Publications

Price discovery without trading: Evidence from limit orders

We analyze the contribution to price discovery of market and limit orders by high‐frequency traders (HFTs) and non‐HFTs. While market orders have a larger individual price impact, limit orders are far more numerous. This results in price discovery occurring predominantly through limit orders. HFTs submit the bulk of limit orders and these limit orders provide most of the price discovery. Submissions of limit orders and their contribution to price discovery fall with volatility due to changes in HFTs’ behavior. Consistent with adverse selection arising from faster reactions to public information, HFTs’ informational advantage is partially explained by public information.

The effects of uncertainty on market liquidity: Evidence from Hurricane Sandy

We test the effects of uncertainty on market liquidity using Hurricane Sandy as a natural experiment. Given the unprecedented strength, scale, and nature of the storm, the potential damages of a landfall near the Greater New York area were unpredictable and therefore uncertain. Using a difference-in-differences setting, we compare the market reactions of Real Estate Investment Trusts (REITs) with and without properties in the widely published evacuation zone of New York City prior to landfall. We find relatively less trading and wider bid-ask spreads in affected REITs. The results confirm theory on the detrimental effects of uncertainty on market functioning.

High frequency trading and extreme price movements

Are endogenous liquidity providers (ELPs) reliable in times of market stress? We examine the activity of a common ELP type—high frequency traders (HFTs)—around extreme price movements (EPMs). We find that on average HFTs provide liquidity during EPMs by absorbing imbalances created by non-high frequency traders (nHFTs). Yet HFT liquidity provision is limited to EPMs in single stocks. When several stocks experience simultaneous EPMs, HFT liquidity demand dominates their supply. There is little evidence of HFTs causing EPMs.

High frequency trading and the 2008 short-sale ban

We examine the effects of high-frequency traders (HFTs) on liquidity using the September 2008 short sale-ban. To disentangle the separate impacts of short selling by HFTs and non-HFTs, we use an instrumental variables approach exploiting differences in the ban’s cross-sectional impact on HFTs and non-HFTs. Non-HFTs’ short selling improves liquidity, as measured by bid-ask spreads. HFTs’ short selling has the opposite effect by adversely selecting limit orders, which can decrease liquidity supplier competition and reduce trading by non-HFTs. The results highlight that some HFTs’ activities are harmful to liquidity during the extremely volatile short-sale ban period.

Trading Fast and Slow: Colocation and Liquidity

We exploit an optional colocation upgrade at NASDAQ OMX Stockholm to assess how speed affects market liquidity. Liquidity improves for the overall market and even for noncolocated trading entities. We find that the upgrade is pursued mainly by participants who engage in market making. Those that upgrade use their enhanced speed to reduce their exposure to adverse selection and to relax their inventory constraints. In particular, the upgraded trading entities remain competitive at the best bid and offer even when their inventories are in their top decile. Our results suggest that increasing the speed of market-making participants benefits market liquidity.

The impact of computerized agents on immediate emotions, overall arousal and bidding behavior in electronic auctions

The presence of computerized agents has become pervasive in everyday live. In this paper, we examine the impact of agency on human bidders’ affective processes and bidding behavior in an electronic auction environment. In particular, we use skin conductance response and heart rate measurements as proxies for the immediate emotions and overall arousal of human bidders in a lab experiment with human and computerized counterparts. Our results show that computerized agents mitigate the intensity of bidders’ immediate emotions in response to discrete auction events, such as submitting a bid and winning or losing an auction, as well as the bidders’ overall arousal levels during the auction. Moreover, bidding behavior and its relation to overall arousal are affected by agency: whereas overall arousal and bids are negatively correlated when competing against human bidders, this relationship is not observable for computerized agents. In other words, lower levels of agency yield less emotional behavior. The results of our study have implications for the design of electronic auction platforms and markets that include both human and computerized actors.

High-frequency trading and price discovery

We examine the role of high-frequency traders (HFTs) in price discovery and price efficiency. Overall HFTs facilitate price efficiency by trading in the direction of permanent price changes and in the opposite direction of transitory pricing errors, both on average and on the highest volatility days. This is done through their liquidity demanding orders. In contrast, HFTs’ liquidity supplying orders are adversely selected. The direction of HFTs’ trading predicts price changes over short horizons measured in seconds. The direction of HFTs’ trading is correlated with public information, such as macro news announcements, market-wide price movements, and limit order book imbalances.

Interactive data: Technology and cost of capital

We examine the introduction of the voluntary filing program (VFP) by the Securities and Exchange Commission (SEC) for the introduction of XBRL (eXtensible Business Reporting Language), or Interactive Data as called in the US. XBRL is a machine-readable standardized format for financial reports. The VFP allowed firms to file annual and quarterly reports using XBRL. This program represents a quasi-natural experiment to isolate the effects of an improvement in the information environment of program participants. We study two documented effects of voluntary disclosure, reduced cost of capital and increased information intermediation. Our results show a decrease in the cost of capital, especially for financial and IT firms, and an increase in information intermediation. These effects support existing literature on the adoption of IT in firms and voluntary corporate disclosure and sheds light on the decision to be an early adopter of XBRL reporting technologies.

Latency, liquidity and price discovery

The speed of trading is an important factor in modern security markets, although relatively little is known about the effect of speed on liquidity and price discovery, two important aspects of market quality. On April 23, 2007, Deutsche Boerse made an important upgrade to their trading system. With the 8.0 release of Xetra, system latency was reduced from 50 ms to 10 ms. Subsequently, both quoted and effective spreads decreased, which are mainly concentrated in small- and medium-sized stocks. This increase in liquidity is due to dramatically lower adverse selection costs that were only partially translated into higher liquidity. We interpret this as a decrease in the competition between liquidity suppliers who are able to increase their revenues by more than 90 million euros. The contribution of quotes to price discovery doubles to 90% post upgrade, indicating that prices are more efficient.

