The impact of the 2008 - 2009 short selling
ban in the United Kingdom on stock prices
and trading volume
Abstract: From September 19, 2008 until January 16, 2009 the Financial Services Authority imposed a ban on short selling on 35 financial companies in the United Kingdom. This thesis analyses the effects of this ban in the United Kingdom on stock prices and trading volume and concludes that the ban had a significantly positive effect of 6.46% on stock prices of financial firms compared to non-financial firms. With first differences an insignificantly positive effect 0.0007 was found. The ban had a significantly negative effect of -44.92% on trading volume for financial firms compared to non-financial firms. With first differences an insignificantly negative effect of -0.0051 was found.
Name: Nick van Delft
University: University of Amsterdam
Faculty: Faculty of Economics and Business
Date: 30 June 2014
Supervisor: M. A. Dijkstra
Bachelor: Finance and Organization
I: Introduction
Short selling is borrowing a stock from a stockowner for a premium and selling the stock at the current market price. The borrower intends to buy the stock back later on the market for a lower price. His profit is the decline in stock price minus the paid premium. Short selling was originally designed to hedge against price declines.
When on September 15, 2008 investment bank Lehman Brothers filed for a chapter 11 bankruptcy stock markets dropped. The FTSE100, the 100 largest companies in the United Kingdom based on market capitalisation, closed on September 15, 2008 -3.9% lower. Because of the size of Lehman Brothers and its involvement with other financial companies, the Financial Services Authority imposed a ban on short selling on financial institutions to stop the decline in stock prices in financial companies. The ban was announced on Thursday September 18, 2008, and will be effective of 00:01 the following day. The ban prohibited the creation or increase of all net short positions in public financial companies in the United Kingdom and all United Kingdoms financial companies traded outside the United Kingdom. The ban included naked and covered short selling, derivatives trading and contracts for differences. Short selling in bonds was still possible. The Financial Services Authority announced that the ban would remain effective until January 16, 2009.
The research question in this thesis is: What are the effects of the short selling ban in 2008 on stock prices and trading volume on financials in the United Kingdom? Daily closing stock prices and daily trading volume before, during and after the ban is used to analyse the effects of the ban. For the FTSE350 financial companies and non-financial companies are compared during 2008 and 2009 by using panel regressions and first difference regressions. A difference-in-difference method is used to see what the effect of the short selling ban in on stock price and trading volume.
Chang, Cheng, and Yu (2007) have researched the Hong Kong stock market and found that stock prices volatility is higher, and returns are more negative when short selling can be done. Boulton and Braga-Alves (2010) looked abnormal returns at the United States stock market in 2008. They found a positive abnormal return during the short selling ban. Because the United Kingdom stock market is one of the largest stock markets in Europe it is interesting to see if the effects of the short selling ban are in line with previous research. Since regulators use short selling bans in times of economic crises, this research is useful to
see if short selling bans have the desired effects. Regulators can learn from this research if they intent to ban short selling in future crises.
Chapter II reviews the current literature on the effects of a short selling ban. In chapter III the hypothesis is described, followed by the methodology and data in chapter IV. Chapter V presents the results. Chapter VI concludes and finally in chapter VII a discussion is given.
II: Literature review
The reason for the Financial Services Authority to impose the short selling ban was to protect the quality of the markets and to protect against further instability in the financial sector. Hector Sants, the chief executive officer of the Financial Services Authority stated that the current circumstances have disrupted the stability of the markets (Financial Services Authority, 2008). The belief of the Financial Services Authority is that short sellers increase the downward pressure on falling stocks, causing instability in financial markets. The Financial Services Authority wants to prevent this instability. This instability exposes financial firms to a higher level of risk. The goal of the ban is to stop the downfall in stock prices of financial institutions.
When researching the effect of a short selling ban on stock prices it is important to consider that there are two types of short selling; naked short selling and covered short selling. Gruenewald, Wagner and Weber (2009) explained that covered short selling is that the seller borrows a stock from a lender for a small premium and which the seller later on a specific date repurchases back from the market and gives it back to the lender. The short seller then sells the borrowed stock on the market so the receives the current market price. A short seller expects a stock price decline, so his profit will be the difference between the current market price and the future market price minus the borrowing premium.
Naked short selling is were the short seller sells the stock to a buyer without owning or lending the stock prior to the sale. The risk of naked short selling is that the seller cannot deliver. This means that the selling is unable to obtain the stock from a lender before the expiration date. An implication of naked short selling is that more then 100% of a company’s stock can be sold. Stocks that don’t exist can be sold. Gruenewald, Wagner and Weber (2009) stated that naked short selling is more aggressive than covered short selling.
