Editor’s Introduction to the Special Issue on Market Microstructure

INTRODUCTION

Market structure is the set of rules that govern order handling, trading, and price determination. Market microstructure, a relatively new field in financial economics, addresses issues regarding how orders are integrated together and turned into trades and transaction prices, and it involves the study, design, and regulation of trading mechanisms. Accordingly, market structure and market microstructure have much in common. In the 12 articles in this special issue on market structure we deal with various topics of interest to asset managers and regulators.

When microstructure analysis emerged in the second half of the 20th century, only a handful of people were working in the yet unnamed field. One of us, Schwartz, was part of that small group, and we were concerned that the field would run out of topics to deal with. We were wrong. Microstructure has attracted a substantial amount of academic research and, today, virtually half a century later, the field has remained robust. How come? The security markets are very complicated, the devil is in the details, technology and regulatory change have been transformative, institutional investing has moved to center stage, and a good deal remains that practitioners, regulators, and academicians do not yet understand.

Following the 1975 Securities Acts Amendments which mandated the development of a National Market System, there has been a robust growth of institutional investing. The impact of technological change has been transformative. In the first article in this special issue, “Equity Market Structure and the Persistence of Unsolved Problems: A Microstructure Perspective,” Robert A. Schwartz, James Ross, and Deniz Ozenbas first present a brief historical perspective on the evolution of equity market structure and then consider implications microstructure analysis has had for improving market structure. The authors describe the structural issues that plague US equity market structure every day, issues that are not caused by external forces like regime change or climate change. Instead, these issues are wholly internal to the design and operations of the equity markets’ systems. Although they may not be dramatic enough to trigger stock market circuit breakers or regulatory intervention, the costs to investors, issuers and a healthy, vibrant free market economy are chronic, persistent, and substantial. Despite the improvements in the equity markets, market impact costs and price discovery noise continue to accentuate short period returns variance. The authors present empirical evidence of how sizable that accentuation has been, focusing on the rules that govern how orders are integrated together and turned into trades and transaction prices.

As markets evolve, the successful implementation of investment strategies requires that traders and portfolio managers need to adapt their trading strategies, their models of transaction costs, and their processes. In “A Practitioner Perspective on Trading and the Implementation of Investment Strategies,” Joseph A. Cerniglia and Frank J. Fabozzi discuss the fundamentals of market structure and trading, along with how these impact the execution of investment strategies. They provide a practitioner perspective for bringing together this information to explore different models and processes that can be employed by traders and portfolio managers to achieve optimal execution of trading strategies. After providing a taxonomy of trading costs, the authors discuss how these costs, particularly market impact costs, are estimated and forecasted.

Tremendous progress in order processing speeds, reduced fees and costs, and new algorithmic tools have driven up market volumes. The resulting growth in trading turnover has been greater than projected yet several high-profile technology developments have failed, and legacy practices based on intermediated floor and telephone trading have proven resilient. Emerging technologies including blockchain and other Web 3.0 developments will disrupt incumbent exchanges and market services providers. These emerging technologies will be used to launch novel alternatives to today’s trading venues and post-trade services organizations. An examination of technology’s current and future influence on trading and the structure of markets is the subject of “Emerging Technologies and the Transformation of Exchange Trading Platforms” by Andy Novocin and Bruce Weber. The authors explain how emerging technologies will support new operating platforms that will threaten today’s exchanges and disrupt intermediaries’ business models. They expect new entrants to leverage three technologies to transform trading and, in their view, create new market structures and disrupt exchange providers and incumbent firms that offer data and post-trade services to market participants: (1) algorithmic trust technology, (2) formative artificial intelligence, and (3) decentralized autonomous organizations (DAOs) that could perform financial markets functions.

Development of securities markets has changed as the application of information technology has changed their operations, the ways to trade, market access, and the speed of trading. Today, market development focuses on market design which refers to and is guided by market microstructure with its market model schemes and classifications, market model related findings, and results on topics of market liquidity. The objects of market design involve the analysis, description, and specification of market models, trading services, and trading systems. Market design as an approach for market development is applied by market operators who as a business build and run trading systems. It is the topic of Martin Reck’s article “Market Design—A Practitioner’s Perspective.” He describes market design in the context of Xetra, a trading system operated by Deutsche Börse that is the leading trading venue for German equities and exchange-traded funds in Europe. Some key concepts and innovations of Xetra are described as example cases for market design and it is shown how the evolution of the hybrid market model contributed to the success of Xetra.

