Not FiTT For Purpose
There has been renewed talk recently among the Democratic Presidential candidates about the long proposed Financial Transaction Tax (conveniently we will refer to as FiTT, though more commonly referred to as FTT). Although it is highly unlikely to pass anytime soon with the current President and Republican Senate majority, we thought we would provide our take on why this proposal is not a good idea. It is not like a sales tax, where you don’t like paying it but it is not going to prevent you from making the purchase anyway.
The FiTT proposal states that for derivative transactions, a 0.1% tax would apply to payments from the buyer under the contract. (For cash it would be on the notional value). The size of this tax would have a significant, potentially catastrophic impact on the market. It would be a game changer. Our analysis suggests that an institutional investor holding a futures position long term would be dramatically more impacted than an HFT firm (on a per lot basis).
This is the exact wording of the proposal: “For purchases of derivatives, the tax would be 0.1 percent of all payments actually made under the terms of the derivative contract, including the price paid when the contract was written, any periodic payments, and any amount to be paid when the contract expires.” This is complicated, especially since that higher daily volatility will contribute to paying more tax even if the overall price move is flat over the whole time the position is held.
Here is a quote from one of the proposal’s sponsors, Democratic Senator for Hawaii, Brian Schatz: “Roughly half of the 8 billion daily trades now are high-frequency trades, and that is increasing volatility in the market; it is allowing a certain category of traders to essentially skim profit off the top.”
The intent of the legislators who support this appears to be to target High Frequency Trading firms (HFT) and decimate market volumes. This is a quote from the Congressional Budget Office: “Trading costs for high-frequency traders tend to be very low—in many cases less than 0.1 percent of the value of the securities traded—so this option would generate a notable increase in trading costs for them”. We agree, although explicit trading costs are much lower than 0.1 percent.
Our expertise as a firm is in minimizing transaction costs for our clients. The FiTT would dramatically increase explicit costs as well as increase implicit costs through wider bid/offer spreads and reduced liquidity. The total cost of trading for institutional investors will be higher – indeed, it can be argued that the end investor will be more adversely affected than the HFT firms, which would be a perverse outcome.
Here are some worked examples for instruments that QB executes for our clients:
Futures
Let’s look at two worked examples, each across three major instruments:
HFT Firm:
Buying and holding 1 contract for one day, assuming no price difference between the buy and the closing sell:
| Instrument | Initial Margin % | Notional Value of Contract ($) | Initial Margin Payment ($) | FiTT ($) | FiTT ($) |
|---|---|---|---|---|---|
| Crude Oil (WTI) Futures | 5 | 62,497 | 3,125 | 3.12 | 0.31 |
| E-mini S&P 500 Futures | 5 | 140,054 | 7,003 | 7.00 | 0.56 |
| 10-Year US Treasury Note Futures | 1 | 120,005 | 1,200 | 1.20 | 0.08 |
Notes:
This is for buying 1 December contract in each instrument and closing it out within the day (not held overnight).
The numbers above are an approximation. Actual initial margin is different for speculator vs. hedger classification and may differ from the above.
This assumes no intraday variation margin and therefore includes only the tax due on the initial margin.
Notional Value is the average notional value of 1 December contract on October 1, 2018 and November 19, 2018 along with the March contract on February 15, 2019.
FiTT is the proposed Financial Transaction Tax, levied at 0.1% of total margin payments made by a buyer of the futures contract.
Institutional Investor:
Buying and holding 1 contract from October 1, 2018 to April 24, 2019, including two roll cycles:
| Instrument | Initial Margin % | Notional Value of Contract ($) | Initial Margin Payment ($) x3 | Variation Margin Payment ($) | FiTT ($) | FiTT ($) |
|---|---|---|---|---|---|---|
| Crude Oil (WTI) Futures | 5 | 62,497 | 9,375 | 1,208,468 | 1,218 | 121.78 |
| E-mini S&P 500 Futures | 5 | 140,054 | 21,008 | 556,490 | 587 | 47.00 |
| 10-Year US Treasury Note Futures | 1 | 120,005 | 3,600 | 111,270 | 115 | 7.35 |
Notes:
The numbers above are an approximation. Actual initial margin is different for speculator vs. hedger classification and may differ from the above.