Interactive Data: Technology and cost of capital

We examine the introduction of the voluntary filing program (VFP) by the Securities and Exchange Commission (SEC) for the introduction of XBRL (eXtensible Business Reporting Language), or Interactive Data as called in the US. XBRL is a machine-readable standardized format for financial reports. The VFP allowed firms to file annual and quarterly reports using XBRL. This program represents a quasi-natural experiment to isolate the effects of an improvement in the information environment of program participants. We study two documented effects of voluntary disclosure, reduced cost of capital and increased information intermediation. Our results show a decrease in the cost of capital, especially for financial and IT firms, and an increase in information intermediation. These effects support existing literature on the adoption of IT in firms and voluntary corporate disclosure and sheds light on the decision to be an early adopter of XBRL reporting technologies.

Technology and market quality: The case of high frequency trading

Technological innovations such as high frequency trading systems (HFT) and algorithmic trading systems have changed financial markets. We discuss technological and economic aspects of HFT and find that HFT have a major impact on all aspects of the internal market structure of exchanges. We use market quality measures on a unique dataset provided by NASDAQ in order to analyze the contribution of HFT to market quality. The empirical results are discussed in the context of external and internal factors of market quality. Our results indicate considerable differences in trading strategies: HFT engage in market making strategies and provide liquidity when it is expensive and demand liquidity when it is cheap. Their trades are more informed than non-HFT trades for stocks with a high market capitalization and therefore make prices more informative, but less informed across the entire sample.

Participation, feedback & incentives in a competitive forecasting community

Macroeconomic forecasts are used extensively in industry and government despite the lack of accuracy and reliability. Prediction markets as a community forecasting method have begun to gain interest in academia industry alike. An open question is how to design incentive schemes and feedback mechanisms to motivate online communities to contribute and thereby increase the predictive power of the market. We design a prediction market for macroeconomic variables that aggregates information from a cross-section of participants. We analyze participation and feedback in this online community. We show that a weekly newsletter that acts as a reminder drives participation. In public goods projects participation feedback has been found to increase participants’ contributions. We find that the competition inherent in markets appears to dominate classical feedback mechanisms. We show that forecast errors fall over the prediction horizon. The market-generated forecasts compare well with the Bloomberg- survey forecasts, the industry standard. Additionally we can predict community forecast error using an implicit market measure.

Price efficiency in futures and spot trading: The role of information technology

During the last years information technology has had a profound impact on financial markets. The speed of trading and the amount of available information has increased substantially. Nearly all exchanges have upgraded their trading systems to meet the demand of investors and enhance their competitive position. However, the impact on liquidity and price efficiency remains unclear. In this paper we present an event study to examine the effects of an infrastructure change at the Deutsche Börse in Germany. On April 23, 2007, Deutsche Börse released an upgraded version of their electronic trading system Xetra. We study the impact that this upgrade had on the efficiency of prices, measured as the pricing gaps between the observed futures prices and their theoretical values based on the underlying cash market. Our results suggest that the system upgrade reduced the pricing gapand thus improved price efficiency.

Mispricing and exchange market systems: The effect of infrastructure upgrades

We study the effects of an infrastructure change at the Deutsche Boerse in Germany. On April 23rd 2007, Deutsche Boerse released an upgraded version of their electronic trading system Xetra. We examine the impact that this upgrade had on the efficiency of prices, measured as the level of mispricing, in the underlying liquid cash market and derivative futures market. Our analysis shows that the level of mispricings is economically and statistically (significant) reduced by the system upgrade.

The effect of automated trading on market quality: Evidence from the New York stock exchange

From the end of 2006 until the beginning of 2007 the NYSE introduced the NYSE Hybrid Market on a rolling basis. The NYSE Hybrid Market significantly changed the NYSE’s market model and supports automated execution for almost unlimited order sizes and different order types. The introduction of the Hybrid Market was driven by fundamental changes in the securities trading industry over the last years. This paper analyzes the effect of the NYSE Hybrid Market on market quality through analyzing different spread measures and price impact. Results show that the introduction of the Hybrid Market reduced trading costs and improved execution quality at the NYSE.

System latency in linked spot and futures markets

We examine the lead-lag effect between DAX index and DAX index futures under asymmetric latency in the exchange infrastructure. Using 1-min high frequency observations in 2006-2007, it is found that the market integration between stock index and stock index futures has significantly grown compared to prior research. While the degree of price discovery in the futures market decreased both markets react mostly contemporaneously towards new information. An event story of latency reduction on Xetra reveals that exchange latency is one important factor explaining this development. We find evidence that smaller asymmetric round-trip-times between Xetra and Eurex lead to a higher degree of market integration.

Know the flow: Sentiment extraction from retail order flow data

Retail investor sentiment has been a subject of interest in the finance literature for a number of years. Using the order flow based on a unique data set with 20.7 million transactions in bank-issued warrants from the European Warrant Exchange, we present the construction of a retail investor sentiment index. We show that retail investors are contrarian, that retail investor sentiment is an important part of the equity pricing process and that we have a good measure of the sentiment. We believe that this information can be used by exchanges to increase transparency in financial markets. As a whole our findings further support a role for retail investor sentiment in the equity pricing process.

Collaborative continuous service engineering: A case study in a financial service environment

We present a methodology to engineer services in real-time information environments. We evaluate, combine and enrich traditional techniques and methodologies such as New Product Development and Adaptive Software Development in order to support the development and continuous improvement of information services. Our methodology is based upon the financial service environment. It is specific enough to be applicable in the field of finance and general enough to be applied broadly in information services development.