By Phil Mackintosh, Chief Economist, Nasdaq
With Covid-19 playing havoc with federal and state government budgets, and Congressional rescue plans on hold, we’re starting to hear more about Financial Transaction Taxes (FTT) again.
Specifically, New Jersey proposed an FTT of 25 mils per share on all financial transactions processed in the state. Based on trading of 10 billion shares average daily volume (ADV), that adds up to more than $6 billion.
It also seems to apply to bonds and foreign exchange trading. But individual investors, who own around two-thirds of the market via funds, ETFs and direct holdings are most likely to be affected by stock trading charges.
The U.S. equity market is the global leader for a reason
This U.S. stock market has significantly outperformed global markets in the past decade. It remains a leading venue for IPOs and capital formation. That’s been great for U.S. investors and pensions.
It helps that the U.S. market is also the most liquid (Chart 1) and least expensive market in the world. In fact, looking at turnover, U.S. stocks are more than twice as liquid as stocks in other regions.
Chart 1: The U.S. has the deepest, most liquid market in the world
Interestingly, many countries in Europe and Asia-Pacific regions have FTTs, which hurts their liquidity and international competitiveness.
The proposed costs are large
Industry participants debate costs quite frequently, but the reality is U.S. equity market costs are very low compared to other countries.
New Jersey officials claim their proposal would raise around $6 billion each year. But these estimates are likely to be greatly exaggerated. That’s partly because this additional trading cost will potentially double the explicit costs of trading in the U.S. equity markets.
Combined, U.S. exchanges’ revenues are around $1.4 billion each per year, less than one-sixth of this proposed tax. Despite that, the industry has been debating exchange data revenues for 15 years.
Even if we add market makers’ revenues, estimated at around $1.8 billion in an average year and SEC fees of $2.3 billion, the total frictions of trading are still below the total of this proposed tax.
In fact, the tax alone adds up to roughly equal what the asset managers spend on research and trading each year.
Chart 2: Comparing the proposed FTT to market frictions
Importantly, the market is already highly tuned to minimize frictions. A tax of this scale will significantly affect trading and will disproportionately affect those who trade a lot—the market makers and arbitrageurs who provide liquidity to investors and keep cross-market prices efficient.
However, because market makers and arbitrageurs only trade when mispricing totals are greater than their transaction costs, a new tax will translate into less liquidity, worse prices and higher costs for mutual funds and investors. The data below (Chart 4) confirms this too.
FTTs hurt liquidity
Over two dozen (mostly emerging) countries currently have FTTs and at least a dozen more (mostly developed) countries have repealed FTTs. So there is substantial data about how FTTs affect markets.
In fact an IMF Summary of FTT studies concluded that FTTs consistently:
- reduced company valuations
- raised the cost of capital for issuers
- increased bid-ask spreads
- reduced order book depth
- reduced trading volume
- slowed price discovery
We combined the results from many studies, (including Angel, Two Sigma, Tax Foundation, IMF, Wrobel) in Chart 3 below. We see from the downward sloping diagonal of the data that:
- Larger cost changes cause a larger reaction in liquidity (costs up = liquidity down) which ultimately reduces tax revenues
- More exemptions, like market maker exemptions, reduce the impact on liquidity but also limits the tax revenues collected
- Market volumes shifting to substitute locations and products would exaggerate the impact on both
Chart 3: Comparing the impact on liquidity of trading cost changes
Source: Nasdaq Economic Research
For example, when Sweden implemented its FTT, 50% of share volume moved to London almost immediately. Futures volumes fell 98%, bond trading fell 85% and the options market shuttered. The tax collected was a mere 3% of estimates. Sadly, even after the tax was removed, it’s considered by some that Swedish markets never recovered their previous status in Europe.
Some markets exempt market makers in an effort to retain market efficiency. Although this impacts volumes to a lesser degree, it also limits taxes collected. For example, recent EU FTTs in France raised less than one-half of its expected revenue, and in Italy less than one-fifth.
