Listen to the whole quant album

20181201 Sunset

It’s 50 years since the Beatles released the White Album. To celebrate it has been reissued with new mixes of the original tracks. I have to admit, it’s rarer these days to listen to an album the whole way through. Instead, I invariably listen to playlists on iTunes which select tracks from all manner of different artists and albums and mix them together. iTunes and Spotify have revolutionized how we listen to music, making it so find a track by nearly any artist you can think of. Whilst, I appreciate the convenience of listening to music in this way, we can sometimes miss the point.


An album, such as the White Album, is designed to take you on a journey, around a certain theme. In a sense, by only sampling a track here or there, we are basically changing that story. It’s like picking up a novel, but randomly reading chapters from it. We evidently lose something in the process. Whilst albums might be pockmarked by filler tracks, which would never make it as a single, there are often also somewhat less commercial tracks, which you’d never discover, as they’d rarely be one of the playlist tracks on iTunes or on a radio station.


I often think that people try to delve into the quant trading and investment in the same way. At first, it can be easy to be seduced by the “top selling singles” the themes which would invariably make it to an iTunes playlist (yes that top selling single called “AI and machine learning”). However, this I don’t think is the best way of doing it. Instead we need to think about quant as a whole, and listen to whole album, to see what it can offer. Here’s my checklist of how to move your investment process towards a more quant based approach.


Ask someone for help at the start of your quant journey (maybe like Cuemacro!)

There are many paths you can choose when starting to get into quant investing. It is easy to waste time getting to the wrong past, spending time developing systems and processes which aren’t helpful for your specific investment process, which needs to be customised for you. It can save time and money getting help at the beginning to avoid these costly mistakes and to help brainstorm ideas with you from a totally independent viewpoint.


It’s not just about machine learning: think about automating spreadsheets too!

I often get asked about machine learning. It’s a great tool, especially when it comes to trying to understand text data. I’ve spent a lot of time doing projects for firms such as Bloomberg examining how machine readable news can be useful for the investment process. However, it is important to understand that quant investing is not purely about machine learning, it is one facet. Creating basic trend or carry models can for example be a good first step along the way to quant investing. Also thinking about how to automate processes is key. I worked in banks for many years, and I can’t begin to think about how much time I spent updating spreadsheets to generate market indices for traders or researchers. These days I try to make sure as much as my research process is automated from downloading data to backtesting, which has helped free up a lot of my time.


Different quant techniques for different investors

I’ve developed models for trading macro assets for many time frequencies from intraday to daily to monthly. For fundamental investors, having access to high frequency trading algorithms might be useful for execution, however, in the grand scheme of things, their investment decisions tend to be longer term. Having an understanding of how the economy is performing and forecasting over months is what really counts. The types of quant tools and analytics needed are somewhat different compared to a high frequency firm.


Quant isn’t┬ájust about systematic trading models

There’s a misconception that quant is only about systematic trading strategies, which tell you to buy and sell. Of course this is what a quant fund does, but we also have many firms with a semi-systematic approach, who use quant tools to help come up with views and validate arguments, but have portfolio managers making the final trading decision. Even for quant funds, I would argue that to a certain extent the whole process is about making a systematic trading model, involves elements of discretion which are made into code and requires a strong understanding of how the market works.



Quant tools can be great for investors starting out in the discretionary world. However, to use them effectively we need to understand that their requirements will differ from for example high frequency traders. We need to listen the whole quant album, and not purely the machine learning track! Ultimately, quant needs to be customised for fundamental investors, and that’s indeed something that Cuemacro can help with. We can help guide you along the right path.