charlemagne: The Seven-Yearly War,Econ-omist, June 30, 2011.

15. Kiefer.

16. Campbell and Froot, p. 277.

17. Previous studies present summaries for vari-ous types of financial transaction taxes aroundthe world over different time periods. See RobertPollin, Dean Baker, and Marc Schaberg, Secu-rities Transaction Taxes for U.S. Financial Mar-kets, Eastern Economic Journal 29, no. 4 (2003):52758; McCulloch and Pacillo; Schwert and Se-guin; Roll; and Matheson.

18. John Maynard Keynes, The General Theory ofEmployment, Interest, and Money (New York: Har-court Brace, 1936).

19. James Tobin, A Proposal for InternationalMonetary Reform.

20. James E. Tobin, On the Efficiency of Fi-nancial Systems, Lloyds Bank Review 153 (1984):115.


20/24

20

21. Milton Friedman, Essays in Positive Economics(Chicago: The University of Chicago Press), 1953.

22. See Eugene F. Fama, The Behavior of Stock-Market Prices, Journal of Business 38 (January1965): 3105.

23. Joseph E. Stiglitz, Using Tax Policy to CurbSpeculative Short-Term Trading, Journal of Fi-nancial Services Research 3 (1989): 101115.

24. Summers and Summers.

25. Keynes.

26. Stiglitz.

27. Sanford J. Grossman and Joseph E. Stiglitz,Information and Competitive Price Systems,

American Economic Review 66 (May 1976): 24653;Sanford J. Grossman, and Joseph E. Stiglitz, Onthe Impossibility of Informationally Efficient

Markets, American Economic Review 70 (June1980): 393408. Grossman and Stiglitz showedthat the stock market provides a mechanismthrough which information is aggregated fromboth uninformed and informed individuals inthe market. They argued that stock prices playno informational role in a firms investment deci-sionmaking and are not likely to play any majorrole in the firms investment decision. They sug-gested that managers do not base their invest-ment decisions on stock prices, although stockprices provide signals to investors that may beread by the managers in the same way. Summersand Summers make the same argument.

28. For examples, see Kiefer; Congressional Bud-get Office, Reducing the Deficit: Spending and Rev-enue Options (Washington: Government PrintingOffice, 1990); and Pollin, Baker, and Schaberg.


30. Pollin, Baker, and Schaberg made the esti-mate based on 1997 levels of market activity forstocks, bonds, and swaps, and the March 1999level of market activity for futures and options.

31. Roll.

32. Raman Uppal, A Short Note on the TobinTax: The Costs and Benefits of a Tax on Finan-cial Transactions, working paper, EDHEC-RiskInstitute, 2011.

33. Andrew W. Lo, Harry Mamaysky, and JiangWang, Asset Prices and Trading Volume underFixed Transactions Costs,Journal of Political Econ-omy 112 (2004): 105491.

34. Stiglitz, p. 12.

35. Edwards.

36. Robin K. Chou and Jie-Haun Lee, The Rela-tive Efficiencies of Price Execution between theSingapore Exchange and the Taiwan FuturesExchange, Journal of Futures Markets 22 (2002)17396.

37. Wen-Liang G. Hsieh, Regulatory Changesand Information Competition: The Case of Tai-wan Index Futures, Journal of Futures Markets 24(2000): 399412.

38. Shinhua Liu, Securities Transaction Taxand Market Efficiency,Journal of Financial Service

Research 32 (2007): 16176.

39. Kiefer.

40. Kupiec, White, and Duffee, pp. 5576.

41. Frank M. Song and Junxi Zhang, Securities

Transaction Tax and Market Volatility,EconomicJournal115 (2005): 110320.

42. Ibid., p. 1105.

43. Paolo Pellizzari and Frank WesterhoffSome Effects of Transaction Taxes under Dif-ferent Microstructures, Journal of Economic Behaviour and Organisation 72, no. 3 (2009): 85063.

44. Kang Shi and Juanyi Xu, Entry Cost, theTobin Tax, and Noise Trading in the Foreign Ex-change Market, Canadian Journal of Economics/

Revue Canadienne dEconomique 42, no. 4 (2009)150126.

45. See generally, Alan Kirman, Epidemics ofOpinion and Speculative Bubbles in FinancialMarkets, in M. Taylor (ed.), Money and Financia

Markets (New York: Macmillan, 1991); ThomasLux and Michele Marchesi, Scaling and Criti-cality in a Stochastic Multi-Agent Model of aFinancial Market, Nature 397 (1999): 498500and Thomas Lux and Michele Marchesi, Vola-tility Clustering in Financial Markets: A Macro-Simulation of Interacting Agents, Internationa

Journal of Theoretical and Applied Finance 3, no. 4(2000): 675702.

46. Yakov Amihud and Haim Mendelson, AssetPricing and the Bid-Ask Spread,Journal of Financial Economics 17, no. 2 (1986): 22349.