Miller (1977) states that short selling increases the supply on the stock market by the total amount of the short position. And when a short selling ban is present, this supply will be lower. Miller shows that when divergence of opinions is present, only optimistic investors, who think the stock price will go up, will trade. While pessimistic investors, who think the stock price will go down, can’t do anything because they aren’t allowed to sell short. This results in that only optimistic beliefs are incorporated in the stock prices and that the stock prices would be overvalued.
Harrison and Kreps (1978) made a model and show that in a situation in which rational investors with differences of opinions and where short selling bans are present, a speculative premium occurs. Because of this speculative premium stock prices are higher, even higher then when stock prices only reflect the optimistic investors views. The theory of Harrison and Kreps (1978) goes beyond the theory of Miller (1977) who stated that stock prices only reflect the optimistic investors view when a short selling ban is present. But in both cases short selling constraints lead to overpricing.
Diamond and Verrecchia (1987) give a model that estimates the effect of constraints on the speed of adjustment of stock prices to private information. They stated that when many people go short on a stock, the return on the stock would go down. Since short selling is costly, it isn’t likely to be done by liquidity traders. They say that short sellers are informed traders. Short selling restrictions then reduce the speed at which prices adjust to private information, especially to negative news. But that in a presence of a derivatives market, informed investors can still go short with options and other derivatives. Because of this no overvaluation is presence.
In the article of Hong and Stein (2003) Differences of opinion, short-sales constraints
and market crashes, they propose a theory and create a heterogeneous agent model that looks
at market crashes and short selling bans. The theory of Hong and Stein (2003) is that when there are a lot of differences of opinions and the presence of a short selling ban, bearish investors with negative views can’t sell a stock short. Because of the short selling ban, the negative information from these investors is not revealed. Hong and Stein (2003) theorizes that this accumulated hidden information comes out when markets begin to decline, which further pressures the markets decline. This could lead to a market crash. Hong and Stein (2003) state that trading volume is a proxy for differences in opinions and that after periods of high trading volume, negative skewness in returns will occur.
Empirical evidence on short selling largely supports the theoretical view Miller (1977) that short selling restrictions lead to overpricing
Boulton and Braga-Alves (2010) researched a short selling ban in the United States in 2008. They looked at the abnormal return in stock prices for banned stocks. The cumulative abnormal returns were calculated using the Fama and French’s (1993) three-factor model. They found a positive abnormal return of 12.9% between the announcement date and the implementation date. However 4 days after the expiration date the abnormal returns becomes negative of -10.9%. This ends up with a mean abnormal return of only 0.7% from the
that short selling constraints lead to overpricing.
Chang, Cheng, and Yu (2007) use an event study to analyse the short selling restriction in the Hong Kong stock market. They analyse the Hong Kong stock market because on the Hong Kong stock market it is only allowed to sell short if the stock is on an official short selling list. They look at the stock returns before and after a stock is added to that list. They find that the volatility increases when short selling is possible. They also find that when short selling is possible the stock returns are less positive.
Figlewski (1981) tests Millers (1977) theory and finds results that support the theory that short selling constraints can lead to overvaluation. Since short selling constraints have different impact to bullish or bearish investors. Investors with positive information can buy a stock, but investors with negative information cannot short the stock, there is more adverse information between investors. This can disturb the price discovery and lead to overvaluation.
Boehmer, Jones and Zhang (2011) look at the U.S. security market. They examine the effect of the ban on market quality, shorting activity and stock prices for 1000 U.S. financial stocks. They made matched pairs of banned stocks and stocks that have never been banned from January 2008 to July 2008. They used firm-pair fixed effect regressions. They find that market quality degrades. They measured this by effective spreads, price impact and realized spreads. They also show a decrease of shorting activities by 77%. Another point was that for banned stocks the stock price increases, but this is not due to the short selling ban, but due to the bailout program.
Bris, Goetzmann and Zhu (2007) look at 46 equity markets from the period 1990 to 2001 and look if short selling restrictions have effect on stock price returns. Their result is a significantly positive correlation between the skewness of market returns and short selling restrictions. They didn’t find this correlation for individual stocks. Another result is that there is no significant association between short selling restrictions and the frequency of negative returns. Bris, Goetzmann and Zhu (2007) also did an event study on five countries from 1990 to 2001 that changed their laws concerning short selling and found that short selling makes market crashes more likely. They show that short selling bans are lifted, the distribution of stocks is more negatively skewed and that more negative returns occur. They note that because only five countries are examined the sample size is small.
Lamont (2004) look the battles between short sellers and firms. He looks at 266 firms in the United States stock market from 1977 to 2002. Because firms use methods to reduce short selling, it creates short selling constraints. The result of Lamont (2004) is that firms that constrain short selling have lower returns the subsequent year. He shows that firms have on average a negative abnormal return of 2% per month.
