- Series: Wiley Trading
- Hardcover: 264 pages
- Publisher: Wiley; 1 edition (February 6, 2017)
- Language: English
- ISBN-10: 1119219604
- ISBN-13: 978-1119219606
- Product Dimensions: 6 x 1.2 x 9.1 inches
- Shipping Weight: 1 pounds (View shipping rates and policies)
- Average Customer Review: 2.8 out of 5 stars See all reviews (7 customer reviews)
- Amazon Best Sellers Rank: #81,109 in Books (See Top 100 in Books)
Enter your mobile number or email address below and we'll send you a link to download the free Kindle App. Then you can start reading Kindle books on your smartphone, tablet, or computer - no Kindle device required.
To get the free app, enter your mobile phone number.
Machine Trading: Deploying Computer Algorithms to Conquer the Markets (Wiley Trading) 1st Edition
Use the Amazon App to scan ISBNs and compare prices.
Prepare for your professional certification with study guides and exam prep tools from Wiley. See more
Frequently bought together
Customers who bought this item also bought
From the Inside Flap
Following up on his widely popular algorithmic trading guides, Quantitative Trading and Algorithmic Trading, this third installment is written for quant traders and investors ready for more advanced examinations and techniques. Machine Trading is your accessible companion for the state-of-the-art of algo-trading in today's complex markets.
Don't worry if you lack trading and finance experienceif you've worked in a quantitative field, such as computer science, engineering, or physics, this step-by-step resource makes the transition into algorithmic trading seamless. It starts out with a comprehensive look at the latest backtesting and trading platforms, the best- rated and most cost-effective vendors' data, and the easiest way to optimize allocations in different assets and strategies. Acquire a firm grasp on options and volatility strategies; factor models, and why they can be useful to short-term traders; and the intricacies of intraday and high-frequency trading, including market microstructure, dark pools, order flow, and backtesting intraday strategies with tick data. There are no canned solutions insideeach prototype trading strategy provides a rock solid foundation for you to customize. Hone your skillset on topics such as:
- Using factor models for long-term returns and short-term trades,including using option prices as factors
- Real-world trading with time series techniques, including ARIMA, VAR, and State Space Models with hidden variables
- Cutting-edge techniques to reduce outfitting in artificial intelligence and machine learning strategies
Every chapter includes hands-on exercises walking you through the critical modifications to make on your own to gain control of the strategies and discover their potential. From stocks to futures and options, foreign exchange, and bitcoins, Machine Trading is your one-stop training ground for finding algo-trading solutions.
From the Back Cover
Praise for Machine Trading
"It is easy to make simple ideas complex. It is far more difficult to make complex ideas seem simple. In this book, Ernie has done exactly that. I cannot think of any trader who would not benefit from reading Machine Trading."
Euan Sinclair, Partner, Talton Capital Management; author, Volatility Trading
"As with his first two books, Ernie delves into the practical matters of automated trading in Machine Trading. He carefully explains everything from where to find quality data, to which platforms to consider, to timely and topical strategies. In the years since his last book, automation of the investment industry has accelerated, due in large part to the recent Cambrian Explosion in financial technology (FinTech). Ernie has long been teaching the hard-to-find methods of automated trading; now he is also a guide to the dizzying array of new companies, technologies, and services for automated trading. Machine Trading takes you deeper into the field with new concepts, while still delivering Ernie's signature explanation of the concrete steps of pursuing your own passion for automated trading."
John Fawcett, CEO and Founder, Quantopian
"Dr. Chan has written another accessible and information-packed book for the quant-minded trader. The book starts with a clear discussion of factor models, advanced time series analysis, and Kalman filters, all of which lead into a detailed description of machine learning and artificial intelligence techniques applied to volatility trading, market microstructure, and even Bitcoins. His focus on finance in a MATLAB context is refreshing and opens the algorithmic trading domain to a whole generation of engineers and quantitative practitioners not familiar with finance but involved in numerate fields like self-driving cars and proteomics. I highly recommend it."
Dr. Taha Jaffer, President, OXOBOXO Inc.; formerly Principal at The Carlyle Group
Browse award-winning titles. See more
If you are a seller for this product, would you like to suggest updates through seller support?