Notional Value is the average notional value of 1 December contract on October 1, 2018 and November 19, 2018 along with the March contract on February 15, 2019.
This is for buying 1 December contract in each instrument on October 1, 2018, rolling to the March contract on November 19, 2018, rolling to the June contract on February 15, 2019 and holding the position until April 24, 2019. (3x initial margin is from entering the original position and two rolls).
The variation margin is calculated from the daily mark to market, summarizing the total notional value of adverse market moves which would constitute a required payment for variation margin. (This does not include variation margin cash received following favourable market moves).
FiTT is the proposed Financial Transaction Tax, levied at 0.1% of total margin payments made by a buyer of the futures contract.
On the run, U.S. Cash Treasuries
0.1% = $1,000 per $1,000,000 notional. That is >100x typical explicit costs for order book execution. Undoubtedly, this will dramatically reduce liquidity and result in even more trading activity transferring to futures as the tax implication is much higher than the equivalent exposure in futures.
So ironically, in futures, the FiTT cost per lot would be orders of magnitude higher for an institutional investor holding a position for 6 months than for an HFT firm. It would be likely that HFT firms would quote these instruments 2 ticks wide, with only occasional natural flow narrowing the bid/offer spread.
The problem is that supporters of FiTT appear to look at high frequency trading (HFT) in a vacuum. HFT is modern day market making, otherwise known as providing liquidity. This is an imperative necessity for any financial market, and a cornerstone of the most liquid, diverse and respected markets in the world, which is what the U.S. is renowned for. HFT serves an important purpose. Yes, they make money but they operate in a business with thin margins, ferocious competition and substantial financial risk. Their presence creates liquid markets with tight bid/offer spreads that benefit investors.
Since QB helps our clients reduce their implicit costs by using our proprietary algorithms, the net effect is that we make it harder for the HFT firms to make money in the trades we are involved with. Two way flow that just crosses the spread is a liquidity provider’s dream. Intelligent algorithms that can capture spread are much more challenging for them. However, whilst we compete with HFT firms on the order book, we wouldn’t be able to do much if they didn’t exist and we appreciate the service that they provide to our markets. Without them, the market simply would not function. 50%* or more of all volume is from HFT firms for a reason – natural buyers and sellers don’t appear at exactly the same time, most of the time.
The other 50% of market participants are mostly institutional investors, who in many cases have very normal people as the beneficiaries of their funds. These end investors will pay a heavy price in the transaction tax itself, and the increased implicit costs caused by the illiquidity (e.g. multi-tick bid/offer spreads). When you think about this comprehensively, it is hard to see whether anyone would be a winner with this proposal, including the U.S. Government who will see repercussions to issuing debt from the greatly increased costs of trading Treasuries.
HFT does not increase volatility in the market. In fact, quite the opposite: greater liquidity reduces volatility. There has been a gross extrapolation and misunderstanding of relatively rare, extreme market events such as “flash crashes” and “flash rallies”. Here is our research on one such example from some time ago, which concluded that HFT was not to blame.
The reduction in liquidity would consequently cause an increase in volatility as the price impact of trades on average would be higher. This itself would lead to more variation margin payments and as a result more tax paid in order to hold a position. This creates a self-destructive cycle; less liquidity, more volatility, more tax. Additionally, market liquidity is asymmetric, it tends to be lower on sharp down days and additional taxes for long holders resulting from these moves would only make it worse. Markets as we know them would not be the same.
*50% is stated as a proxy. We do not have an exact number for the proportion of market volumes attributable to HFT firms. However, the CFTC have updated their report on CME automated trading strategies which is related. You can read that here.
Quantitative Brokers
New York
May 6, 2019