Danielsen and Sorescu (2001) test the theory of Diamond and Verrecchia (1987) who say that short selling restrictions have an impact on the speed of how stock prices are adjusted by positive and negative information. Danielson and Sorescu (2001) use a large sample of United States stocks that also have listed options before and prior to 1981. They used standard deviation analysis and used proxies for dispersion and for beliefs. They stated that the introduction of options relax short selling bans. Their result is a negative abnormal return of 2% when options are introduced. This indicates that negative information is embedded slowly into stock prices when a short selling ban is present.
Saffi and Sigurdsson (2011) conduct a study by using lending supply and loan fees as proxies for short selling bans. The study is based on a global dataset of 12,600 stocks from 26 different countries. The time period of this study is from 2005 to 2008. They state that short selling bans have no effect on price stability or negative returns. The result Saffi and
Sigurdsson (2011) is a positive effect of short selling bans on the skewness and kurtosis of stock returns. However they say that it is caused by the frequency of extreme positive returns. This is because due to the short selling restrictions arbitrageurs can’t correct for
overvaluation. Saffi and Sigurdsson (2011) also looked at the volatility of the stocks returns affected by short selling bans. They used the standard deviation of weekly returns as measure for volatility. Their result was that stocks affected by short selling bans have higher volatility.
III: Hypothesis
The main article on which this thesis is based is Miller (1977). Miller states that when that a ban on short selling removes negative investors from the market. These negative investors are no longer allowed to sell the security short. Since this negative information is not incorporated in the stock prices anymore, so stock prices only reflect optimistic beliefs, stock prices will be overvaulted.
Hypothesis 1: Stock prices of financial companies increase when the short selling ban is imposed.
Miller (1977) also stated that the supply of stocks is higher when no restrictions are present. Since short selling in financial companies is no longer allowed, there is less supply and as result a lower trading volume. The theory is that the short selling ban in the United Kingdom will result in a lower trading volume.
Hypothesis 2: financial stocks have lower trading volume when the short selling ban is imposed.
IV: Methodology and Data
IV.I. Methodology
The main model used to measure the effects will be:
Yit = b0+b1Finit+b2Banit+b3Finit*Banit+a1LRit+a2Assetsit+a3Empit+a4ROIit+eit
Were Yit is the dependent variable for either stock price or trading volume measured as the total traded volume. The variables b1, b2 and b3 are the main variables. The coefficient Fin is a dummy that equals 1 if the company is a financial firm and 0 if the company isn’t a financial firm. Ban is a dummy that equals 1 if the ban is active and 0 if the ban isn’t active.
The coefficient Ban*Fin is the interaction of these two. It captures the interaction between Fin and Ban. The coefficient Ban*Fin is the difference-in-difference estimator and it explains the effect of the ban on financial firms. The variables LR, Assets, Emp and ROI will be the relevant control variables. As control variables leverage ratio measured as the total debt/total capital (LR) was used. The total debt is measured as the long-term debt + the short-term debt. The total capital is measured as common equity. Total assets (Assets) are measured in pounds. Employees (Emp) are the total number of employees including fulltime and part-time employees. The return on investment (RoI) is measured in percentages as the (net profit/investment)*100.
The hypothesis consists of 2 sub-hypothesis. The first one was the stock price.
The coefficient b1 captures the effect that effect of any potential difference between financials and non-financials. B2 captures the effect that can cause changes in the stock price even if a ban is not active. The coefficient b3 is the interaction coefficient. This one only applies if Ban and Fin are both 1. This coefficient is the difference-in-difference estimator and shows what the effect of the short selling ban on financials is on the stock price. The expectation is that the short selling ban has a positive effect on stock price. If this is the case then b3 should to be larger than 0.
The second part of the hypothesis was to look at the trading volume. b1 now captures the effect between financials and non-financials on the trading volume, b2 captures the effect of changes in trading volume if a ban is active or not. And b3 now captures the difference-in-difference of the ban on financials on trading volume. The expectation is that the trading volume would go down during a ban period.
The hypothesis will be tested using ordinary least squares regressions and first difference regressions.
IV.II. Data
The sample contains all the firms of the FTSE350 (the largest 350 companies in the United Kingdom based on market capitalization) from January 1 2008 to December 31 2009. There are a total of 359 firms in the sample including the stocks that were affected by the ban. In the sample 25 stocks were affected by the short sell ban. Table 1 shows the affected firms and if the firm is traded on the FTSE350 or not. A total of 9 firms were too small to be traded on the FTSE350 and they are traded on the FTSE Small Cap or the Alternative Investment Markets.
