Intra Day Bid‐Ask Spreads, Trading Volume and Volatility: Recent Empirical Evidence from the London Stock Exchange

With the benefit of very high frequency (25 million 1 minute observations) and recent data (2001) for the UK, this paper explores a number of intra day patterns of stock market behaviour. More specifically, a distinct reverse J shaped bid‐ask spread pattern is noted for SETS securities, a declining bid‐ask spread pattern for non‐SETS securities, a two hump pattern for trading volume and a U‐shaped pattern for returns volatility for all securities. In terms of complementing the existing literature, the paper shows that differences in trading systems may affect the bid‐ask spread patterns, while differences in market environments (i.e. US and UK markets) seems to affect the trading volume pattern. The paper suggests avenues for future research, in particular, the need to consider what factors are significant in determining intra day patterns for different trading systems and the need for additional cross‐market comparisons to identify how institutional factors affect the behaviour of investors on an intra day basis.     

A Note on the Intraday Patterns of Initial Public Offerings: Evidence from Hong Kong

This paper examines the intraday patterns of IPOs in Hong Kong during the period 1995–1998. The results reveal that the well‐known under‐pricing phenomenon of IPOs occurs only at the opening trading of new issues and vanishes afterwards. The return volatility of IPOs is found to be high during the first trading session, and declines rapidly during the rest of the first trading day until the end of the trading day. The intraday return volatility of IPOs is found to follow a double U‐shape pattern, which is similar to that of the general market. A great deal of trading activity was recorded during the first five minutes of the trading day. Consistent results are obtained for IPOs registered during the pre‐crisis and post‐crisis periods. This paper has practical implications for investors. Investors can benefit from the under‐pricing only if they subscribe for new shares in the primary market. There is, however, no profit‐making opportunity for day traders who buy shares on the first trading day. This shows that the Hong Kong market is efficient in adjusting for the IPO under‐pricing. In addition, it is likely that, because of Hong Kong's share allotment method, only big investors who apply for large numbers of shares can benefit from this under‐pricing phenomenon.     

FOREX Risk: Measurement and Evaluation Using Value‐at‐Risk

Abstract:   We measure and evaluate the performance of a number of Value‐at‐Risk (VaR) methods using a portfolio based on the foreign exchange exposure of a small open economy (Ireland) among its trading partners. The sample period highlights the changing nature of Ireland's exposure to risk over the past decade in the run‐up to EMU. Our results offer an indication of the level of accuracy of the various approaches and discuss the issues of models ensuring statistical accuracy or more conservative leanings. Our findings suggest that the Orthogonal GARCH model is the most accurate methodology while the EWMA specification is the more conservative approach.     

Tests of Weak‐Form Efficiency of the Dhaka Stock Exchange

Abstract:   Conflicting evidence on weak form efficiency of the Dhaka Stock Market appears to stem from the use of monthly versus daily data, structural changes after the 1996 market crash, and the use of tests with or without heteroscedasticity adjustment. Heteroscedasticity‐robust tests indicate short‐term predictability of share prices prior to the crash, but not afterwards. Although a heteroscedasticity‐robust Box‐Pierce test was used by Lo and MacKinlay (1989) in their simulations, our study appears to be the first to apply this test to stock prices. Typical rejection of weak‐form market efficiency by the usual autocorrelation tests may be reversed by a heteroscedasticity‐robust test.     

A note on transaction costs and the interpretation of dividend drop‐off ratios

In a recent edition of this Journal, Bartholdy and Brown (1999) presented an analysis of the ex‐dividend share price behaviour of shares listed on the New Zealand Stock Exchange. The authors conclude that their results are consistent with the tax clientele effect (driven by long‐term investors) and that there is little or no support for the short‐term trading hypothesis. Our purpose is to highlight the importance of transaction costs in analyses such as Bartholdy and Brown's. We argue that their results have an alternative interpretation because their analysis excludes the impact of transaction costs. We extend their model to include transaction costs and show that their results are not necessarily inconsistent with the short‐term trading hypothesis. A critical point of our analysis is that, in the presence of transaction costs, the equilibrium drop‐off ratio for dividend strip traders will be less than one, and, in some cases, can be less than the equilibrium drop‐off ratio for long‐term investors.     

Odd‐lot Costs and Taxation Influences on Stock Dividend Ex‐dates

Abstract:   Past research has revealed significant abnormal ex‐date returns for stock dividends even though the ex‐date is known in advance and the distribution contains no new information. Various researchers have suggested that the higher transaction cost of selling odd‐lot share parcels compared to round‐lot share parcels is a key driver in the abnormal returns. However, no study to date has directly compared the ex‐date price reaction of stock dividends distributed when odd‐lot transaction costs were charged to those issued when odd‐lot costs were not evident. As odd‐lot trade costs were eliminated from the New Zealand Stock Exchange on 1 October, 1991, the New Zealand market provides a unique opportunity to directly test the role, if any, that odd‐lot transactions costs have in explaining stock dividend ex‐date returns. We find that prior to October 1991 stock dividend ex‐dates exhibit significantly positive returns, however, we do not find any significant ex‐date return once the higher odd‐lot transaction costs were removed. The New Zealand market also enables us to examine an imputation tax based argument of the ex‐date price reaction and we find evidence that imputation tax credits have a value greater than zero.     

