The Great Debate©: High Frequency Trading and Transaction Taxes

by Bradley G. Lewis, Prof. of Economics, Union College and Michael Sankowski, Traders Crucible

This Great Debate© comes from an online discussion that occurred in an MMT discussion forum hosted by Roger Erickson.

Thinking about high-frequency stock trading and transactions taxes on trades

by  Bradley G. Lewis

Two bills (H.S. 3313 and S. 1787) now before Congress would impose a small tax (0.03% of the value of the security) on most purchases of securities and transactions involving derivatives beginning January 1, 2013.  There are some exceptions—e.g., no tax on the initial issuance of stock or debt securities, or on trading in debt instruments (e.g., some Treasury bills) that have fixed maturities of no more than 100 days, or to currency transactions.  The attempted reach of the tax is significant: it would apply to trading within the U.S. but also to transactions done outside the U.S. if any party to the transaction is a U.S. corporation, partnership, or individual.

While raising revenue is arguably one aim, clearly a major purpose of imposing such a tax is to reduce the profitability of certain strategies that are common in high-frequency trading (HFT), which uses systems that can make trades in milliseconds.  Since common strategies using these systems reportedly have profit margins of about 0.01 percent per trade, the transactions tax might eliminate the benefits of these strategies, and possibly the trades themselves.

Would this be good or bad?

To get a preliminary answer to that question, let’s focus on stocks but also acknowledge that both the proponents and opponents are focusing not just on the technical ability of HFT to execute transactions quickly but also on the way in which that speed is currently being used, with impunity.

Why the speed arms race over the last decade?  Because the fastest providers can and do in fact use their advantage largely to accomplish a simple result: HTF exists to use information to “cut in front of the lines” of orders queued up for execution, so to speak:  at the very least, you can get your transactions milliseconds ahead of others in line and collect what amounts to a tax on those who are slower.

It’s easiest to see this by taking a simple but powerful example: if you know several others have entered buy orders at market price, but their systems are slower to execute than yours, you can buy ahead of them, let their orders push the price up higher, and then, using your superior speed, sell what you just bought at the higher price their orders created and take a small profit.  Do it regularly and the pennies add up.  And while your HFT system might cost a lot to build and maintain (an estimated $1.2 – 1.6 billion per year reportedly has been spent in the last five years to improve them) extra trades cost you very little: high volume means high profits.

Opponents of the tax argue that by making these trades unprofitable, the proposed transaction tax would simply chase away trading to friendlier venues, reduce liquidity in U.S. markets, and (in some more apocalyptic views), raise the cost of capital to U.S. companies.

That seems a remarkably gloomy forecast for anyone whose aim in buying stock is to hold it for dividends and possible capital gains. To put it in perspective, that transaction tax of 0.03 percent would add less than a penny (to be exact, 6 tenths of a cent) to the cost of acquiring a share of a stock selling for $20.  Anyone who puts in a market order for such shares would do so knowing that it might move by a few cents.  Even a mutual fund that holds shares for any significant amount of time would, it seems, find that a minor incentive to move traffic overseas.  Would this be likely to chase money out of U.S. stocks and thereby raise the cost of capital to U.S. firms?  Given the order of magnitude, it seems highly unlikely.

To me, it seems at least equally likely that such a tax might reduce average transactions costs, reduce financial market volatility, and benefit longer-term investors in stock.

Let’s consider what some publicly available facts and simple logic would suggest.

1.      Those who use HFT in the ways mentioned are not long-term holders of stock: the strategies only work because they can capture the results of pushing others to the back of the line temporarily, or outguessing other HFT traders by programming and maintaining superior algorithms.  Cash flows to pay for the systems alone are high, and while precise amounts attributable to this kind of trading are hard to come by, some information is very revealing.  On August 8, 2011, when the Dow dropped 635 points and New York Stock Exchange composite volume was the fourth-biggest on record, an estimate by Tabb Group was that HFT traders made $60 million.

2.      At the time, Tabb also estimated that HFT total profits for 2011 would be $5 billion.  Record profits in 2009 for HFT, it says, were about $7.2 billion.  In October, the New York Times estimated that the profits for HFT firms as a group for 2010 and 2011 combined would be $12.9 billion.    I wouldn’t weep for the HFT traders in 2011: $5 – $7 billion profit is a lot of money for a group that doesn’t even want to hold the stock.