47. Vinay T. Datar, Narayan Y. Naik, and RobertRadcliffe, Liquidity and Stock Returns: An Al-ternative Test, Journal of Financial Markets 1, no2 (1998): 20319.

48. Michael J. Brennan and Avanidhar Sub-rahmanyam, Market Microstructure and Asset


21/24

21

Pricing: On the Compensation for Illiquidity inStock Returns, Journal of Financial Economics 41(1996):44164.

49. He used market impact cost as the measureof liquidity in the study. Yakov Amihud, Illiquid-ity and Stock Returns,Journal of Financial Markets

5 (2002):3156.

50. Edwards.


52. Grundfest and Shoven, p. 411.

53. Countries have indicated their concernabout a transaction tax that is not global andare aware that relative competitiveness may bechanged by the burden of a nonglobal transac-tion tax. See Ramstad and Zweig, in which he dis-cusses the proposal for a global transaction taxon flash trading.

54. Edwards, Table 2, p. 84.

55. Stiglitz, p. 14.

56. Edwards, p. 85

57. The above theoretical arguments for a FTTare incorporated in different types of theoreticalmodels: traditional, game theoretical, and zerointelligence agent models. For traditional theo-retical models, see Paul H. Kupiec, A SecuritiesTransaction Tax and Capital Market Efficiency,Contemporary Economic Policy 12, no. 1 (1995):10112; Song and Zhang; Pellizzari and Wester-hoff; Charles Noussair, Stephane Robin, and Ber-nard Ruffieux, The Effect of Transaction Costson Double Auction Markets, Journal of Economic

Behavior and Organization 26, no. 2 (1998): 22133;Joel Hasbrouck and Gideon Saar, Limit Ordersand Volatility in a Hybrid Market: The IslandECN, Stern School of Business Working PaperFIN01-025, 2002; and Adrian Buss, Raman Uppal,and Grigory Vilkov, Asset Prices in General Equi-librium with Transactions Costs and RecursiveUtility, working paper, Ecole des Hautes EtudesCommerciales and Goethe University, Frankfurt,2011. For neo-traditional models, see MarkusHaberer, Might a Securities Transaction Tax

Mitigate Excess Volatility? Some Evidence fromthe Literature, CoFE Discussion Paper 04-06,Center of Finance and Econometrics, Universityof Kontanz, 2004; Shi and Xu; Kirman; Lux andMarchesi, Scaling and Criticality in a StochasticMulti-Agent Model of a Financial Market; FrankH. Westerhoff, Heterogeneous Traders and theTobin Tax, Journal of Evolutionary Economics 13,no. 1 (2003): 5370; Frank H. Westerhoff, and Ro-berto Dieci, The Effectiveness of Keynes-TobinTransaction Taxes When Heterogeneous Agents

Can Trade in Different Markets: A Behavioral Fi-nance Approach,Journal of Economic Dynamics andControl30, no. 2 (2006): 293322; and Thomas I.Palley, Speculation and Tobin Taxes: Why Sandin the Wheels Can Increase Economic Efficiency,

Journal of Economics 69, no. 2 (1999): 11326. Forgame theoretical models, see Johannes Kaiser,

Thorsten Chmura, and Thomas Pitz, The TobinTaxA Game Theoretical and an ExperimentalApproach, working paper, University of Bonn,Germany, 2007; Ginestra Bianconi et al., Effectsof Tobin Taxes in Minority Game Markets, Jour-nal of Economic Behaviour and Organisation 70, no.12 (2009): 23040. For zero intelligence agent mod-els, see McCulloch and Pacillo; G. Ehrenstein, F.Westerhoff, and D. Stauffer, Tobin Tax and Mar-ket Depth, Quantitative Finance 5, no. 2 (2005):21318; Katiuscia Mannaro, Michele Marchesi,and Alessio Setzu, Using an Artificial FinancialMarket for Assessing the Impact of Tobin-LikeTransaction Taxes, Journal of Economic Behaviorand Organization 67, no. 2 (2008): 44562; and J.Doyne Farmer, Paolo Patelli, and Ilija I. Zovko,The Predictive Power of Zero Intelligence in Fi-nancial Markets, Quantitative Finance Papers 102,no. 6 (2006): 225459.

58. A tick size is the smallest increment (tick) bywhich the price of financial instrument can move.

59. Thomas W. Epps, The Demand for Broker-age Services: The Relation between Security Trad-ing Volume and Transaction Cost, The Bell Jour-nal of Economics 7 (1976): 192.

60. P. Jackson and A. ODonnell, The Effects of

Stamp Duty on Equity Transactions and Prices inthe UK Stock Exchanges, Discussion Paper no.25, Bank of England, 1985, p. 14.