Compustat Global is used to download the FTSE350 constituent list. The sample includes a total of 523 trading days including the 86 trading days when the ban was active. The ban period start at September the18th, not the 19th to eliminate any announcement effects. In the period 2008-2009 some firms went bankrupt. These firms were included in the sample to eliminate any survivorship bias. Stock prices, trading volume, leverage ratio, assets, employees and return on investment are based on a daily basis. Daily stock prices, trading volume, leverage ratio, assets, employees and return on investment are downloaded from Thompson’s DataStream.
This thesis looks at the effect of the short sell ban in the United Kingdom using panel data. The difference-in-difference estimation model used in this thesis is based on the model of Roberts and Whited (2011). The effect of the short sell ban is measured by comparing the pre-ban period and post-ban period with the ban period.
Since the data consisted of multiple panels with gaps, a unit root test couldn’t be done. This means that it couldn’t be tested whether the time series is non-stationary. It can be assumed that price and volume consist a unit root. To overcome this problem, regressions using first differences are also dome. First difference regressions look at the changes from one period to the next period. If Y(t) denotes the value of the time series Y at period t, that the first difference of Y at period t is equal to Y(t)-Y(t-1).
Descriptive statistics:
Figure 1 shows the daily average stock prices of the financials and the non-financials in 2008 and 2009. This figure shows a decline in stock prices for both the financials and the non-financials during the ban period and shows that stock price drop for financial firms was a little bit steeper that for non-financials. The figure also shows that the stock prices become less volatile during 2009, especially when the ban is lifted. The up-and-down swings of financials become much less than before the ban and during the ban.
Table 2a, b, c, d and e show descriptive statistics for the used variable in the ban period and the non-ban period for both financials and non-financials. What figure 1 already suggested is sustained by the descriptive statistics. By looking at the changes in prices a drop of 17.94% for financials during the ban period and a drop in prices of 18.79% for non-financials during the ban period.
Figure 2 shows the daily average trading volume for the period 2008 to 2009 for financial firms and non-financial firms. The trading volume in financials is more then three times higher than the trading volume in non-financials. For non-financials the trading volume doesn’t change much during the ban period compared before and after the ban, but the trading volume for financials declines during the ban. After the ban is lifted the trading volume in financials rises again. This suggests that the short selling ban has a great impact on the trading volume of financial firms.
Table 2a, b, c, d and e show the descriptive statistics for trading volume. The table shows that the short selling ban has an effect on the trading volume. The average trading volume for financials firms drops with 23.62% compared to an increase of 18.22% in trading volume for non-financial firms. This result is in line with the drop in trading volume showed in figure 2.
V: Results
V.I. Price
Regression analysis
Table 3 presents the results of the panel regression on the natural logarithm of stock prices. The difference between the models (1), (2), (3), (4), (5) and (6) is that a control variable is added each time. By looking at model (6) it is shown that a short selling ban has a significantly negative impact of -27.89% on stock prices whether it are financial firms or non-financial firms. This is in line with the expectations from figure 1 were a stock price decline for both financial a non-financial firms was seen.
The coefficient Fin is highly significant and has a negative impact on stock prices. The stock prices of financial firms are worth 139.45% less then non-financial firms. This effect isn't in line with figure 1.
The interaction coefficient ban*fin captures the effect of the short selling ban on financial firms. The estimated value is 0.0643 and is highly significant. This means that the stock prices of financial firms are valued 6.46% higher then non-financial firms during a ban. This is in line what Miller (1977), who stated that stock prices of banned firms would be overvalued in times of a short selling ban. The short selling ban in the United Kingdom had a positive effect on the financial companies stock prices compared to non-financials. The overall effect of a short selling ban on stock prices is negative, also for financial firms.
Since the data consisted of multiple panels with gaps, a unit root test couldn’t be done. This means that it couldn’t be tested whether the time series is non-stationary. It can be assumed that price and volume consist a unit root. To overcome this problem, regressions using first differences are also done. Table 4 presents the estimations results using first differences.
Only the variable ban is highly significant in this case. The result is that a ban has a negative impact on stock prices whether it are financial firms or non-financial firms. This is in line with the panel regression were similar results were found. Both times the outcome was that a ban has a negative impact on stock prices.
In the case of first differences the variable Fin isn’t significant anymore. This implies that financials are not valued more then non-financials. This in contrast of the panel regression, were the coefficient Fin was highly significant.
The interaction coefficient Ban*Fin is also insignificant in the case of first differences. This implies that a ban on financial firms has no effect on the stock prices of the financial firms. This result would reject Millers (1977) theory that stock prices would be overvalued during a short selling ban, and would also contradict the findings in the panel regression, were a highly significant effect was found.
