Skip to main content
SpringerLink
Log in

A Worldwide Examination of Exchange Market Quality: Greater Integrity Increases Market Efficiency

  • Published:
Journal of Business Ethics Aims and scope Submit manuscript

Abstract

We develop a framework for assessing security market quality (MQ), relating five elements of market design to three metrics of market integrity and two metrics of market efficiency. We empirically implement this integrity–efficiency MQ framework by testing a hypothesis that trade-based ramping manipulation at the close (MTC) raises execution costs on 24 security markets worldwide. Estimating a simultaneous equations model of ramping incidence, spreads, and the probability of deploying real-time surveillance (RTS), we show that quoted bid-ask spreads are positively related to the incidence of MTC across seven liquidity deciles. The magnitude is economically significant; improving market integrity by cutting MTC in half reduces spreads 6–11 %. Allowing direct market access in conjunction with RTS, conducting auctions at the close, and developing regulations that require surveillance, all reduce MTC and thereby lower spreads, assuring better market integrity and enhancing market efficiency. Introducing circuit breakers or prohibiting shorts poses integrity–efficiency tradeoffs.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2

Similar content being viewed by others

Notes

  1. The other two are insider trading ahead of price-sensitive announcements and the front running of clients with whom an agency relationship exists.

  2. Trade-based manipulation is thereby distinguishable from other types of market manipulation such as the spreading of false rumors (Van Bommel 2003) and opportunistic disclosures of earnings forecasts and pro forma non-GAAP earnings (Miller 2008). Aitken et al. (2013) analyze the simultaneous equations relationship between ramping manipulation, spreads, and another integrity violation--namely, trading ahead of price-sensitive announcements (Bhattacharya and Daouk 2002).

  3. Trades at the offer are shown in red, and trades at the bid are shown in green. The size of the trade is indicated by the diameter of the circle. Blue circles represent off-market trades that can be negotiated at prices away from the continuous auction price beyond a minimum size. Trading volume is shown in the bar diagrams.

    SMARTS, a wholly owned subsidiary of the NASDAQ/OMX Group, is the leading equity market surveillance company world-wide, deployed in 40+ national exchanges and regulators and 150+ brokers across 35 countries. See www.smartsgroup.com.

  4. Aitken et al. (2007) provide empirical evidence supporting the implications of Foucault’s theory. They hypothesize that hedge funds and proprietary trading desks tend to have short-lived information about valuation and/or the state of the market. As a result, these classes of traders face higher costs of non-execution and lower picking-off risk than insurance companies and mutual funds. They hypothesize that the order placement strategy of proprietary trading desks and hedge funds will, therefore, be more aggressive than insurance companies and mutual funds. Using a large sample of trading desk instructions, they show that insurance companies and mutual funds do tend to be less aggressive order placers, ceteris paribus.

  5. Further discussion of recent integrity alert data appears in Aitken and Harris (2011).

  6. We thank a referee for suggesting this interpretation tying few observed manipulations in thinly traded stocks to the information cascade theory of trade-based manipulation.

  7. Relative liquidity is what really matters to trade-based manipulation incidence and detection. The least liquid stocks are much the same everywhere, and trade under much the same conditions everywhere. The most liquid stocks are somewhat less homogeneous across exchanges. But still the first liquidity decile stocks in New York are much more akin on spreads and trading behavior to the first decile stocks in Singapore than the first decile stocks in New York (or Singapore) are to the ninth or tenth decile stocks in New York (or Singapore). Table 3 makes this point. It shows NYSE decile 10 effective/and quoted spreads 11 and 21 times greater than NYSE decile 1 effective and quoted spreads, respectively. Similarly, Singapore decile 10 effective/quoted spreads are 21 and 39 times greater than Singapore decile 1 effective and quoted spreads, respectively. But decile 1 effective and quoted spreads in Singapore (82 and 86 b.p.) are only 5 times greater than in New York (16 and 15 b.p.). So, while there is some lack of homogeneity across exchanges, the greater concern in grouping data for microstructure analysis of AI and spreads is across the liquidity deciles. See Harris and Cook (2008).

  8. This observed 6.56 % mean estimate differs from the lognormal theoretical \(\mu_{\text{ES}}\)f 6.64 % because our sample differs slightly from an exact lognormal distribution.

  9. We elaborate this reasoning further in the second and third sections of the Data Appendix.

  10. Because the present study focuses on trade-based manipulation rather than insider trading or false rumor dissemination, the topic of information generation and disclosure (Pagano and Roell 1996; Lang and Lundholm 1996; and Oved 2002) is excluded from our research.

