High-Frequency Trading: A Practical Guide to Algorithmic Strategies and Trading Systems (Wiley Trading)

This book gives the reader a broad introduction to the controversial and highly-competitive world of high-frequency trading. It is written in language clear enough for non-technical readers to benefit while dipping sufficiently deep into information technology and trading mathematics to satisfy those seeking more detail on the methods and mechanics involved in HFT. On the other hand, based on comments by other reviewers, it does not appear to satisfy the demands of the top technologists in the field for a magnum opus on the transmutation of C# code into gold. The book is well organized and illustrated with numerous graphs, tables and formulas that facilitate comprehension of the subject matter. Speculators familiar with algorithmic trading may see direct application to their current strategies as implemented on available commercial trading platforms. Professional money managers get a glimpse behind the smokescreen that that shrouds HFT enabling development of risk-mitigation strategies suited to the evolving world of high-frequency trading. Regulatory personnel can familiarize with the terminology, technologies and strategies to ask better informed questions leading the way to effective policies that limit market manipulation while minimizing harmful unintended consequences. As a private speculator with experience programming and operating algorithmic trading systems on somewhat longer timeframes than microseconds, I find Irene Aldridge's "High-Frequency Trading: A Practical Guide to Algorithmic Strategies and Trading Systems" an informative and useful reference book on the subject. Whether or not the shortcomings noted by other reviewers are technically correct they do not detract from the overall value of the material towards a better understanding of the field of high-frequency trading. I rate this book five stars on the basis that it should be on every trader's bookshelf.

I am a computer engineer and have used this book to improve my trading software. The sections on minimizing market impact and exchange microstructure are invaluable when implementing large scale trading operations. Aldridge does a great job extracting ideas from advanced research papers and making them more digestible. This book is essential for any quant, even if it's just for quick-references. Many of the concepts in this book can also be applied to various asset classes (e.g. commodities, stocks, crypto, etc).

This book presents an excelent introduction for High-Frequency Trading. You will need to get expirience to understand the area, but this book will give you a good start point. Some books will introduce unnecessary mathematical details in an initial vision. This is not the case of this book. It presents the concepts in a good flow of ideas. I recommend.

Dear All High Frequency Enthusiasts, This book is an amazing book. There are very few books which impact our wisdom and this is one of them. The author has done an incredible job and she did her best to make sure that people with less mathematical and financial background can understand it without complication. I see from comments that Practitioners also benefitted from this book and to any one else who is just thinking to purchase this incredible book, please don't think just purchase it. I highly recommend it.

Great content. Even when I work in financial Services and I' ve Belén teaching Finance for 5 years, this book vives Great insights into HFT. Good for starters and product manager's in the industry trying to grasp this complicated new business core. Better if you have some básic programming skills or IT background

I think the content is concrete and useful.

I spend a lot of money on math and finance books, so I try to carefully judge whether these expensive books will be worth while purchases. One resource is Google books, which sometimes has excerpts when Amazon doesn't. This was the case with Irene Aldridge's book "High Frequency Trading, Second Edition". Unfortunately only the first 28 pages are available on Google books. But one section, on page 24 I found disturbing since it is wildly inaccurate. I have quoted this section below: "FPGAs are programmed using special programming languages, such as Verilog or VHDL. The languages are similar to C programming language [sic] and are easy to learn. A special FPGA programming device translates Verilog or VHDL into Assembly language understood by the FPGA chips. In the absence of FPGAs, trading programs need to be compiled and translated to the computer chips like CPUs during program run time, requiring additional computer operations eating into latency. The process of programing an FPGA is rather straightforward and inexpensive." Page 24 For a small section of text, this includes a remarkable number of errors. Verilog does resemble C, but VHDL most certainly does not. It resembles Ada more than anything else. When Verilog and VHDL are used to program FPGAs they are not translated into assembly language (the text form form of the CPU instruction set). These language are translated into logic "net lists", which describe the logic gates and interconnections for the logic design. The FPGA manufacturer usually provides a program that compiles the logic net list into the FPGA binary. Verilog and VHDL are referred to as hardware design languages. You use these languages not to express a software algorithm, but a hardware design. I have a discussion of this on my web site, which with a little searching you can find (Amazon will not allow URLs). Learning to design efficient hardware is not easy or straight forward. As Ms Aldridge notes a bit later in this text, electrical engineering students do learn to use the languages, but I suspect that most don't find them easy to learn. For a realization of an algorithm in an FPGA to be fast there must be parallelism. Not all algorithms have this parallelism, in which case they will run faster on a CPU. The only acknowledgment of the issue of parallelism and FPGAs is to "a format known as massively parallel processor array configuration", which in this case seems to be a usage without understanding. One might forgive Ms. Aldridge for not understanding FPGAs and modern hardware design. But the most egregious error in this short text is the description of compiled code relative to FPGAs: "In the absence of FPGAs, trading programs need to be compiled and translated to the computer chips like CPUs during program run time, requiring additional computer operations eating into latency". This shows an almost complete ignorance of software. Computer programs, written in languages like C++ are translated by a compiler into executable binary machine code and stored in one or more files. This translation is not done at run time, but when the program is compiled. It incurs no run-time overhead at all and has no effect on latency. High Frequency Trading is a technological arms race. HFT traders rent computer rack space that is physically close to the exchanges to minimize the lag caused by network cables. The issue of FPGAs comes up because FPGA based systems are used to reduce the overhead of the software "stack" that processes the network transactions that underlie market buy, sell and quote operations. There are groups that are building specialized long distance computer networks to connect the Chicago and New York exchanges. One group is using fiber optics and the other microwave relays. HFT relies heavily on "bleeding edge" technology, so it is disturbing when a book on HFT does not get the technology right. As I noted, I have only read the part of the book that is available on Google books. So it may be unfair to judge a book on this basis. However, I have rarely seen such a short section of text that contains so many errors. If the rest of the book is no more factual than this, then the book would be a poor investment.