V.II. Volume
Regression analysis
Table 5 shows the results using panel regressions. The estimation for the coefficient ban is 0.0579. This implies that the trading volume rises significantly with 5.79% when a ban is active. This coefficient only shows the effect of a ban on trading volume, not the effect of
financials during a ban. This isn’t in line with the expectations that during a short selling ban the trading volume would drop. This might have to due because the coefficient captures both financials and non-financials, and since non-financials weren’t affected by the short selling ban they could be still traded short.
Table 5 also shows the difference between financial firms and non-financial firms. The trading volume for financial firms is significantly larger than the trading volume for non-financial firms. Financial firms are traded on average 62.34% more then non-non-financial firms. This is in line with figures 2 were it could be seen that the trading volume for financials is much higher than for non-financials. It is also in line with the literature that financial firms are traded more actively than non-financial firms.
The interaction coefficient Ban*Fin captures the effect of the short selling ban in the United Kingdom on the financial firms. The result is that during a short selling ban on financials, the trading volumes in financial firms drop significantly with 44.92%. Again this is in line with figures 2 were this drop during a short selling ban was seen. Moreover this result is also in line with the literature, were it was stated that a short selling ban would have a negative effect on the trading volume of the affected firms.
Table 6 shows the results using first differences. The results are much different than the results found in table 5. Were in table 5 all the coefficients were significant, now only the coefficient Ban is significant. The estimated value in model (6) is -0.0269. This is a remarkable result since the coefficient is now negative. In the previous model the effect was positive. This result implies that a ban has a negative effect on trading volume on both financials and non-financials. This result is in line with the literature that stated that trading volume would drop when a short selling ban is active.
The coefficient Fin isn’t significant. This result implies that there is no difference between the trading volume of financial firms and non-financial firms. This contradicts the findings in table 5 were a significant difference between financial firms and non-financial firms was found.
The interaction coefficient Ban*Fin isn’t significant in the case of first differences. The estimated value in model (6) is -0.0051. This would imply that the trading volume of financial firms don’t differ during a short selling ban. But since the estimated coefficient is still negative, the effect of a short selling ban on financials on trading volume is negative. Only it can’t be said that this effect is significant. This result is in line with table 5 were a negative effect on trading volume was found, and the literature that stated that the effect of a short selling ban on trading volume was negative.
VI: Conclusion
This thesis researched the effect on stock prices and trading volume of the short selling ban on the United Kingdom on financial companies. The ban was imposed on September 18, 2008 and was lifted on January 16, 2009. Both naked and covered short selling was banned. In total 34 United Kingdom financials were affected by the short selling ban. This research looked at the FTSE350, the 350 largest companies by market capitalization, of which 25 financial companies were affected by the short selling ban. By using dummy variables for whether the company is a financial company or not and whether the ban is active or not. A difference-in-difference analysis is used to obtain the results.
The short selling ban on financial firms had a significantly positive effect on stock prices. The stock prices were valued 6.42% higher then non-financials during the short selling ban. To control for unit roots in the data set first differences are also used, which resulted a non-significant positive effect on stock prices. Because this result is insignificant it cannot be concluded that stock prices for financial firms increase when a short selling ban in imposed.
Further the short selling ban had a negative effect on the trading volume for financials. The total trading volume was positive, but lower that without a short selling ban. During the ban, the trading volume for financial firms dropped significantly with 44.92% compared with non-financials. Also for the case of trading volume first differences were taken. The result was not significant, but a small negative effect on trading volume for financials compared with non-financials was found.
VII: Discussion
Future research could try to look at volatility and liquidity. Bid-ask spreads can be used as measurement for liquidity. Some form of regulation with respect to short selling should exist to prevent the markets for financial crises. In cases of financial crises, where there is a downward pressure on stock prices, and especially financials stocks, governments should be able to put restrictions on short selling to ease the downward pressure. Because financial companies are important for a civilization, since these companies provides the people with money and other financial services. When stock prices of financial firms are dropping, these companies can get in trouble and this could start panic among the people. Short selling should not be banned permanently, but in times of financial crisis’s short selling should be banned on a temporary basis like the short selling ban in the United Kingdom to prevent for market wide panics.
References
Bai, Y., E. C. Chang and J. Wang. 2006. Asset prices under short-sale constraints. Working paper, University of Hong Kong and MIT.
Boehmer, E., C.M. Jones and X. Zhang (2009) “Shackling Short Sellers: The 2008 Shorting Ban”, working paper, Available at SSRN: http://ssrn.com/abstract=1412844.