  11. The indirect effect on market efficiency of this improved market integrity should be tighter spreads, ceteris paribus, by H1.

  12. We retain in each specification all exchange dummies found to be significant at α < 0.05 in OLS regressions.

  13. This simultaneous structural equations model structure (I–III) for market quality research is debated by a panel of experts in Harris et al. (2011).

  14. Details of these Stock-Yugo test results for the AI and QSPR equations for each of the three sets of liquidity deciles are available from the authors.

  15. Table 5 conveys the cross-sectional moments for spreads, alert incidence, volatility, and turnover in 2005. These data are demonstrably non-normal as in the 2000–2005 sample reported earlier. For example, the quoted spread has a mean of 5.36 %, and a standard deviation of 0.0956 with skewness of 3.4166 and kurtosis of 13.1504. After a natural log transform, we observe the distribution of \({ \ln }({\text{Quoted Spread}})\) to be approximately normal \(\mu_{\text{LES}}\) (−3.9911 and \(\sigma_{\text{LQS}}\) 1.4776) with skewness of 0.2425 and kurtosis of −0.6729.

  16. Again, we thank the referees for suggesting this line of reasoning that variation in spreads traditionally represents variation in phenomena that affect liquidity.

  17. The same is true of CloseAuction and ShortProhibition, the other pair of collinear regressors with high r = −0.688 and double-digit variance inflation factors (again, see the OLS Appendix), but again both variables are statistically significant in all deciles.

  18. In other research, we have shown this market integrity approach extends to metrics of information leakage—i.e., trading ahead of price-sensitive announcements (Aitken and Harris 2011; Harris and Di Marco 2012; Aitken et al. 2013), and potentially to metrics of broker-agency conflict—i.e., front running and payment for order flow (e.g., Battalio and Loughran 2007).

References

  • Aggarwal, R., & Wu, G. (2006). Stock market manipulation. Journal of Business, 79(4), 1915–1953.

    Article  Google Scholar 

  • Aitken, M. J., Almeida, N., Harris, F. H. de B., & McInish, T. H. (2007). Liquidity supply in electronic markets. Journal of Financial Markets, 10, 144–168.

  • Aitken, M. J., Aspris, A., Foley, S., Harris, F. H. de B. (2013). The effects of algorithmic trading on security market quality. SSRN Working paper.

  • Aitken, M. J., & Harris, F. H. de B. (2011). Evidence-based policy making for financial markets: A fairness and efficiency framework for assessing market quality. Journal of Trading, 6, 22–31.

  • Allen, F., & Gale, D. (1992). Stock-price manipulation. The Review of Financial Studies, 5, 503–529.

    Article  Google Scholar 

  • Allen, F., Titov, L., & Mei, J. P. (2006). Large investors, price manipulation, and limits to arbitrage: An anatomy of market corners. Review of Finance, 10, 645–693.

    Article  Google Scholar 

  • Angel, J. (2013). Fairness in financial markets: The case of high frequency trading. Journal of Business Ethics, 112, 585–595.

    Article  Google Scholar 

  • Angel, J., & McCabe, D. (2010). The ethics of speculation. Journal of Business Ethics, 90, 277–286.

    Article  Google Scholar 

  • Battalio, R., & Loughran, T. (2007). Does payment for order flow to your broker help or hurt you. Journal of Business Ethics, 80, 37–44.

    Article  Google Scholar 

  • Bernhardt, D., & Davies, R. J. (2005). Painting the tape: Aggregate evidence. Economic Letters, 89, 306–311.

    Article  Google Scholar 

  • Bhattacharya, U., & Daouk, H. (2002). The world price of insider trading. Journal of Finance, 7(1), 75–108.

    Article  Google Scholar 

  • Brown, P., Beekes, W., & Verhoeven, P. (2011). Corporate governance, accounting and finance: A review. Accounting and Finance, 51(1), 96–172.

    Article  Google Scholar 

  • Carhart, M. M., Kaniel, R., Kusto, D. K., & Reed, A. V. (2002). Leaning for the tape: Evidence of gaming behavior in equity mutual funds. Journal of Finance, 57(2), 661–693.

    Article  Google Scholar 

  • Chakraborty, A., & Yilmaz, B. (2004). Manipulation in a market order market. Journal of Financial Markets, 7, 187–206.

