Proponents of securities transaction taxes (STTs) argue that they can help to
reduce asset market volatility by discouraging short-term speculation by
poorly informed and destabilising ‘noise traders’. New research by Professors
Frank Song and Junxi Zhang, published in the October 2005 issue of the
Economic Journal, shows that the impact of an STT depends on which kinds
of traders are discouraged and the extent to which trading volume is reduced.
If rational, stabilising traders are discouraged and/or the market becomes too
‘thin’, an STT may in fact lead to increased market volatility.
The fact that the effect of an STT on volatility is contingent on market conditions
suggests that in countries with higher volatility and greater participation of noise
traders – as in many emerging markets – an STT may help to reduce volatility.
But in markets with relatively lower volatility and fewer noise traders, an STT
may have the opposite effect, contrary to what regulators will have intended.
The idea of a tax levied on financial transactions can be traced back to Keynes,
who said that a transaction tax could strengthen the weight of long-range
fundamentals in stock market pricing as against those of short-term
speculations. Similarly, James Tobin proposed taxing foreign exchange
transactions to ‘throw sand in the wheels’ of super-efficient financial vehicles.
Later scholars have questioned the view that an STT reduces market volatility
by discouraging the trading activity of destabilising short-term traders:
• • First, while noise traders may increase market volatility, informed
traders may more than offset their effects: an indiscriminate STT on
both informed and noise traders would have ambiguous effects on
volatility.
• • Second, an STT may increase volatility by reducing market liquidity
and increasing the price impact of trades.
• • Third, an STT may reduce asset values, decreasing market
efficiency and driving trading to other countries.
• • Finally, there are significant implementation problems across
different types of securities markets.
Few studies have empirically documented the impact of an STT on market
volatility. Some have found that transaction taxes are inversely but
insignificantly correlated with volatility across countries. Others have found
that volatility does not decline in response to the introduction of an STT
although stock price levels and turnover do. In the context of Asian markets,
researchers have found that an STT reduces stock prices but has no
significant effect on market volatility and market turnover. Overall, there
seems to be no evidence that an STT can reduce noise trading and volatility.
Song and Zhang’s paper examines the effects of an STT on the set of noise
traders and market volatility. They argue that an STT may not only discourage
the trading activity of destabilising traders but also discourage rational and
stabilising traders. The net effect of an STT on volatility will depend on the
change of composition of traders arising from the tax – the trader composition
effect.
Moreover, an STT may reduce significantly trading volume and hence the
depth of the market. The resulting thin market may exhibit greater volatility.
The potential effect of an STT on market volatility through its impact on
liquidity is labelled the liquidity effect. The net impact of an STT on financial
market volatility will depend on the relative magnitude and interaction of the
two effects.
In the framework of the study, both rational, well-informed investors and
erratic, poorly informed traders participate in securities markets. Since the
latter type of trader adds a considerable amount of noise to security prices, an
STT that restrains their trading may help reduce the market volatility.
But by participating in the market, the noise traders also contribute to its
depth. Because of its high liquidity, a thick market tends to be associated with
relatively low volatility. Hence it is entirely possible for an STT, which
discourages noise traders, to result in thin markets, leading to less liquidity
and greater volatility.
The findings in this paper have important implications. First, the result that the
effect of STT on market volatility is contingent on market conditions sheds some
light on why empirical researchers have failed to document any significant
impact of STTs on volatility.
Second, by formally articulating the potential channels through which an STT
affects market volatility, the study helps regulators to formalise important
regulations that may potentially affect securities markets.
Finally, the framework in this paper can be easily incorporated into studies of
other important securities regulations, such as margin requirements and price
limits.
ENDS
Notes for editors: ‘Securities Transaction Tax and Market Volatility’ by Frank
Song and Junxi Zhang is published in the October 2005 issue of the
Economic Journal.
The authors are in the School of Economics and Finance at the University of
Hong Kong.
For further information: contact Frank Song on +852-285-8507 (email:
fmsong@econ.hku.hk); or RES Media Consultant Romesh Vaitilingam on
0117-983-9770 or 07768-661095 (email: romesh@compuserve.com).