Appropriate liquidity monitoring and management tools and programs can help asset managers and asset allocators navigate a variety of liquidity events. The practical aspects of managing liquidity within funds and multi-asset portfolios are explained in “A Primer on Liquidity from an Asset Management and Asset Allocation Perspective” by Harshdeep Ahluwalia, Anatoly Shtekhman, Venky Venkatesh, and Yu Zhang. The authors complement existing research by providing a primer on the topic of liquidity from the unique perspectives of both asset managers and asset allocators. For asset managers, they provide an overview of the importance of managing liquidity and regulations surrounding it, a liquidity classification of broad assets and sub-asset categories, and liquidity measurement, monitoring, and management programs. For asset allocators, the importance of incorporating liquidity assumptions of illiquid assets into an asset allocation optimization approach and the selection of a rebalancing method framework and liquidity considerations are discussed. The authors rank most liquid to least liquid assets but emphasize that the level of liquidity can vary quite drastically within their segments of fixed income and equities, where specific events can cause liquidity shocks. The importance of a custom-built asset allocation framework accounting for unique asset characteristics is explained and how the optimal allocation to illiquid asset changes once their unique characteristics are accounted for, highlighting the importance of its allocation being aligned with the investor’s risk tolerance and liquidity requirements.

While dealers continue to play an important role in corporate bond trading, new technologies have created more competitive and transparent markets. In “Corporate Bond Trading: Finding the Customers’ Yachts,” Maureen O’Hara and Xing (Alex) Zhou provide a survey of some of the most important changes in the corporate bond market, drawing on the large body of academic research investigating corporate bond trading, and supplementing this with current data on the state of the market. The most apparent change is the development of electronic trading. However, the authors argue that both dealer and customer behavior are also experiencing fundamental shifts in four specific areas—execution costs, customer trading networks, dealer behavior and inter-dealer trading, and electronic trading. They also identify some areas where change is perhaps slower than expected, highlighting the reasons that corporate bond markets have unique risks that influence market structure. Despite the number of bond dealers being half of what they were over the past decade, resulting in an increase in market concentration, they report that bond market volume has soared while execution costs are approximately 70% lower than in the past. O’Hara and Zhou find that small bond trades now execute at approximately the same transaction costs as large block trades. While there has been improvement in trading in the corporate bond market, the authors note that they are not perfect for several reasons. First, the corporate bond market is susceptible to periodic illiquidity due to the market practice of match making (which now makes up over 30% of bond trading volume) as opposed to market making. Second, bond issues are more subject to information leakage, and those that are highly illiquid, still trade almost entirely with dealers. Third, there remains limited pre-trade transparency. On balance, however, the authors argue that corporate bond markets have improved.

Call auctions play a critical role in markets. They are one of the oldest trading mechanisms, and yet one of the newest because non-electronic call auctions died out in the pre-computer age but have made a reentrance as electronic venues in an electronic marketplace. Periodic call auction trading contrasts sharply with continuous market trading where a transaction is made any time a buy and sell order meet or cross in price. Call auctions batch orders together for execution in multilateral trades at specific points in time when the market for a stock is “called.” Electronic calls auctions were introduced by the Nasdaq and the NYSE in 2004 and 2007, respectively, and have combined their electronic call auctions and continuous trading facilities in hybrid market structures that use the calls to open and to close their 9:30–16:00 continuous intraday markets. Despite the critical importance of electronic call auctions, they need to be better understand. Deniz Ozenbas and Robert A. Schwartz in “The Return of the Call Auction” explain why the electronic call is an important innovation in market structure, and has unique benefits concerning liquidity provision, intra-day volatility control, and price discovery accuracy. Using data between 2010 and 2021 for the Nasdaq and NYSE electronic call auctions, the authors assess how the opening and closing call auction volumes compare to the daily continuous trading volume and investigate the factors that impact the use of call auctions. They find that electronic call auctions are attracting substantial order flow, particularly the closing call which accounts for more than 7% of total daily continuous trading volume.