  • Chakraborty, A., & Yilmaz, B. (2008). Microstructure bluffing with nested information. American Economic Review, 98(2), 280–284.

  • Chakravarty, S., Harris, F. H. de B., & Wood, R. A. (2009). The changing state of the NYSE market: New common factors, players, and impulse responses. Journal of Trading, 4(1), 68–94.

  • Chamberlain, T. W., Chueng, C. S., & Kuan, C. C. Y. (1989). Expiration-day effects on index futures and options: Some Canadian evidence. Financial Analysts Journal, 45, 67–71.

    Article  Google Scholar 

  • Chiou, J. S., Wu, P. S., Wang, A. W., & Huang, B. Y. (2007). The asymmetric information and price manipulation in stock markets. Applied Economics, 39, 883–891.

    Article  Google Scholar 

  • Comerton-Forde, C., & Rydge, J. (2006). The influence of call auction algorithm rules on market efficiency. Journal of Financial Markets, 9, 199–222.

  • Comerton-Forde, C. & Putnins, T. (2013). Stock price manipulation: Prevalence and determinants. Review of Finance, 17(6), 1–44.

  • Comerton-Forde, C., & Putnins, T. (2011). Measuring closing price manipulation. Journal of Financial Intermediation, 20, 135–158.

    Article  Google Scholar 

  • Cumming, D., & Johan, S. (2008). Global market surveillance. American Law and Economics Review, 10, 454–506.

    Article  Google Scholar 

  • Cumming, D., & Li, D. (2011). Exchange trading rules and stock market liquidity. Journal of Financial Economics, 98, 651–671.

    Article  Google Scholar 

  • Cushing, D. (2001). The hidden cost of trading at the close. Journal of Trading, 2001, 12–19.

    Google Scholar 

  • Cushing, D., & Madhavan, A. (2000). Stock returns and trading at the close. Journal of Financial Markets, 3, 45–67.

    Article  Google Scholar 

  • Eleswarapu, V., & Venkataraman, K. (2006). The impact of political and legal institutions on equity trading costs. Review of Financial Studies, 19, 1081–1111.

    Article  Google Scholar 

  • Felixson, K., & Pelli, A. (1998). Day end returns—stock price manipulation. Journal of Multinational Financial Management, 9, 95–127.

    Article  Google Scholar 

  • Foucault, T. (1999). Order flow composition and trading costs in a dynamic limit order market. Journal of Financial Markets, 2, 99–134.

    Article  Google Scholar 

  • Harris, F. H. de B., & Cook, R. (2008). Execution costs and market design worldwide. Journal of Trading, 3(1), 9–25.

  • Harris, F. H. de B., Aitken, M., Berekeley, A., Overdahl, J., & Venkataraman, K. (2011). A panel discussion of a proposed framework for assessing market quality. Journal of Trading, 6, 69–89.

  • Harris, F. H. de B., & DiMarco, E. (2012). European market quality pre/post MiFID1. Journal of Trading, 7(4), 7–26.

  • Hart, O. D., & Kreps, D. M. (1986). Price destabilizing speculation. Journal of Political Economy, 94(6), 927–952.

    Article  Google Scholar 

  • Hillion, P., & Souminen, M. (2004). The manipulation of closing prices. Journal of Financial Markets, 7, 351–375.

    Article  Google Scholar 

  • Ji, S. (2009). Security Market Manipulation and the Assurance of Market Integrity. Ph.D. dissertation, University of New South Wales.

  • Lang, M. H., & Lundholm, R. J. (1996). Corporate disclosure policy and analyst behavior. The Accounting Review, 71(4), 467–493.

    Google Scholar 

  • Miller, J. S. (2008). Opportunistic disclosures of earnings forecasts and non-GAAP earnings measures. Journal of Business Ethics, 89, 3–10.

    Article  Google Scholar 

  • Murray, M. P. (2006). Avoiding invalid instruments and coping with weak instruments. Journal of Economic Perspectives, 20(4), 111–132.

    Article  Google Scholar 

  • Murray, M. P. (2011). Avoiding the pitfalls of instrumental variable estimation with few or many instruments. SSRN Working Paper.

  • Oved, Y. (2002). Information disclosure costs and the choice of financing source. Journal of Finance, 4(1), 3–21.

    Google Scholar 

  • Pagano, M., & Roell, A. (1996). Transparency and liquidity: A comparison of auction and dealers markets with informed trading. Journal of Finance, 51(2), 579–611.

