Book Reviews

Kevin J. Davey – Building Winning Algorithmic Trading Systems – Review

Kevin J. Davey - Building Winning Algorithmic Trading Systems - ReviewKevin Davey is a professional systematic trader with over 25 years of experience. He has achieved triple-digit annual returns in three World Cup of Futures Trading Championships. He won the competition in 2006, and got 2nd place in 2005 and 2007. I have listened to him speak on many podcasts. Kevin is a respected trading system developer in the industry. I came upon his book online and decided to add it to my reading list. Like the title says, the book is about algorithmic trading. It was published in 2014.

Getting started in trading

Davey starts the book with some personal life and disastrous losses in family. He then goes back to early ’90s and describes all his newbie mistakes in trading. Like many beginning traders, he was naive that success would be easy and some methods presented in a book would work well. He had around 10 years of miserable trading experience when he learned that adding to losers was bad, making decisions on gut feel was bad, automatic trading systems were easy to get curve-fitted and high win rate with small wins was not a good trading strategy. In 2004, Kevin was onto something. His first working strategy was a trend-following breakout system with volatility risk mechanism.

World Cup of Futures Trading Championships

The author describes the trading system he used for three years at the World Cup of Futures Trading Championships illustrated with equity curves of each contest. Though the end result was over 100% annual return each time, it didn’t come without drawdowns. There were 40-50% drawdowns and Kevin admits that trading at a competition is somewhat different to trading one’s regular account. The risk taking part is definitely different. He stresses the importance of discipline and following the process.

Going full-time

Next, Kevin shares what he did right and wrong at the start of pursuing trading full-time in 2008. I think one of the most asked questions by new traders is how much capital a full-time trader needs. While this obviously depends on many factors and is different for individuals, Kevin admits that starting with a low six-figures account was optimistic, though he had separate savings for 3-5 years of living expenses which feels more than enough to me.

Testing trading systems

Trading systems can be tested using four ways: historical backtesting, out-of-sample testing, walk-forward and real-time testing. The author describes each of these methods. Some of Kevin’s points I noted about evaluating a trading system:
– At least 5-10 years of data and test 30-100 trades per each trading rule.
– Slippage and commission can change the end result significantly.
– Monte Carlo simulation works if trades are not dependent on each other.
– If testing a trading idea gives bad results, try doing (backtest) the opposite.

Kevin goes in details how he plans and develops a trading system. He writes about the necessary parts of a trading system: entry and exit rules, market selection, time frames, programming and data considerations. The author comments on a data example that crude oil prices had never been negative (2014), well now by 2021 they actually have; markets do change all the time.

His steps on how to test an entry with random exits, an exit with random entries, then entries and exits with random ones are useful to think about. Kevin calls it the monkey test where he compares these parts of his edge to random occurrences to see if at least 70% of the random iterations are worse than his baseline outcome. Such process is being run regularly to see if the trading system still has its edge. Once preliminary check is done, the strategy will be put to in-depth testing and walk-forward analysis. Here the goal is not to tweak and optimize too much to keep the system robust. If the strategy passes this step, it goes to Monte Carlo simulation to see a thousand variations how the trades could have played out in different distributions. The goal is not to have risk of ruin that could wipe a trader out. Kevin wants to see the annual return to maximum drawdown ratio at least 2.

More topics in the book

There is more in this book about diversification, position sizing, taking a strategy live and monitoring it. Davey presents two intraday euro strategies that he is trading with real money at the time of writing. He walks readers through developing a trading system based on these real life examples. Kevin’s approach to position sizing is somewhat opposite to what I do myself. He first creates a futures trading system around one contract and then calculates how much money the account needs to have so that the max drawdown in dollars would make sense in percentage-wise. In that way he has many small accounts to trade separate systems. He takes more risk (maxDD 25% and more) to catch over 50% annual returns. In this book example, he sets a target and develops the entry rule in an empirical way rather than theoretical. At the end of the book Davey also provides TradeStation code for these two trading systems.

Summing it up

Overall, it’s an easy to follow book about systematic trading, where Davey walks the reader through a step by step process for creating a trading system. Kevin’s approach seems risky to me with short-term trading systems having only one year perspective. His risk of ruin in the examples provided is up to 10% that the account equity falls over 50% and he stops trading it. Like mentioned, he uses a separate small account for each system.

For someone who wants to actively develop and retire trading systems based on market conditions, this is a useful book about algorithmic trading. Though Kevin Davey’s style is more return with more risk (drawdowns), I think traders with all kinds of systematic goals will find useful points in this book. I recommend this book to people who are struggling to develop a process for creating their own algorithmic trading systems.

Building Winning Algorithmic Trading Systems by Kevin J. Davey book link

Share this post

Leave a Reply

Your email address will not be published. Required fields are marked *

This website uses cookies and collects visitors' data. By continuing to use the site, you agree to Privacy Policy

The cookie settings on this website are set to "allow cookies" to give you the best browsing experience possible. If you continue to use this website without changing your cookie settings or you click "Accept" below then you are consenting to this.