Boulton, T. J. and M. V. Braga-Alves (2010). "The skinny on the 2008 naked short-sale restrictions." Journal of Financial Markets 13(4): 397-421.
Bris, A., W. N. Goetzmann and N. Zhu. 2007. Efficiency and the bear: Short sales and markets around the world. Journal of Finance 62 (3): 1029–1079.
Chang, E. C., J. W. Cheng and Y. Yu Cheng. 2007. Short-sales constraints and price discovery: Evidence from the Hong Kong market. Journal of Finance 62 (5): 2097–2121. Danielsen, Bartely, and Sorin M. Sorescu, 2001. “Why do Option Introductions Depress Stock Price? A Study of Diminishing Short Sale Constraints.” Journal of Financial and Quantitative Analysis, Vol. 36, No. 4, 451-484.
Diamond, D. W. and R. E. Verrecchia. 1987. Constraints on short selling and asset price adjustment to private information. Journal of Financial Economics 18: 277–312.
Figlewski, Stephen, 1981. “The Informational Effects of Restrictions on Short Sales: Some Empirical Evidences.” Journal of Financial and Quantitative Analysis, Vol. 16, No. 4, 463- 476.
Financial Services Authority, 2008. “Short Selling.” Available at:
http://www.fca.org.uk/firms/markets/international-markets/eu/short-selling-regulations/notifications-disclosures
Gruenewald, S. N., A. F. Wagner and R. Weber. 2009. Short selling regulation after the financial crisis: First principles revisited. International Journal of Disclosure and Regulation 7 (2): 108–135.
Harrison, M. J., & Kreps, D.M. (1978). Speculative investor behavior in a stock market with heterogeneous expectations, Quarterly Journal of Economics.
Hong, Harrison, and Jeremy C. Stein, 2003, Differences of opinion, short-sales constraints, and market crashes, Review of Financial Studies 16, 487-525.
Lamont, O. (2004). Go down fighting: Short sellers vs. firms, National Bureau of Economic Research.
Miller, E. M. 1977. Risk, uncertainty, and divergence of opinion. Journal of Finance Roberts, Michael R. and Toni M. Whited, 2011, Endogeneity in Empirical Corporate FinanceHandbook of the Economics of Finance, vol. 2, Elsevier.
Saffi, Pedro, and Kari Sigurdsson, 2010, Price efficiency and short selling, Review of Financial Studies.
Appendix
Table 1: List of banned stock
This table shows the list of all the stocks that were affected by the short selling ban from September 18, 2008 to January 16, 2009.
Table 1: list of firms that were affected by the 2008-2009 short selling ban in the United Kingdom
Firms in FTSE350 Firms not in the FTSE350
1 ABERDEEN ASSET MANAGEMENT PLC 26 ARBUTHNOT BANKING GROUP PLC
2 ADMIRAL GROUP PLC 27 CHESNARA PLC
3 ALLIANCE & LEICESTER PLC 28 EUROPEAN ISLAMIC INVESTMENT BANK PL
4 ALLIANCE TRUST PLC 29 HIGHWAY INSURANCE GROUP PLC
5 AVIVA PLC 30 ISLAMIC BANK OF BRITAIN PLC
6 BARCLAYS PLC8 31 JUST RETIREMENT HOLDINGS PLC
7 BRIT INSURANCE HOLDINGS PLC 32 LONDON SCOTTISH BANK PLC
8 CLOSE BROTHERS GROUP PLC 33 NOVAE GROUP PLC
9 F&C ASSET MANAGEMENT PLC 34 TAWA PLC
10 FRIENDS PROVIDENT PLC
11 HBOS PLC
12 HSBC HOLDINGS PLC
13 INVESTEC PLC
14 LEGAL & GENERAL GROUP PLC
15 LLOYDS TSB GROUP PLC
16 OLD MUTUAL PLC
17 PROVIDENT FINANCIAL LTD
18 PRUDENTIAL PLC
19 RATHBONE BROTHERS PLC
20 ROYAL BANK OF SCOTLAND GROUP PLC
21 RSA INSURANCE GROUP PLC
22 SCHRODERS PLC
23 ST JAMES'S PLACE PLC
24 STANDARD CHARTERED PLC
Figure 1: Average stock price
This figure shows the average stock prices for financials and non-financials in from January 1, 2008 to December 31, 2009. Firm’s stock prices are in pounds.
Figure 2: Average trading volume
This figure shows the average trading volume for financials and non-financials from January 1, 2008 to December 31, 2009.