Insider Trading and Pay‐Performance Sensitivity: An Empirical Analysis

Abstract:   We examine whether the sensitivity of pay to performance is associated with the amount of insider trading that managers undertake. Because insider trading profits represent an alternative form of compensation, we expect that firms will consider the compensation component provided by insider trading when designing remuneration contracts. Employing a proxy for insider trading that captures the degree to which managers trade on private information, we find evidence that an increased (a decreased) level of insider trading is associated with a decreased (an increased) pay‐performance sensitivity.     

Are Economically Significant Stock Returns and Trading Volumes Driven by Firm‐specific News Releases?

This paper explores whether firm‐specific information events drive economically relevant positive and negative stock price changes and trading volume and, if so, the nature of such information. We find that no less than 65% of significant price changes and trading volume movements in our sample of FTSE 350 companies can be readily explained by public domain information contradicting the thesis that corporate news is not a primary driver, and that share price changes and trading volume activity are driven by factors unrelated to information flows per se. In addition, we find that a parsimonious set of news categories represent the key drivers. Sell‐side analyst stock recommendations and earnings forecast revisions as a class, unaccompanied by other news releases, dominate all other news categories in terms of significant market reaction. However, taking into account the relative magnitude of market response to different news releases, firms' formal accounting disclosures dominate within this domain. As such, we conclude these are not fully anticipated by apparently more timely market disclosures, and that the existence of news services and the activities of the sell‐side analyst are not substitutes for a firm's interim and preliminary results.     

Re‐assessing the long‐term underperformance of UK Initial Public Offerings

Previous work has identified that IPOs underperform a market index, and the purpose of this paper is to examine the robustness of this finding. We re‐examine the evidence on the long‐term returns of IPOs in the UK using a new data set of firms over the period 1985–92, in which we compare abnormal performance based on a number of alternative methods including a calendar‐time approach. We find that, using an event‐time framework, there are substantial negative abnormal returns to an IPO after the first 3 years irrespective of the benchmark used. However, over the 5 years after an IPO, abnormal returns exhibit less dramatic underperformance, and the conclusion on negative abnormal returns depends on the benchmark applied. Further if these returns are measured in calendar time, we find that the (statistical) significance of underperformance is even less marked.     

The Impact of Mandated Cash Flow Disclosure on Bid‐Ask Spreads

Abstract:   This study provides evidence that mandatory cash flow disclosure required by Approved Australian Accounting Standard AASB 1026, Statement of Cash Flows (June 1992) was associated with a decline in bid‐ask spreads following the introduction of the regulation, even after controlling for changes in trading volume and price volatility. More pronounced decreases in bid‐ask spreads were associated with firms having lower correlations between reported CFO and estimates of CFO using balance sheet reconstructions. We conclude that mandatory cash flow disclosure reduces information asymmetry across market participants.     

Write‐Down Timeliness, Line‐of‐Business Disclosures and Investors' Interpretations of Segment Divestiture Announcements

Abstract:   This paper examines investors' anticipation and subsequent interpretations of asset write‐downs accompanying segment divestitures. Examining long‐window returns cumulated over the two years preceding the year of divestiture, we hypothesize and find that investors anticipate write‐downs of segment operating assets before divestiture and recognition occurs, with anticipation conditional on the timeliness of the write‐down and prior disclosure of the segments' operating results under segment reporting rules. Short‐window returns cumulated over the three days surrounding the announcement of the divestiture confirm that investor interpretations of asset write‐downs are similarly contingent on write‐down timeliness and prior disclosure.     

Net Present Value-Consistent Investment Criteria Based on Accruals: A Generalisation of the Residual Income‐Identity

Abstract:   In recent years, many firms have favoured residual income for value based management. One main argument for this measure is its identity with the net present value rule and that this compatibility with the net present value rule holds true for all possible depreciation schedules selected. In this article, we analyse whether there are other, undiscussed, accrual accounting numbers that enable net present value‐consistent investment decisions for all possible depreciation schedules. Our analysis provides an if‐and‐only‐if characterisation of the entire class of net present value‐consistent investment criteria, based on accounting information. This provides new insights into the residual income concept, hurdle rates, opening and closing error conditions achieved by applying more common performance measure structures, and allocation rules. Moreover, our analysis shows the limits of constructing such investment criteria.     

