!(../../../../../../../../../../var/folders/34/zq18d8kx7kbgby0j06p_j6t40000gn/T/TemporaryItems/NSIRD_screencaptureui_EM2XPo/Screenshot 2022-01-04 at 17.01.16.png)
- Financial Markets
The Global Financial Crisis revived the idea of using transaction taxes to discourage short-term speculative trades. Such trading is often blamed for causing excess volatility in financial markets. Tobin (1978) proposed the tax more than 40 years ago, to “throw some sand in the wheels of speculation”, specifically for currency trading. The idea has been extended to all forms of financial transactions.
In September 2011, the European Commission proposed a harmonised Financial Transaction Tax for the EU, for the purpose of stabilising financial markets and raising additional tax revenue from financial institutions. The proposal is being implemented by 11 EU member states, whereas other member states do not share the belief that the tax is beneficial.
Economists are also divided on the stabilising effect of raising financial transaction costs, often referred to as Tobin’s tax. Recent work in behavioural finance demonstrates that uninformed short-term trading may drive prices away from fundamentals and raise price volatility. By curtailing noise trading, Tobin’s taxes can reduce excess volatility (e.g. Stiglitz 1989, Summers and Summers 1989). On the other hand, Subrahmanyam (1998) and Amihud and Mendelson (2003) show theoretically that a financial transaction tax could reduce the amount of informed trading and harm market quality.
Empirical evidence on the market-stabilising effect of Tobin’s tax is not conclusive. Studies that report statistically-significant evidence generally find Tobin’s tax to be associated with increased volatility (Umlauf 1993, Jones and Seguin 1997, Hau 2006, Pomeranets and Weaver 2011). However, it is unclear whether this increased volatility is due to the withdrawal of informed traders.
The application of transaction taxes goes beyond the trading of financial instruments. For example, many Asian countries have used stamp duty to ‘cool down’ property markets. While there are noticeable differences between real estate and financial markets, the effect of stamp duty on real estate markets can shed light on the possible effects of a transactions tax on financial markets. In a recent paper, we analyse the change in stamp duty policy in Singapore in 2006 (see Fu et al. 2013).The stamp-duty concession in Singapore
In Singapore, developers can sell their condominium before completion. These are usually referred to as pre-sale transactions, as opposite to resale transactions that pertain to buying and selling of completed (and often used) condominium units. In the wake of the Asian Financial Crisis in the late 1990s, the Singapore government gave a stamp duty concession to homebuyers in uncompleted new condominium projects. Buyers could defer the payment of the stamp duty – 3% of the home price – until the property was completed (typically three to four years from the launch of the project) or sold, whichever was sooner. The intention was to stimulate demand for new residential development. The policy measure had no impact on resale transactions, i.e. the spot market.
Pre-sale contracts – sale and purchase agreements for uncompleted properties – could be acquired from developers with a downpayment of 10–20% and subsequently traded in the pre-sale market. Speculators could ‘flip’ them – buy and then sell them at a profit before the properties were ready for occupancy. Note that there are no property taxes or management fees for uncompleted homes, unlike in completed projects. Given their low-cost nature, pre-sale contracts have been popular among short-term speculators in the property market. Thus, the stamp duty deferral substantially reduced the investment commitment and the transaction cost for speculators.
To dampen property speculation, in December 2006 the government withdrew the stamp-duty concession without warning – obliging buyers of uncompleted homes to pay stamp duty upfront. The removal of the stamp duty concession significantly increased transaction costs for short-term speculators in the pre-sale market. As before, the policy did not affect buyers of completed homes.Empirical investigation
To investigate the policy impact, we compare the changes in the pre-sale market (the treatment group) to changes in the spot market (the control group) using a differences-in-differences approach. We focus on two key measures – the trading turnover rate and volatility.
The trading turnover rate is defined as the number of transactions in a condominium project in a given month, scaled by the total number of units in that project. Volatility is a monthly measure for individual pre-sale condominium projects (each involving 40 to over a thousand apartment units). It is defined as the difference between the highest and lowest unit transaction prices in natural logs, after adjusting for the regional market price level, macro factors as captured by time dummies, differential hedonic attributes among the apartments, and time to completion pertaining to pre-sale contracts, within each project in a month.