61. Umlauf, pp. 22930.

62. Jackson and ODonnell, p. 14.

63. Campbell and Froot, p. 298.

64. Jackson and ODonnell; Campbell and Froot.

65. See generally Wang, Yau, and Baptiste; GeorgeH. K. Wang and Jot Yau, Trading Volume, Bid-AskSpread, and Price Volatility in Futures Markets,

Journal of Futures Markets 20, no. 10 (2000): 94370;Robert Z. Aliber, Bhagwan Chowdhry, and ShuYan, Some Evidence That a Tobin Tax on ForeignExchange Transactions May Increase Volatility,

European Finance Review 7 (2003): 481510; RobinK. Chou and George H. K. Wang, Transaction Taxand Market Quality of the Taiwan Stock Index Fu-tures,Journal of Futures Markets 26, no. 12 (2006):11951216; C. Johan Bjursell, George H. K. Wang,and Jot Yau, Transaction Tax and Market Qual-ity of U.S. Futures Market: An Ex Ante Analysis,


22/24

22

Review of Futures Markets (2012), forthcoming; andPravakar Sahoo and Rajiv Kumar, The Impact ofCommodity Transaction Tax on Futures Tradingin India: An Ex-Ante Analysis, Singapore Economic

Review 56, no. 3 (2011): 42340.

66. Bjursell, Wang, and Yau.

67. Aliber, Chowdhry, and Yan.

68. Chou and Wang, p. 1210.

69. J. Harold Mulherin, Regulation, Trad-ing Volume and Stock Market Volatility, Revue

Economique 41, no. 5 (1990): 92337.

70. Charles M. Jones and Paul L. Seguin, Trans-action Costs and Price Volatility: Evidence fromCommission Deregulation, American Economic

Review 87 no. 4 (1997): 72837.

71. Roll, p. 221.

72. Shing-Yang Hu,The Effects of the StockTransaction Tax on the Stock MarketExperi-ences from Asian Markets, Pacific-Basin Finance

Journal6 (1998): 35860.

73. Ragnar Lindgren and Anders Westlund,Transaction Costs, Trading Volume, and Price

Volatility on the Stockholm Stock Exchange,Skandinaviska Enskilda Banken Quarterly Review 2(1990): 3035.

74. Steven R. Umlauf, Transaction Taxes andthe Behavior of the Swedish Stock Market, Jour-nal of Financial Economics 33, no. 2 (1993): 22740.

75. Ibid., p. 233.

76. Ibid., p. 239.

77. The Taiwanese government reduced the taxlevied on futures traded on the Taiwan futuresexchange from 0.05% to 0.025% on May 1, 2000.Chou and Wang, p. 1212.

78. Shinhua Liu and Zhen Zhu, TransactionCosts and Price Volatility: New Evidence from theTokyo Stock Exchange,Journal of Financial Service

Research 36 (2009): 7577.

79. Hendrik Bessembinder, Tick Size, Spreads,and Liquidity, Journal of Financial Intermediation9, no. 3 (2000): 233.

80. Hendrik Bessembinder and SubhrenduRath, Trading Costs and Return Volatility: Evi-dence from Exchange Listings, working paper,University of Utah, 2002.

81. Harald Hau, The Role of Transaction Costs

for Financial Volatility: Evidence from the ParisBourse, Journal of the European Economic Association 4, no. 4 (2006): 86290.

82. Edwards.

83. Wang and Yau; Bjursell, Wang, and Yau; Sa

hoo and Kumar, pp. 42340.

84. Robert Z. Aliber, Bhagwan Chowdhry, andShu Yan, Some Evidence that a Tobin Tax onForeign Exchange Transactions May Increase

Volatility,Review of Finance 7, no. 3 (2003): 498.

85. Markku Lanne and Timo Vesala, The Effectof a Transaction Tax on Exchange Rate Volatil-ity, International Journal of Finance and Economics15, no. 2 (2010): 12333.

86. Badi H. Baltagi, Dong Li, and Qi Li, Transaction Tax and Stock Market Behavior: Evidencefrom an Emerging Market, Empirical Economics

31 (2006): 393408.

87. Chou and Lee, The Relative Efficiencies ofPrice Execution between the Singapore Exchangeand the Taiwan Futures Exchange.Journal of Futures Markets 22 (2002): 173196.

88. Hsieh.

89. Badi Baltagi, Dong Li, and Qi Li, Transaction Tax and Stock Market Behavior: Evidencefrom an Emerging Market,Empirical Economics 31no. 2 (2006): 405406.

90. For summaries of the arguments, see KieferCongressional Budget Office; Pollin, Baker, andSchaberg.

91. Congressional Budget Office, p. 12.

92. Summers and Summers, p. 285.

93. The estimate was based on 1997 levels ofmarket activity for stocks, bonds, and swaps, andthe March 1999 level of market activity for futuresand options. Pollin, Baker, and Schaberg, p. 528.

94. Stiglitz, p. 211. Summers and Summers sug-gested the same amount (p. 276).

95. Uppal, p. 5.

96. E

transaction cost important

Documents