Table 2a, b, c, d and e: Descriptive statistics
These tables show the descriptive statistics for the financials and the non-financials in the ban period and in the non-ban period. The ban period is from September 18, 2008 to January 16, 2008. Mean, median, standard deviation, minimum and maximum are given for price, trading volume, natural logarithm of assets, natural logarithm of employees, RoI and the debt/capital ratio.
Table 2a Financials
Ban period
Mean Median Stan. Dev. Min Max
Price 410,211 287,000 325,823 31,000 2205,000 Volume 14692,050 3481,900 34107,860 1,200 711217,800 lnAssets 17,547 17,426 2,512 13,645 21,596 lnEmployees 9,211 8,916 1,949 5,710 12,654 RoI 2,564 5,050 21,000 -45,050 56,480 Debt/capital 47,958 50,390 32,250 0,050 95,910 Table 2b Financials Non-ban period
Mean Median Stan. Dev. Min Max
Price 499,866 362,375 452,201 19,990 3717,970 Volume 19236,000 4217,300 42125,550 2,800 669517,800 lnAssets 17,512 17,426 2,518 13,645 21,596 lnEmployees 9,188 8,947 1,973 5,663 12,654 RoI 7,544 7,610 18,132 -45,050 56,480 Debt/capital 45,703 45,660 31,267 0,050 95,910 Table 2c Non-financials Ban period
Mean Median Stan. Dev. Min Max
Price 419,033 273,000 447,134 1,500 3502,680 Volume 4912,257 1387,900 16010,800 0,200 496005,900 lnAssets 14,396 14,110 1,444 10,793 19,118 lnEmployees 8,601 8,888 1,959 1,609 13,296 RoI 16,393 9,350 107,314 -210,820 1376,470 Debt/capital 39,789 36,920 32,780 -64,770 323,170
Table 2d Non-financials
Non-ban period
Mean Median Stan. Dev. Min Max
Price 515,969 340,220 551,092 1,480 5847,170 Volume 4155,106 1335,400 12283,370 0,000 914842,300 lnAssets 14,393 14,110 1,426 10,793 19,118 lnEmployees 8,608 8,848 1,962 1,609 13,296 RoI 29,265 12,030 279,505 -229,390 7206,450 Debt/capital 37,021 34,480 39,371 -1070,000 323,170
Table 2e Percentage change between ban period and non-ban period
Financials Non-Financials Price -17,94% -18,79% Volume -23,62% 18,22% lnAssets 0,20% 0,02% lnEmployees 0,25% -0,08% RoI -66,01% -43,99% Debt/capital 4,93% 7,48%
Table 3: Panel regression of stock price
This table reports the results of the panel regressions for the stocks in the sample. The dependent variable is the natural logarithm of stock price. The independent variables are the natural logarithm of assets, natural logarithm of employees, the return on investment, the debt/capital ratio and the dummy variables Ban, Fin and Ban*Fin.
Table 3: Panel regression of stock price
lnPrice (1) (2) (3) (4) (5) (6) Ban -0.2388 *** -0.2428 *** -0.2616 *** -0.2708 *** -0.2692 *** -0.2789 *** (-106.37) (-104.24) (-114.7) (-109.93) (-109.73) (-113.26) Fin -0.2388 0.0354 -1.4549 *** -1.8042 *** -1.4681 *** -1.3945 *** (0.22) (0.17 (-7.3) (-8.84) (-7.26) (-6.95) banFin 0.0557 *** 0.0527 *** 0.0646 *** 0.0616 *** 0.0643 *** (6.36) (6.19) (7.36) (7.17) (7.51) lnassets 0.4848 *** 0.6056 *** 0.4931 *** 0.4584 *** (67.26) (72.46) (59.03) (54.65) lnemployees 0.0186 *** 0.0181 *** 0.0008 (5.84) (5.81) (0.24) RoI -0.0001 *** -0.0001 *** (-13.75) (-13.88) Debt/Capital 0.0051 *** (32.78) Constant 5.7744 *** 5.7751 *** -1.2153 *** -3.1626 *** -1.5165 *** -1.0592 *** (106.73) (106.74) (-10.41) (-23.9) (-11.44) (-7.98) Number of obs 184619 184619 170305 155008 149467 149467 R2 0.0579 0.0581 0.0999 0.1059 0.1001 0.1065
Table 4: First difference regression of stock price
This table reports the results of the first difference regressions for the stocks in the sample. The dependent variable is the natural logarithm of stock price. The independent variables are the natural logarithm of assets, natural logarithm of employees, the return on investment, the debt/capital ratio and the dummy variables Ban, Fin and Ban*Fin.