    Article  Google Scholar 

  • Pagano, M. S., & Schwartz, R. A. (2003). A closing call’s impact on market quality at Euronext Paris. Journal of Financial Economics, 68, 439–484.

    Article  Google Scholar 

  • Stoll, H. R., & Whaley, R. E. (1987). Program trading and expiration-day effects. Financial Analysts Journal, 43(2), 16–28.

    Article  Google Scholar 

  • Van Bommel, J. (2003). Rumors. Journal of Finance, 58, 1499–1521.

    Article  Google Scholar 

Download references

Acknowledgments

Tom Smith, Peter Swan, Laurent Fournier at NYSE Euronext Paris Research, Frank Hathaway at NASDAQ, Wei-man Liu, and Doug Cumming at the 2012 CFA conference provided invaluable feedback. A panel discussion on “Market Integrity” at the 2010 FMA Meetings in New York with James Overdahl and Kumar Venkataraman as well as presentations at the SEC, NASDAQ, NYSE Euronext Paris, Six Group, and the Italian CONSOB have helped refine our proposed market quality framework. Two anonymous referees made insightful criticisms that substantially improved the paper.

Author information

Authors and Affiliations

  1. Capital Markets Cooperative Research Centre and Australian School of Business, University of New South Wales, Sydney, NSW, Australia

    Michael J. Aitken

  2. Wake Forest University School of Business, Winston-Salem, NC, USA

    Frederick H. de B. Harris

  3. Capital Markets Cooperative Research Centre, Sydney, NSW, Australia

    Shan Ji

Authors
  1. Michael J. Aitken
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Frederick H. de B. Harris
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Shan Ji
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Michael J. Aitken.

Appendices

Data Appendix

Appendix 1: Calculating Alert Incidence (AI)

Alert Input

  • Last trade price of the day (LPd).

  • Last available trade price at 15 min (LPd,−15 min) before the continuous trading period ends.

  • The best bid price and the best ask price at the end of the continuous trading period.

  • Close price (CPd).

  • Opening price of the next trading day (OPd+1).

Alert Logic

Part 1 Detect Abnormal End-of-day Price Change Instances

  • For exchanges without closing auctions, end-of-day price change (Pchgt) is defined as the log change of price during the last 15 min of trading.

    $${\text{Pchg}}_{\text{d}} = \, \ln \left( {{\text{LP}}_{\text{d}} /{\text{LP}}_{{{\text{d}}, - 15\hbox{min} }} } \right).$$

  • For exchanges with closing auctions, end-of-day price change (Pchgt) is defined as the log change from the last available trade price 15 min before continuous trading ends day (LPd.−15 min) to the close price (CPd).

    $${\text{Pchg}}_{\text{d}} = \, \ln \left( {{\text{CP}}_{\text{d}} /{\text{LP}}_{{{\text{d}}, - 15\hbox{min} }} } \right).$$

  • For each stock and trading day, the end-of-day price change is compared to a distribution of historical end-of-day price changes, which occurred during the last month of trading. When an end-of-day price change exceeds 3 standard deviations away from the mean of the historical distribution, that trading day is flagged as an instance of abnormal end-of-day price change.

Part 2 Detect Abnormal Closing Auction (for exchanges with closing auction)

  • Calculate the pre-close to close price change as the log change from last trade price of the day (LPd) to the close price (CPd).

    $${\text{PchgClose}}_{\text{d}} = { \ln }\left( {{\text{CP}}_{\text{d}} /{\text{LP}}_{\text{d}} } \right).$$

  • For each stock and trading day, the pre-close to close price change is compared to a distribution of historical pre-close to close price changes, which occurred during the last month of trading. When a pre-close to close price change exceeds 3 standard deviations above (below) the mean of the historical distribution and the close price is above or at (below or at) the best ask (bid) price at the end of the continuous trading, that trading day is flagged as an instance of abnormal closing auction.

Part 3 Detect Ramping the next trading day

  • A trading day of abnormal end-of-day price change or abnormal pre-close to close price change, which is followed by a price reversion (PRd+1) of 50 % at the opening of the next trading day, is considered a successful ramping attempt.