3.      There are claims that for large stocks, at least, this system provides “better liquidity” because one can easily buy or sell at any time at a price close to the current one.  But if buying prices for such investors are routinely pushed up and selling prices routinely pushed down, it seems unlikely they are better off.  Recall that the transactions tax would add about 6/10 of a cent to the cost of $20 share of stock.  Could the effects of someone pushing the price up or down via HFT strategies be equal to that?  It seems possible. At least some of this is a social loss—an extra amount of money for a group playing the equivalent of computer games.  Some indeterminate amount, but I suspect nowhere near the total extracted, is a payment for maintenance of liquidity.

4.      While the last few years have seen crises and shocks aplenty, it seems highly unlikely that the HFT day traders have reduced volatility.  Some evidence suggests that they regularly enter the markets when profits from their strategies are easy to find and exit when they are not.  This has little to do with stabilizing the market.

5.      The high costs of playing in this market suggest that much of the money is sticking to fewer hands.  Some HFT traders use relatively simple systems, but the largest players are the ones making the biggest investments in shaving a few more milliseconds off already short times.  Do we prefer ever more concentrated profits if they do not enhance the usefulness of markets?

6.      There are common experiences of extra volatility and poor trade execution that turn out to be worse than when HFT was not used.  In the long run, HFT seems likely to discourage those who actually are interested in holding stocks for their fundamental value.   Discouraging those who normally would hold such stocks certainly would push up the cost of capital for U.S. publicly traded firms.

A transactions tax would not and need not get rid of the quick execution times that are now physically possible, but it would make it much harder to simply trade in and out by cutting to the front of the line.

Could other solutions such as re-regulation or changes to periodic auctions accomplish the same thing?  Maybe, but so far neither has happened. Nor have HFT traders regularly been hauled into court so far for what used to be called front-running (or “first look”) and much of the public assumes they won’t be.  In the absence of changes on either front, the transaction tax is arguably the best option.

Some high-frequency trading provides very efficient service at a low cost and should not be taxed

by Michael Sankowski,

HFT (High Frequency Trading) is really no worse than the old specialist/floor trader system, and probably better for the average retail customer.   HFT is the most logical response to moving exchanges to electronic matching.

I know for a fact – a 100% fact – that the old floor trades would see orders come into the pit and then try to buy all the orders in front of that order. I know this because I did it myself, along side 100’s of other people doing exactly the same thing.

It’s possible to complain about market manipulation by HFT crowd.  But any floor trader at the CME or the CBOT knows the term “Stop Hunt.”

A stop hunt is where the entire pit would band together and push prices in one direction, looking to trigger stop orders above (or below) the market, and make easy money.  We didn’t pass out flyers at the door, but a few conversations and just random market talk, add in some deductive reasoning and common assumptions, and you could guess where stops might be.

I don’t see how this is different than what happens today with the big HFT players.  Of course these players move the market around a bit – essentially, this is their job and how they make money.

However, providing liquidity is valuable.  It’s valuable enough that we should probably not try to reduce liquidity.

Why is liquidity valuable?

I’d say there are five major reasons:

1. Easy, low cost transactions draw legitimate hedgers into the markets

2. Real world users of goods can better estimate their real world costs and earnings

3. Lower cost transactions increase the profits to real world users of products

4. Owners of firms get a better estimate of the value of their holdings

5. Owners of firms can increase or reduce their holdings easily

These are big reasons, and in many cases, HFT provides a legitimate value by providing liquidity at a low cost.

The money collected from providing liquidity is very, very high. Prior to electronic trading, these excess returns were distributed among thousands or even 10’s of thousands of individuals.

The old Eurodollar pit held many hundreds of traders. The rental cost for a badge to wear in the pit was over $8,000 a month at times – so you can imagine what profits had to be in the pit. This doesn’t include the per contract trading fees.

You needed to make $100,000 a year trading just to pay to be able to trade on the floor – that’s the level of profits that come from providing liquidity.

There was a lot of money being made by many, many traders in those days.  There were so many traders it didn’t seem like the profits were all that large when each trader was considered individually.

Fast forward a decade, and instead of thousands of traders and support staff, you have a few dozen HFT firms. Those few dozen firms are dominated by a small handful of firms who do a bulk of the HFT business.

With the job of providing liquidity concentrated in fewer hands, it seems as though these firms are making more money.   I’d say the same total money (or perhaps less) is being made by these fewer firms.  Instead of thousands of floor traders making several hundred thousand  each (totaling a few billion dollars), you have 20 firms responsible for liquidity and sharing  similar revenues.

But the value provided by liquidity is extremely high as well.  I already went through a list of reasons liquidity is valuable.