HOLIDAY CLOSINGS AND SECURITY RETURNS

This paper documents unusual return patterns for securities around holiday closings. Returns for trading days immediately before holiday closings (pre-holiday trading days) are unusually high regardless of weekday, year, or holiday closing. Returns for trading days following holiday closings (post-holiday trading days) are high only if they occur at the end of the week. Tests indicate that pre-holiday returns do not respond to a closing effect, and that the post-holiday returns do not result from a time-diffusion process. Holiday trading day returns question the tax-loss selling explanation of the turn-of-the-year effect and display a significant small firm effect outside of January.     

Decomposing and Testing Long‐term Returns: an Application on Danish IPOs

An improved method for measuring and testing long‐run returns is proposed. The method adjusts for the right‐skewed distribution of long‐run buy‐and‐hold by decomposing average cross‐sectional buy‐and‐hold returns into mean components and volatility components. The method is applied to initial public offerings in Denmark. The mean‐component under performance of initial public offering stocks compared to the market is 30% and significant after 5 years. Compared to matching firms the under performance of IPO stocks is 13% after 5 years but insignificant.     

Accruals, Cash Flows and the Post‐Earnings‐Announcement Drift

Abstract:   Several prior studies have shown that cash flows have significantly greater impact on stock prices than accruals. We examine the implications of these findings for the post‐earnings‐announcement‐drift anomaly. We argue that, if investors under‐react to earnings news, then the larger price impact of cash flows causes the cash flow component of earnings news to predict future returns better than the accruals component. Consistent with this argument, we show that unexpected cash flows are more positively related to future returns, than are unexpected accruals. Also, unexpected cash flows are found to predict future returns above and beyond that predicted by earnings surprises. Finally, we show that a strategy that decomposes earnings news into its components significantly outperforms strategies based on earnings news alone. The results support under‐reaction explanations for the drift.     

Information Asymmetry and the Bid‐Ask Spread: Evidence From the UK

Abstract:   The generally accepted factors that determine the bid‐ask spread are volatility, trading volume and market value ( Atkins and Dyl, 1997 ; Glosten and Harris, 1988 ; and Menyah and Paudyal, 2000 ). Following Kim and Verrecchia (1994) we include a measure of the disagreement in analysts' earnings forecasts in our model of the bid ask spread. This measure serves as a proxy for the informational disadvantage of market makers with respect to informed traders. Market makers respond to the additional risk by increasing the bid‐ask spread. We find that the disagreement amongst analysts is significant for horizons up to and including six months (and with the hypothesised sign) in explaining FTSE 100 company spreads, rendering strong empirical support for our model.     

The effect of US holidays on the European markets: when the cat’s away…

This paper presents evidence of the existence of a return effect on European stock markets coinciding with New York Stock Exchange (NYSE) holidays, which is particularly marked after positive closing returns on the NYSE the previous day. The effect is large enough to be exploited by trading index futures. This anomaly cannot be explained by seasonal effects, such as the day of the week effect, the January effect or the pre‐holiday effect, nor is it consistent with behavioural finance models that predict positive correlation between trading volume and returns. However, examination of factors such as information volume or investor mix provides a reasonable explanation.     

Not‐for‐Profit Annual Reports: What do Museum Managers Communicate?

This study addresses: (1) What disclosures are provided in annual reports of not‐for‐profit entities?(2) What characteristics of the reporting entities explain variations in the quantity of financial disclosure?(3) How do not‐for‐profit disclosures compare with those in for‐profit corporate reports? The annual reports of 170 not‐for‐profit museums were examined. The reports were highly variable. Some contained no financial data and only 22 percent included complete financial statements with footnotes. Regression analysis indicated that the amount of museum financial data was positively associated with museum size, a larger number of pages of donor disclosures, and museum type (art and history, but not science, natural history, or general).     

Intraweek and intraday trade patterns and dynamics

This paper investigates informed traders' order-splitting strategies on different days of the week and times of the day for a sample of stocks traded on the Australian Stock Exchange. Based on cumulative price changes, we document that informed traders tend to use medium size trades. We find that informed investors concentrate their strategic trading on Mondays and particularly during the first trading hour. In addition, informed investors also use large size trades around market opening and closing, as well as on days other than Mondays and Fridays. These results are more pronounced for the large market capitalization stocks.     

Market Microstructure Empirical Regularities: Behavior of the Bid-Ask Spread and Closing Prices

A substantial number of last reported transactions for stocks trading on the New York Stock Exchange occur inside the quoted closing bid-ask spread. The tendency to close inside the spread results in price change magnitudes much smaller than those predicted from binomial models. Moreover, although the change magnitude is biased by the underlying trend of the market, the distribution of next day price change relatives is largely unaffected. The result is a systematic regularity between the location of today's close and tomorrow's close relative to the bid-ask spread.