After finishing my first book, Trading Thalesians: What the ancient world can teach us about trading today, I said to myself that I would never write a book again. After a few months, I pretty much forgot saying anything of the sort, and tried to think of something else to write. The difficulty though was picking the right subject! Luckily, Alexander Denev approached me earlier this year, suggesting we co-author “The Book of Alternative Data” for finance. The book is due to be published by Wiley in early 2020 and should be available for pre-order on Amazon before the end of 2019.
Alexander has done a lot of work in the area of alternative in his previous roles, including at IHS Markit. On my side, I’ve done several research projects on alternative data, including one with Bloomberg, showing how their machine readable news can be used to trade FX. Hence, tackling the topic of alternative data made a lot of sense for us.
We started working on the book several months ago. We’ve written just under half the book, so I thought it would be worthwhile giving a sneak preview of what we’re aiming at. Admittedly, this article doesn’t go into every chapter, but hopefully it’ll give a nice overview.
So who is the book for? We’ve writing it to appeal to folks working in financial firms who have an interest in alternative data and are trying to understand how it could fit in with their process. This can for example include portfolio managers, traders, data strategists and so on. It also equally suitable for firms looking to understanding how they can monetise their data. Yes, there are certain sections which are more technical and likely to be more for data scientists and quants, but we think so far we have struck the right balance.
In the introduction, we seek to understand how data is produced and who consumes it. We try to define alternative data and discuss its various characteristics. We discuss the topic of “exhaust data”. We write about the challenges associated with alternative data, such as cleaning it and addressing missing data, as well as the importance of structuring these datasets. We also flag questions associated with the legal aspects of distributing data and also the very important question of trying to value data. The very important question of how to find alternative data is also discussed.
In later chapters, we drill down into a number of subjects like factor investing, giving concrete examples of how alternative data can be applied. A large part of the book drills down into the various types of alternative data out there at present, with various case studies for traders. These include satellite imagery, text based sources (web, social media and news), consumer transactions, alternative market data, crowdsourced data etc. We are still very much in the process of trying to find fun datasets to include in our book. Examples we’ve added so far, include showing how to use Twitter to help forecast payrolls and understanding how measures of investor attention can help trade VIX futures. Another example we’re currently working on is a case study looking at how car park data from satellite imagery can be used to help forecast earnings per share for retail companies. For quants, we go into some detail into natural language processing, which is a crucial part of structuring text data and we also give an overview of how images can be structured.
We still have a lot of the book left to go, but we’re on target to finish the book in a couple of months. If you’ve got any questions about the book, or think you can help in any way, let us know. Or if you’re a data company who’d be interested in me doing an in depth case study to help market your dataset let me know too!