As a contra indicator, look at bond markets vs. stock markets. Bond markets are far less liquid.  Bond markets aren’t liquid – and that’s where all of our real financial problems are originating.  Part of this is because of lack of liquidity means very few people are looking at specific bonds…

Now why are the profits to liquidity so high?

You have to remember there is a huge disconnect between the needs of a hedger and long term trader and the money involved. This economic disconnect exists no matter what. It’s an economic reality.

Here’s an example that will make it clear what this disconnect means and why providing liquidity is so valuable.

Imagine you have a company who wants to hedge their foreign currency exposure. They get a large purchase order from a German company, and they will get paid in Euros instead of U.S. Dollars. They decide to hedge their exposure of $100m USD in the currency futures markets.

The 0.03% for a hedger of a $100m purchase order in the current markets isn’t any money at all to the company hedging their currency exposure.  During the bidding process, they probably planned on a wide band (say 1% or 2%) for the currency.

But this .03% of $100m is $30,000.  While a company won’t even notice the difference, that’s a ton of money for an individual.

The company is worried about losing $1,000,000 or more. They are glad to get the hedge off at nearly any small price within a rather wide band.

I call this “dumb money” in my head, but a more descriptive term is “indifferent money”.   And there are piles of this kind of money sloshing around the markets every day.

Is there an alternative to the Tobin Tax?

Reducing the amount of high frequency trading can all be done without any transaction tax at all.  Remember, I maintain HFT isn’t a problem in the first place. There are other problems that are worse.

For example, Dark Pools and “first look” systems (front-running) are worse problems and should be regulated out of existence.

But if you want to reduce the amount of HFT without a tax and maintain liquidity, I’d do this:  Switch to an auction system every second, or every 5 seconds. There would still be decent liquidity because there are ways to trade this kind of market and make money.  This comes very close to replicating the old floor based system, and would provide enough trading opportunities for liquidity providers that legitimate market users would be happy with the amount of liquidity.

The exchange world works right now on a first come first served, continuous matching system. What this means is that every order hits the computer and is either immediately executed or placed into a queue that is ranked by price and then time of receipt.

Also, for “power users”, it is possible to see the order book –the resting orders in the market at different price levels.

So it makes sense to fire off thousands of orders per second and cancel them. You can confuse others and gain an advantage.

In an auction setting, this behavior makes less sense.  So implement an auction rather than a transaction tax.


I’d be very careful about the targets we choose. The entire financial industry isn’t evil or even parasitic. Some significant part of it is actually beneficial to our world.  Providing reasonable liquidity is something very helpful and reduces investor costs.  Those who provide the cost reducing liquidity deserve to be compensated for that effort and the risk involved.

Here is why liquidity is so important.  If liquidity goes away, more companies will be private,   like Koch Industries is private.  Private companies have far less scrutiny on them from a securities regulation and public stand point.  Mostly because their wealth derives from a private company, many people in the United States don’t know the Koch Brothers are among the richest people in the U.S. and financially active politically (rabid right wingers).

Taking away the liquidity of the stock market reduces the incentive to go public.  Part of going public is for the owners to cash out today – but another part is so they can retain part of their stake and easily sell it later at a moment of their choice.

Related Article

Will Greece be Colonized? by Bradley G. Lewis

About the Authors

Brad Lewis is Professor of Economics at Union College in Schenectady, NY, where he teaches courses in financial markets and institutions, international trade and finance, monetary economics, and urban redevelopment. He has also held senior-level visiting positions for a term or more at Carleton College and Skidmore College in the United States and at Kansai Gaidai University in Japan. He is a long-time member of and has occasionally co-chaired the Columbia University Seminar in Economic History and is Vice Chair of the Schenectady (N.Y.) Metroplex Development Authority.  Curriculum Vitae.

Mike Sankowski, the host of Econintersect Forex Trader, is a financial writer and trader. His 20 year financial background spans working for hedge funds, trading at the CME, and designing trading and clearinghouse products. His website, is where he comments on monetary theory, economics, and trading. He is the creator of the TC rule for fiscal policy. He is a CFA and CAIA Charterholder. He spends his free time with his family driving the kids to hockey, and is part of the Box Set Authentic musical collective.  Mike is the proprietor of Generate FX which offers data, charting, training and trading signals services for currency traders.

One reply on “The Great Debate©: High Frequency Trading and Transaction Taxes”

  1. how about front running? can we tax that first? especially since it is illegal…let’s start by lobbing a…one trillion dollar tax liability on Bloomberg News for example. We’ll call it “the American Khoderkofsky Act” or something.

Comments are closed.