This ‘range’ volatility measures the idiosyncratic variation in transaction prices across units in the same project. The range volatility could be due to the arrival of new information about the asset fundamentals, such as a change in the location’s appeal, the builder’s reputation, or the development of competing projects or complementary infrastructures. It could also be due to noisy speculation. Given the hedonic factor adjustments we made, the incremental change in the volatility measure of the treatment group (relative to that of the unaffected neighbouring projects) is likely attributable to the policy’s effect on noise trading.
We first find that the total trading activity decreased significantly in the pre-sale market – but not in the spot market – after the transaction cost increase. However, the volatility measure increased in the pre-sale market in the post-policy period. Moreover, among the pre-sale projects, these measures rose more where pre-intervention speculative trading was more active.
Our results suggest that the elevated transaction cost drives out informed speculators more than uninformed speculators. However, the results could still be subject to other interpretations. It would thus be desirable to study a setting in which the relative prevalence of informed versus noise speculators can be instrumented for ex ante.Changes in volatility for over- and underpriced projects
We therefore further sort the pre-sale projects according to their prices shortly before the policy intervention relative to their long-term price trends. Operationally, we sort pre-sale projects – using two-month-prior prices adjusted for hedonic factors – into the top 30%, labeled ‘overpriced’, and into the bottom 30%, labeled ‘underpriced’. Due to their inability to short-sell real estate, informed speculators would presumably be more likely to pick underpriced projects and avoid overpriced ones, whereas the uninformed noise traders’ transaction picks would be more randomly distributed.
Panel A of Table 1 below shows a sharper fall in trading activity in the previously underpriced projects than in the previously overpriced projects after the policy intervention of December 2006, while at the same time trading activity in the (unaffected) spot market remained unchanged. Panel B shows that after the policy intervention – although there was an overall upward trend in volatility as reflected in the (unaffected) spot market – the previously underpriced projects registered a much stronger increase in volatility. These observations strengthen the case that the heightened volatility reflects a disproportionate retreat of informed speculators relative to noise speculators.
In the 12-month post-policy intervention period, the previously overpriced projects experienced a small increase in volatility, but much less than that experienced in the previously underpriced pre-sale market or the spot market. This observation is consistent with some noise speculators being dissuaded from the market. These results are robust – we obtained the same results when we used more conventional volatility measures (i.e. the standard deviation of monthly returns).Conclusions
Overall, these observations are consistent with the idea that raising transaction costs drives out informed speculators more than noise speculators from the market. The disproportionate withdrawal of the former leads to a rise in volatility. Our research findings thus highlight the possibility that financial transaction taxes can have an adverse effect by damaging the quality of asset prices.
Table 1. Differences-in-differences results
Amihud, Y (2002), “Illiquidity and stock returns: cross-section and time series effects”, Journal of Financial Markets 5: 31–56.
Amihud, Yakov and Haim Mendelson (2003), “Effects of a New York State Stock Transaction Tax”, unpublished.
Fu, Yuming, Wenlan Qian, and Bernard Yeung (2013), “Speculative Investors and Tobin’s Tax in the Housing Market”, NBER Working Paper 19400.
Hau, H (2006), “The role of transaction costs for financial volatility: Evidence from the Paris bourse”, Journal of The European Economic Association 4(4): 862–890.
Jones, Charles M and Seguin, Paul J (1997), “Transaction Costs and Price Volatility: Evidence from Commission Deregulation”, The American Economic Review 87(4): 728–737.
Pomeranets, Anna and Daniel G Weaver (2011), “Security Transaction Taxes and Market Quality”, Bank of Canada Working Paper 2011-26.
Stiglitz, J E (1989), “Using tax policy to curb speculative short-term trading”, Journal of Financial Services Research 3: 101–115.
Subrahmanyam, A (1998), “Transaction taxes and financial market equilibrium”, Journal of Business 71: 81–118.
Summers, L H and V P Summers (1989), “When financial markets work too well: a cautious case for a securities transaction tax”, Journal of Financial Services Research 3: 261–286.
Tobin, James (1978), “A proposal for international monetary reform”, Eastern Economic Journal 4: 153–59.
Umlauf, Steven R (1993), “Transaction Taxes and the Behavior of the Swedish Stock Market”, Journal of Financial Economics 33(2): 227–40.