To assess the capacity of a trading strategy, portfolio managers need to estimate trading costs. Incorporating trading costs and estimating trading strategies’ capacity are critical tasks for quantitative asset managers. Portfolio managers overestimating the amount of capital their trading strategy is able to withstand (i.e., their capacity) tend to grow their assets under management to levels where the strategy is no longer profitable. According to the findings of recent studies, market impact decays slowly through time. What is the impact of such slow decay on a trading strategy’s capacity? Hector Chan in “Market Impact Decay and Capacity” investigates this question. He proposes a numerical procedure to estimate the capacity of trading strategies and explains how the procedure offers a portfolio manager flexibility in incorporating any specification of market impact. The author uses this procedure to estimate capacity, taking into consideration the slow decay of market impact through time. That is, the fact that prices are still impacted many days after a given trade. More specifically, as trades tend to be more autocorrelated when capital devoted to a trading strategy increases, capacity is sensitive to assumptions on market impact decay. The author finds that the slow decay of market impact leads to trading strategy capacity estimates that are significantly lower than previous studies.

In “Endogenous Dynamics of Intraday Liquidity,” Mikolaj Bińkowski and Charles-Albert Lehalle investigate the endogenous information contained in four liquidity variables (traded volume, bid-ask spread, volatility and the volume at first limits of the order book) at a five-minute time scale on four equity markets (US, UK, Japan, and Hong Kong). Based on Granger causality, the authors measure the level of information for 300 stocks from the four equity markets for a period of over five years. From the viewpoint of these four liquidity variables, the authors find that natural memory of the endogenous dynamics of liquidity spans from half an hour to a few hours, depending on some characteristics of the considered stock. There are three other key findings that they report. First, although for smaller bid-ask spreads it is more difficult to predict the bid-ask spread, the easier they find it is to predict the volume at first limits using autoregressive (AR) models. Second, for most liquidity variables, the existence of a correlation between a stock’s market capitalization and the intensity of autoregession is independent of the geographical zone. Finally, the out-of-sample coefficient of determination of AR models on intraday volatility is largest for the US, followed by Europe and then Asia. A similar geographical ranking results when vector autoregressive models are used in lieu of AR models to predict traded value every five minutes.

Recent academic studies claim that price impacts associated with index rebalancing can be large and may represent a “hidden cost” to fund investors. Because virtually all stocks are included in some index, the approach used in these studies to examine the impact of additions and deletions to an index needs to be modified. In “Demystifying Index Rebalancing: An Analysis of the Costs of Liquidity Provision” Ananth Madhavan, Jason Ribando, and Nogie Udevbulu discuss the practical realities of index rebalancing using a large database of 18,871 rebalances related data from 2016 to 2022. Because the indexes analyzed, the Russell indexes, are major benchmarks for index managers, the price effects around these rebalances are of considerable interest. The key difference of the authors’ approach from previous studies is that they estimate net flows for each stock aggregated across all Russell indexes rather than the gross flows into each index. This is crucial since stocks migrating from one index to another may see positive or negative flows depending on the assets tracking those indexes. The authors report many instances where the costs of liquidity provision, as measured by the beta-adjusted temporary impact, are negative, a finding that is consistent with a highly competitive market for liquidity provision.

In “Mean-Variance Optimization for Simulation of Order Flow,” Petter N. Kolm and Nicholas Westray propose an order flow simulator for meta orders such as those originating from the trading activity of buy-side firms. The proposed simulator has three design goals: (1) it should be simple to use and integrate into different applications, (2) it must be computationally efficient so that it can deal with a large number of securities, and (3) the simulated order flow should possess statistical properties similar to those of order flow observed in the market. Kolm and Westray provide two empirical illustrations. The first illustration demonstrates some of the statistical properties of the order flow simulator. The second illustration demonstrates how the proposed simulator can be applied to estimate the number of trades required to determine whether order flow from a buy-side firm has alpha.

Social networks’ impacts on capital markets have become more important today because large-scale online social networks allow for easy communication among individuals as social interactions disseminate important news releases into prices. Traditional models of financial markets have not modeled the effect of social networks, despite their crucial importance. Although social networks improve a firm’s access to institutional capital and increase the firm’s valuation and its stock liquidity, one problem with social networks is that they can generate divergent beliefs and result in persistent excessive trading. Furthermore, social interactions can magnify investors’ behavioral biases, and can influence investors’ attraction to lottery-type stocks. In “Social Networks, Trading, and Liquidity,” Lin Peng, Qiguang Wang, and Dexin Zhou review the latest research that uses large scale, representative, real-world social network data to study social networks’ influences on trading, liquidity, and valuations of stocks. The authors provide examples to illustrate why the roles of social networks are of particular importance to market participants.

Frank J. Fabozzi and Robert A. Schwartz

Editors