Table 4: Panel regression of first differences in stock prices
lnPriceDiff (1) (2) (3) (4) (5) (6) Ban -0.0047 *** -0.0047 *** -0.0048 *** -0.0048 *** -0.0048 *** -0.0049 *** (-10.48) (-10.12) (-10.54) (-10.41) (-10.55) (-10.64) Fin -0.0002 -0.0003 -0.0007 -0.0007 -0.0008 -0.0008 (-0.37) (-0.36) (-0.95) (-0.99) (-1.15) (-1.19 ) banFin 0.0001 0.0005 0.0006 0.0007 0.0007 (0.06) (0.27) (0.35) 0.41) ( 0.43) difflnassets 0.1291 *** 0.1407 *** 0.1190 *** 0.1235 (80.39) 74.02) (60.08) (60.96) difflnemployees -0.0237 *** -0.0211 *** -0.0181 *** (-14.63) (-12.74) (-10.79) diffRoI -0.0001 *** -0.0001 *** (-10.94) (-10.19) diffDebt/Capital -0.0013 *** (-10.55) Constant 0.0008 *** 0.0008 *** 0.0008 *** 0.0007 *** 0.0007 *** 0.0007 *** (4.25) (4.23) (4.07) (3.49) (3.69) (3.76) Number of obs 184619 165186 160722 146246 141169 141169 R2 0.0006 0.0002 0.0003 0.0002 0.0003 0.0003
Table 5: Panel regression of trading volume
This table reports the results of the panel regressions for the trading volume in the sample. The dependent variable is the natural logarithm of trading volume. The independent variables are the natural logarithm of assets, natural logarithm of employees, the return on investment, the debt/capital ratio and the dummy variables Ban, Fin and Ban*Fin.
Table 5: Panel regression of trading volume
lnVolume (1) (2) (3) (4) (5) (6) Ban 0.03545 *** 0.0668 *** 0.0696 *** 0.06434 *** 0.0664 *** 0.0579 *** (7.67) (13.95) (14.48) (12.61) (12.83) (11.12) Fin 1.0747 *** 1.1481 *** 0.6063 ** 0.5117 ** 0.5656 ** 0.6234 *** (3.34) (3.58) (2.49) (2.2) (2.45) (2.7) banFin -0.4358 *** -0.4464 *** -0.4418 *** -0.4516 *** -0.4492 *** (-24.42) (-24.91) (-24.36) (-24.97) (-24.85) lnassets 0.171 *** 0.134 *** 0.1234 *** 0.0953 *** (11.75) (8.19) (7.47) (5.73) lnemployees 0.1034 *** 0.1015 *** 0.0867 *** (15.88) (15.65) (13.19) RoI -0.0001 *** -0.0001 *** (-8.54) (-8.59) Debt/Capital 0.0044 *** (13.44) Constant 7.1111 *** 7.1061 *** 4.6493 *** 4.4401 *** 4.5997 *** 4.9655 *** (83.12) (83.25) (21.19) (18.36) (18.81) (20.2) Number of obs 169202 169202 164624 149797 144600 14460 R2 0.0237 0.0241 0.2980 0.3562 0.3500 0.3159
Table 6: First difference regression of trading volume
This table reports the results of the first difference regressions for the trading volume in the sample. The dependent variable is the natural logarithm of trading volume. The independent variables are the natural logarithm of assets, natural logarithm of employees, the return on investment, the debt/capital ratio and the dummy variables Ban, Fin and Ban*Fin.
Table 6: Panel regression of first differences in trading volume
lnVolumeDiff (1) (2) (3) (4) (5) (6) Ban -0.0266 *** -0.0265 *** -0.0266 *** -0.0266 *** -0.0269 *** -0.0269 *** (-6.08) (-5.82) (-5.80) (-5.58) (-5.54) (-5.54) Fin 0.0025 0.0028 0.0029 0.0034 0.0038 0.0038 (0.39) (0.41) (0.42) (0.48) (0.54) (0.54 ) banFin -0.0020 -0.0050 -0.0049 -0.0051 -0.0051 (-0.12) (-0.29) (-0.29) (-0.30) ( -0.30 ) difflnassets -0.3348 * -0.2022 -0.2119 -0.2068 (-1.92) (-0.97) (-1.01) (-0.99) difflnemployees -0.024 -0.0343 -0.0228 (-0.24) (-0.34) (-0.23) diffRoI -0.0002 * -0.0002 ** (-1.94) (-1.96) diffDebt/Capital -0.0032 (-0.96) Constant -0.0084 *** -0.0084 *** -0.0086 *** -0.009 *** -0.0089 *** -0.0089 *** (-4.55) (-4.53) (-4.59) (-4.62) (-4.51) (-4.51) Number of obs 165186 184618 170272 154957 149401 149401 R2 0.0002 0.0006 0.0372 0.0393 0.0296 0.0303