In the U.K., the tax only applies to around 30% of trading. That’s partly because the industry has engineered a contract for difference (CFD) market to help investors minimize trading costs. Arguably, CFDs may have added to systemic risks, including helping to hide multi-billion-dollar ETF losses at SocGen and UBS.
Trading costs also hurt valuations
Increasing trading costs hurts valuations and vice-versa. Research including our own, Amihud, and Brugler show lower costs help, and Lin, Angel, IMF, Wrobel show higher costs hurt. That is because trading costs affect the “after trade” returns of stocks, and stocks with lower returns are given lower valuations.
Summarizing the results of multiple studies in Chart 4 shows the larger the additional costs, the more stocks valuation falls (diagonal slope). That reduces pension funding for all individuals and states.
It also reduces capital gains tax collections, which is one reason why the CBO estimated the first year impact of a U.S.-wide FTT would be negative. In a related study, Vanguard estimates that a 10-basis-points tax translates to 19% of capital gains over 20 years.
Chart 4: Valuations are also affected by trading costs
Increasing trading costs hurts the economy
Lower valuations also increase the cost of capital for companies, which in turn has been shown to cause companies to invest less ultimately reducing GDP. In fact, economists for the European Union estimated a proposed 0.10% transaction tax would lower GDP by 1.76% while raising revenue of only 0.08% of GDP. In other words, the cost to the economy is far greater than the revenue raised.
That’s consistent with U.S. Congressional Budget Office estimates that found a decrease of 1% in the cost of capital leads to a 0.7% increase in investment.
Designed to hurt small businesses and retail traders
Data also shows that the average cost of the tax is much higher for smaller U.S. companies. Those with market capitalizations under $50 million pay a rate that is 36 times higher than large cap companies.
It will also hurt individual investors more than institutions, as they trade smaller cap stocks more. Based on an average trade size of $10,000 shows small cap trades could see more than $5 of FTT per trade.
Chart 5: Small companies are hurt by FTT taxes more than larger companies
A big miscalculation by New Jersey
Clearly, revenues raised are almost certain to be lower than expected, but there are a few reasons for that.
First, volumes this year are elevated by the uncertainties around the Coronavirus pandemic. “Normal” volumes are closer to 7 billion shares each day, or 30% lower than current volume, which would reduce expected revenues to $4.2 billion, but that’s not all they have missed.
Second, based on the reaction to FTT around the world, a more-than-doubling of transaction costs would likely cause trading to fall even further. Data in Chart 3 suggests it could fall around 50%, reducing expected revenues to $2.1 billion.
Finally, the existence of substitutes makes this proposal more like Sweden’s 1984 proposal. Given investors have the choice to trade in another state for better prices, investors will almost certainly send trades outside the Garden State.
Most of the industry already employs “hot-hot” backup exchanges in Illinois, and some have already said they will move there if investors demand it. Best-ex economics says they will. It is possible revenues raised from the New Jersey FTT may land close to zero.
Longer term, it may be hard to bring equity markets back to New Jersey. Chicago is already a major competitor in the financial markets. Almost all futures and many options markets are based in Chicago, thanks to its history as a major commodity trading center located near mid-west farms. Data shows S&P500 futures are already at least as liquid than the whole U.S. stock market. But Chicago also hosts bonds and commodities markets.
Ironically, moving equities to Chicago reduces the need to maintain costly microwave and data connections to New Jersey too.
The point being, once lost, it may not be possible for New Jersey to win equity markets back.
Chart 6: After adding all this research together, it’s likely the costs of the tax outweigh the revenues
It is quite possible, based on our calculations, that New Jersey is making a shortsighted miscalculation.
Over time, some of the over 30,000 finance industry jobs in New Jersey would also likely leave the state. That, in turn, will reduce spending, property valuations and income tax receipts.
It wouldn’t be the first time a financial center moved states for good.
Econ 101 and the damage of Financial Transaction Taxes (FTTs) was originally published by Nasdaq.
