The Science of Algorithmic Trading and Portfolio Management

Robert Kissell

• Publisher: Academic Press
• Year: 2013
• eBook ISBN:
  9780124016934
• Topic: Negocios y empresa
• Subtopic: Inversiones y valores

The Science of Algorithmic Trading and Portfolio Management, with its emphasis on algorithmic trading processes and current trading models, sits apart from others of its kind. Robert Kissell, the first author to discuss algorithmic trading across the various asset classes, provides key insights into ways to develop, test, and build trading algorithms. Readers learn how to evaluate market impact models and assess performance across algorithms, traders, and brokers, and acquire the knowledge to implement electronic trading systems. This valuable book summarizes market structure, the formation of prices, and how different participants interact with one another, including bluffing, speculating, and gambling. Readers learn the underlying details and mathematics of customized trading algorithms, as well as advanced modeling techniques to improve profitability through algorithmic trading and appropriate risk management techniques. Portfolio management topics, including quant factors and black box models, are discussed, and an accompanying website includes examples, data sets supplementing exercises in the book, and large projects. - Prepares readers to evaluate market impact models and assess performance across algorithms, traders, and brokers. - Helps readers design systems to manage algorithmic risk and dark pool uncertainty. - Summarizes an algorithmic decision making framework to ensure consistency between investment objectives and trading objectives.

Chapter 1

Algorithmic Trading

This chapter introduces readers to algorithmic trading. We provide a description of the electronic trading environment and discuss issues required to make proper algorithmic trading decisions. We present and critique the major theories of algorithmic trading, and provide further insight into where change may continue to expand. We describe the current state of trading algorithms (both single stock and portfolio algorithms) and provide a classification system to assist investors and buy-side traders navigate the ever-changing algorithmic landscape. The chapter ends with a discussion of the recent market changes that have been accompanied with algorithmic trading.

Keywords

algorithms; High Frequency Trading (HFT); Volume Weighted Average Price (VWAP); dark pools; grey pools; crossing networks; Direct Market Access (DMA); Auto Market Making (AMM); smart order routers; limit order models; implementation shortfall; arrival price; Transaction Cost Analysis (TCA)

Introduction

Algorithmic trading represents the computerized executions of financial instruments. Algorithms trade stocks, bonds, currencies, and a plethora of financial derivatives. Algorithms are also fundamental to investment strategies and trading goals. The new era of trading provides investors with more efficient executions while lowering transaction costs—the result, improved portfolio performance. Algorithmic trading has been referred to as “automated,” “black box” and “robo” trading.

Trading via algorithms requires investors to first specify their investing and/or trading goals in terms of mathematical instructions. Dependent upon investors’ needs, customized instructions range from simple to highly sophisticated. After instructions are specified, computers implement those trades following the prescribed instructions.

Managers use algorithms in a variety of ways. Money management funds—mutual and index funds, pension plans, quantitative funds and even hedge funds—use algorithms to implement investment decisions. In these cases, money managers use different stock selection and portfolio construction techniques to determine their preferred holdings, and then employ algorithms to implement those decisions. Algorithms determine the best way to slice orders and trade over time. They determine appropriate price, time, and quantity of shares (size) to enter the market. Often, these algorithms make decisions independent of any human interaction.

Similar to a more antiquated, manual market-making approach, broker dealers and market makers now use automated algorithms to provide liquidity to the marketplace. As such, these parties are able to make markets in a broader spectrum of securities electronically rather than manually, cutting costs of hiring additional traders.

Aside from improving liquidity to the marketplace, broker dealers are using algorithms to transact for investor clients. Once investment decisions are made, buy-side trading desks pass orders to their brokers for execution using algorithms. The buy-side may specify which broker algorithms to use to trade single or basket orders, or rely on the expertise of sell-side brokers to select the proper algorithms and algorithmic parameters. It is important for the sell-side to precisely communicate to the buy-side expectations regarding expected transaction costs (usually via pre-trade analysis) and potential issues that may arise during trading. The buy-side will need to ensure these implementation goals are consistent with the fund’s investment objectives. Furthermore, it is crucial for the buy-side to determine future implementation decisions (usually via post-trade analysis) to continuously evaluate broker performance and algorithms under various scenarios.

Quantitative, statistical arbitrage traders, sophisticated hedge funds, and the newly emerged class of investors known as high frequency traders will also program buying/selling rules directly into the trading algorithm. The program rules allows algorithms to determine instruments and how they should be bought and sold. These types of algorithms are referred to as “blackbox” or “profit and loss” algorithms.

For years, financial research has focused on the investment side of a business. Funds have invested copious dollars and research hours on the quest for superior investment opportunities and risk management techniques, with very little research on the implementation side. However, over the last decade, much of this initiative has shifted towards capturing hidden value during implementation. Treynor (1981), Perold (1988), Berkowitz, Logue, and Noser (1988), Wagner (1990), and Edwards and Wagner (1993) were among the first to report the quantity of alpha lost during implementation of the investment idea due to transaction costs. More recently, Bertsimas and Lo (1996), Almgren and Chriss (1999, 2000), Kissell, Glantz, and Malamut (2004) introduced a framework to minimize market impact and transaction costs, as well as a process to determine appropriate optimal execution strategies. These efforts have helped provide efficient implementation—the process known as algorithmic trading¹.

While empirical evidence has shown that when properly specified, algorithms result in lower transaction costs, the process necessitates investors be more proactive during implementation than they were previously utilizing manual execution. Algorithms must be able to manage price, size, and timing of the trades, while continuously reacting to market condition changes.