  • Price reversion for exchanges without closing auction is defined as

    $${\text{PR}}_{{{\text{d}} + 1}} = ({\text{LP}}_{\text{d}} - {\text{OP}}_{{{\text{d}} + 1}} )/({\text{LP}}_{\text{d}} - {\text{LP}}_{{{\text{d}}, - 15\hbox{min} }} ) \, * \, 100\;\% .$$

  • Price reversion for exchanges with closing auction is defined as

$${\text{PR}}_{{{\text{d}} + 1}} = ({\text{CP}}_{\text{d}} - {\text{OP}}_{{{\text{d}} + 1}} )/({\text{CP}}_{\text{d}} - {\text{LP}}_{\text{d}} )*100\;\% .$$

Appendix 2: False Positives in Ramping Manipulation Alerts

To decipher which ramping alerts are unrelated to manipulation and which should be referred to regulators for investigation, exchange operators typically use broker ID and order-level position data. We wish to base the testing of our manipulation-spreads hypothesis on publicly available data alone. Nevertheless, attention to this well-known problem in the surveillance industry is certainly warranted if false positives measurement error is systematically related to the frequency of manipulation.

An inescapable fact of the random sampling of returns to identify ramping manipulation alerts is that unmanipulated stocks are occasionally included in alerts, and then manipulation incidence is overestimated. Similarly, manipulated stocks are identified only when their manipulation return is statistically rare in magnitude, so their alert incidence is underestimated. The rate of false positives in manipulation alerts is, therefore, negatively related to the true frequency of manipulation. This systematic measurement error in ramping alerts is only a problem for testing H10 if it is systematically related to spreads.

When a manipulator seeks to ramp and then dump (unwind) a position by executing a sequence of buyer-initiated then seller-initiated trades at the offer and then at the bid, this causes trade prices to walk the book and leads to returns that tend to be statistically rare. So, effective spreads do temporarily increase for that security-day in a mechanical ex post way. We instead test H10 ex ante by modeling the effects of manipulation on quoted spreads. Ex ante, manipulation might either discourage participation harming liquidity, thereby triggering lots of price protection and widening the quoted spread. Or alternatively, ex ante, manipulation might attract knock-on day traders to the manipulation and improve liquidity, thereby narrowing the quoted spread. So, it is an empirical question which of these forces dominates. Quoted spreads are, therefore, not mechanically related to manipulation alerts.

Appendix 3: Information Arrivals and False Rumors

In theory, information arrivals that are quickly reversed would lead to no persistent change in quoted spreads, and controlling for induced volatility, no change in effective spreads either. We can, therefore, argue that any ramping alert incidence (AI) mistakenly capturing false information arrivals that are quickly reversed, would be unrelated to spreads. Nevertheless, there are some observational errors of measurement inherent in surveillance practice. For example, false rumors can trigger trade-based ramping manipulation alerts. Therefore, we prescreen the Thomson Reuters database for information announcements the day prior to the ramping event to remove instances where a misleading or partial disclosure could if corrected or elaborated overnight lead to a statistically rare closing return followed by a next-day reversal.

In contrast, confirmed information arrivals (both positive and negative) trigger herding and other information-based trading that cause market prices to trend (rather than mean revert). Markets that trend do exhibit increased spreads. This is for two reasons: (1) quasi market-makers “go flat” reducing their open trading positions, and (2) liquidity providers protect themselves against the risk of their bids or offers being picked off by more informed traders. So, we expect spreads to be positively correlated with the natural or inherent volatility set off by frequent information arrivals. Inherent price volatility, therefore, proves to be an important control variable in our empirical study. But this is distinct conceptually from the artificial price volatility set off by market manipulation which we hypothesize will widen quoted (not just effective) spreads.

Appendix of OLS and Probit Estimates

See Table 7.

Table 7 Panels A, B and C present OLS and Probit parameter estimates and robust standard error tests for the various specifications of the three structural equations for ramping alert incidence (AI), quoted spread (QSPR), and the probability of adoption of real-time surveillance Prob(RTS) reported in Table 6. Observations are grouped by thickly traded deciles 1, 2, 3, less liquid deciles 4, 5, 6, 7, and thinly traded deciles 8, 9, 10. The fixed effects column identifies all exchange-specific dummy variables significant at 95 % in the OLS and Probit estimates. Variance inflation factors are displayed in italics below the OLS t-scores
Full size table

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Aitken, M.J., de B. Harris, F.H. & Ji, S. A Worldwide Examination of Exchange Market Quality: Greater Integrity Increases Market Efficiency. J Bus Ethics 132, 147–170 (2015). https://doi.org/10.1007/s10551-014-2294-5

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10551-014-2294-5

Keywords

JEL Classification

Search

Navigation