Top Customer Reviews
A few more specific comments--
If you plan on replicating any of the scripts, be prepared to purchase Matlab and several libraries, if you do not already have a license. The author says on many occasions that Matlab is better and faster than R. I don't really agree, but nevertheless, the additional investment might be worthwhile if the code was clearly presented and edited. Unfortunately, much of the code was not even runnable in its current state. There were referenced functions with no comments on how to source them (I later found many in the utilities folder). There were several data files that were sourced and not included in the site link. There are files with intraday 1 min data on the order of 26 million sample points, that take a long time to run, assuming they don't suffer from other issues. Having a lot of coding experience, I felt that I was troubleshooting much of the code and trying to get around many of the idiosyncrasies in order to get scripts to run. I don't think it is really oriented towards complete novices as advertised. There were a few interesting code illustrations, if you wanted a starting point to investigate ideas using Matlab, but I felt that they were hastily thrown together and unorganized, without much thought towards verifying them.
Much of the concepts were unorganized and inconsistent. Some examples had in sample results, others, in sample and out of sample for a few years, without really clearly illustrating the importance of using several folds for validation over long periods. Yes, there were some explanations of cross-validation, but it would have been better illustrated towards applying it in the code examples with many more years of data (even if the data was not available for free, a nice table of results would have been very useful for each of the experiments).
Aside from the code, I didn't really get the sense that the ideas presented were able to really simplify the more complex concepts much. I felt that if you already had background in much of the subject matter (e.g. time series, financial engineering, machine learning), some of the ideas were accessible, by having prior exposure to them. But I think there are better, clearer texts, that help the novice understand many of the concepts, and do not really see a novice picking up those concepts here.
All that being said, I do get the sense that the author understood a lot of the concepts that he discussed, I just feel that he could have presented the material in a much clearer and organized manner, and put more effort into editing prior to release. Again, I felt that lots of the ideas explored were interesting and ambitious, but would prefer to see less concepts explained, with more exposition and clarity. Also, the code could use a lot of editing and clear explanations of where data is and is not available, along with where it can be accessed for specific script examples.
The book did seem like more of a collection of thinking out loud and collecting and sharing of interesting research ( a blog extension ), than an introductory text on machine trading, targeted towards novices. Also, even though I think the code could use a lot of work, you'd be hard pressed to find similar concepts and code shared in other texts. Don't expect to find the holy grail within, but if you already have some experience in the topics I mentioned, it might give you some quantitative ideas to explore.
The next few chapters look at analysis using MATLAB as the programming language. The code snippets are well described and fairly easy to understand or translate to another language.
Another chapter which I enjoyed was the one on Artificial Intelligence. Actual techniques were shown in a simple manner and the techniques progressed from simple to more complicated. A good point was made how you want to try the simplest machine learning techniques first, before trying more advanced one. It was shown how neural networks generally do not perform that well on financial data. There was an example given showing that a neural network with only layer performed better than networks with too many layers.
Later there is a detailed section about how orders work in the stock market, and how closing prices are often "consolidated". The consolidated closing (and opening) prices can cause trouble for back-testing strategies. Explanations were clear about potential pitfalls of using different types of market or limit orders.
Another surprise was a chapter on bitcoin which had possible trading strategies on the crypto-currency. One of the advantages of bitcoin for analysis is that the order book is open and each trade specifies whether it was a buy or sell.
The book wraps up with a discussion about maintaining a career in trading. I was impressed with the content, the explanations, and details that went into this book.
I would recommend having some basic knowledge of three areas: linear algebra, a scripting language (MATLAB, Python), and how the stock market works. Text books are written on all of those topics and Chan did an excellent job of staying on point without going on tangents to explain underlying principles, but still using principles that are simple enough to understand. I found the MATLAB Machine Learning e-book a great complement to the Artificial Intelligence chapter for understanding the goal of each algorithm. (Found on the mathworks website.)
He does not, however, assume knowledge of algorithmic trading, so if you are just starting out, this is a great resource. The first chapter lays out where to find data, how to test and verify algorithms, and different metrics for evaluation. The rest of the book is well organized and advances in complexity from time series analysis to AI technology to the latest high frequency (< 1 ms) methods. The book is rich with examples of how to implement each method and Chan even provides a link to the MATLAB code for readers who would like to see more.
Machine Trading cites Chan’s first two books frequently, but this can be read on its own. That said, this was so good I immediately bought his other two books!
Like I said, this was my first book on algorithmic trading, and I think it was exceptional because of its clarity, but also something I will be returning to frequently for examples and more advanced ideas. Highly recommend for anyone interested in the topic.