An excellent book and the writer is a very knowledgeable and experienced

This book describes in a deep mode the High Frequency Trading Systems. Includes many references to others sources where you can find more specific information.

Imagine buying a $80 book (compare to US Federal min wage of $7.25/hour) about HFT and getting "insights" like "Return = (Price_today-Price_yesterday)/Price_yesterday" in return. I have read this book till the end hoping that it will get better. But no. Maximum superficial with little value. It's like another amazon user said "it's looks like an MBA thesis". That's definitely right.

Das Buch wirkt wie eine MBA-Diplomarbeit. Die Autorin referiert die Standardliteratur. Nachdem es zum Thema HFT noch relativ wenige akademische Arbeiten gibt, kommt HFT nur in einer Epsilon Umgebung von Null vor. Das Buch ist primär eine Einführung in die klassische Investitionstheorie. So gibt es ein eigenes Kapitel über alle jemals erfundenen Performance- und Risikokennzahlen. Das ist ganz nett gemacht (dafür gibt es auch den 2-ten Stern), nur hat es so gut wie nichts mit HFT zu tun. Dementsprechend sind auch die meisten Literaturangaben aus der ferneren Vergangenheit in der HFT noch gar nicht erfunden bzw. technisch nicht möglich war. Dort wo sich Aldridge auf technische Details einlässt, wird es peinlich. Für HFT werden FPGA-Karten als eine Art Coprozessor verwendet. Ich habe eine der ersten Anwendungen dieses Konzepts für das High-Performance-Computing entwickelt ([1],[2]) und kenn mich daher auf diesem Gebiet etwas aus. Es ist offensichtlich, dass Aldridge nicht die geringste Ahnung von FPGAs hat. Es wirken auch Sätze wie "Genetic algorithms learn from past forecasts via the so called Bayesian approach" sehr eigenartig. Beide Konzepte haben direkt nichts miteinander zu tun. Aldridge erklärt auch nicht wie das Lernen im Detail funktioniert. Wahrscheinlich wirft sie nur mit chicen Begriffen um sich. Im einzigen HFT spezifischen Kapitel geht die Autorin auf die Vorwürfe und Anschuldigungen gegen das HFT ein. Wobei sie auch hier sehr oberflächlich bleibt und im wesentlichen PR-Phrasen von sich gibt. HFT versorgt den Markt mit Liquidität und das kann ja nur gut sein. Meiner Meinung nach versteht niemand den Mechanismus des Finanzcasinos. Nach dem Standardmodell der Geometrischen Brownschen Bewegung müsste es ständig krachen, da eine Brownsche Bewegung nicht stationär ist. Im Grunde müsste man nach jeden Tag, an dem die Börse leidlich funktioniert hat, eine Untersuchungskommission einsetzen, die dieses Wunder untersucht. Ohne brauchbare Theorie sind daher die Pro- und Contra-Argumente ohne seriöse Grundlage. Nachdem die Börse ein Nullsummenspiel ist, werden die Gewinne der HFT-Trader aus den Taschen der untermotorisierten Teilnehmer bezahlt. Es agieren allerdings auch die anderen Spieler nicht zum Wohle der Realwirtschaft oder gar der Menschheit. Faktisch der gesamte Handel ist Spekulation. Es zocken auch am Pokerserver die Profis die Amateure ab. Wer über die böshen HFT-Onkelz jammert, soll sich halt den Chrilly engagieren. Der baut ihm schon eine Börsenhydra. Ähnlich dürftig wie das Buch ist die beworbene Webseite. Es fängt damit an, dass man nach einem Passwort auf Seite 320 gefragt wird. Nur gibt es in der 2.ten Auflage keine Seite 320. Es hat sich offensichtlich bisher niemand die Mühe gemacht die Webseite auf den neuesten Stand zu bringen. Die Autorin führt in den References ein paar eigene working-papers an. Es gehört wohl zum Minimum einer derartige Webseite, dass man diese Working-Papers herunterladen kann. Es findet sich aber nur ein Bruchteil davon am Server. Man kann aber davon ausgehen, dass die Entropie dieser papers ohnehin nahe am Thermodynamischen Gleichgewicht ist. [1] Ch. Donninger and U. Lorenz. The Hydra Project. XilinX Journal (selected paper), Issue 53, 2005 [2] Ch. Donninger, U. Lorenz. The Chess Monster Hydra. Proc. of 14th International Conference on Field-Programmable Logic and Applications (FPL), 2004, Antwerp – Belgium, LNCS 3203, pp. 927 – 932, eds. J. Becker, M. Platzner, S. Vernalde

The practitioners will probably find this book lacking depth but it provides a good overview of high frequency strategies/trading. I would recommend this book for people willing to learn about HFT but have no real experience in the field.