Advantages

Algorithmic trading provides investors with many benefits such as:

Lower Commissions. Commissions are usually lower than traditional commission fees since algorithmic trading only provides investors with execution and execution-related services (such as risk management and order management). Algorithmic commissions typically do not compensate brokers for research activities, although some funds pay a higher rate for research access.

Anonymity. Orders are entered into the system and traded automatically by the computer across all execution venues. The buy-side trader either manages the order from within his firm or requests that the order is managed by the sell-side traders. Orders are not shopped or across trading floor as they once were.

Control. Buy-side traders have full control over orders. Traders determine the venues (displayed/dark), order submission rules such as market/limit prices, share quantities, wait and refresh times, as well as when to accelerate or decelerate trading based on the investment objective of the fund and actual market conditions. Traders can cancel the order or modify the trading instructions almost instantaneously.

Minimum Information Leakage. Information leakage is minimized since the broker does not receive any information about the order or trading intentions of the investor. The buy-side trader is able to specify their trading instructions and investment needs simply by the selection of the algorithm and specifications of the algorithmic parameters.

Transparency. Investors are provided with a higher degree of transparency surrounding how the order will be executed. Since the underlying execution rules for each algorithm are provided to investors in advance, investors will know exactly how the algorithm will execute shares in the market, as algorithms will do exactly what they are programmed to do.

Access. Algorithms are able to provide fast and efficient access to the different markets and dark pool. They also provide co-location, low latency connections, which provides investors with the benefits of high speed connections.

Competition. The evolution of algorithmic trading has seen competition from various market participants such as independent vendors, order management and execution management software firms, exchanges, third party providers, and in-house development teams in addition to the traditional sell-side broker dealers. Investors have received the benefits of this increased competition in the form of better execution services and lower costs. Given the ease and flexibility of choosing and switching between providers, investors are not locked into any one selection. In turn, algo providers are required to be more proactive in continually improving their offerings and efficiencies.

Reduced Transaction Costs. Computers are better equipped and faster to react to changing market conditions and unplanned events. They are better capable to ensure consistency between the investment decision and trading instructions, which results in decreased market impact cost, less timing risk, and a higher percentage of completed orders (lower opportunity cost).

Disadvantages

Algorithmic trading has been around only since the early 2000s and it is still evolving at an amazing rate. Unfortunately, algorithms are not the be all and end all for our trading needs. Deficiencies and limitations include:

Users can become complacent and use the same algorithms regardless of the order characteristics and market conditions simply because they are familiar with the algorithm.

Users need to continuously test and evaluate algorithms to ensure they are using the algorithms properly and that the algorithms are doing what they are advertised to do. Users need to measure and monitor performance across brokers, algorithms and market conditions to understand what algorithms are most appropriate given the type of market environment.

Algorithms perform exactly as they are specified, which is nice when the trading environment is what has been expected. However, in the case that unplanned events occur, the algorithm may not be properly trained or programmed for that particular market, which may lead to sub-par performance and higher costs.

Users need to ensure consistency across the algorithm and their investment needs. Ensuring consistency is becoming increasingly difficult in times where the actual algorithmic trading rule is not as transparent as it could be or when the algorithms are given non-descriptive names that do not provide any insight into what they are trying to do.

Too many algos and too many names. VWAP, volume weighted average price, is an example of a fairly descriptive algorithmic name and is fairly consistent across brokers. However, an algorithm such as Tarzan is not descriptive and does not provide insights into how it will trade during the day. Investors may need to understand and differentiate between hundreds of algorithms, and keep track of the changes that occur in these codebases. For example, a large institution may use twenty different brokers with five to ten different algorithms each, and with at least half of those names being non-descriptive.

Price Discovery. As we discuss in Chapter 2 (Market Microstructure) the growth of algorithms and decline of traditional specialists and market marker roles has led to a more difficult price discovery process at the open. While algorithms are well versed at incorporating price information to determine the proper slicing strategy, they are not yet well versed at quickly determining the fair market price for a security.

Changing Trading Environment

The US equity markets have experienced sweeping changes in market microstructure, rapid growth in program trading, and a large shift to electronic trading. In 2001, both the New York Stock Exchange (NYSE) and NASDAQ moved to a system of quoting stocks in decimals (e.g., cents per share) from a system of quoting stocks in fractions (e.g., 1/16th of a dollar or “teenies”). As a consequence, the minimum quote increment reduced from $0.0625/share to $0.01/share. While this provides investors with a much larger array of potential market prices and transactions closer to true intrinsic values, it has also been criticized for interfering with the main role of financial markets, namely, liquidity and price discovery.

The decrease in liquidity shortly after decimalization has been documented by Bacidore, Battalio, and Jennings (2001), Nasdaq Economic Research (2001), and Beesembinder (2003). This was also observed in the US equity markets...

Table of contents

• Cover image
• Title page
• Table of Contents
• Copyright
• Dedication
• Preface
• Acknowledgments
• Chapter 1. Algorithmic Trading
• Chapter 2. Market Microstructure
• Chapter 3. Algorithmic Transaction Cost Analysis
• Chapter 4. Market Impact Models
• Chapter 5. Estimating I-Star Model Parameters
• Chapter 6. Price Volatility
• Chapter 7. Advanced Algorithmic Forecasting Techniques
• Chapter 8. Algorithmic Decision Making Framework
• Chapter 9. Portfolio Algorithms
• Chapter 10. Portfolio Construction
• Chapter 11. Quantitative Portfolio Management Techniques
• Chapter 12. Cost Index & Multi-Asset Trading Costs
• Chapter 13. High Frequency Trading and Black Box Models
• References
• Index