Tap, tap, tap. It’s impossible not to tap your feet to a good track. Music can come in all forms, from Beethoven and Mozart, to Jay Z and Taylor Swift. There’s the good, the bad, the ugly: music which uplifts, depresses, infuriates. Music is about triggering a reaction in the listener. It’s about hearing that humming that riff. It’s about waiting for the chorus. It’s about pausing during the silences. In this age of Spotify and iTunes, machine learning algorithms seek to find new music we’d like, which infuriatingly, is similar to what we’ve heard before, or on. That’s not really the point, is it? The whole point is to find new music, you’d never have even thought of before, rather than hearing some variation on a theme. Finding new music at random is what’s most rewarding: stumbling upon a documentary about 1970s music, hearing something you’ve never heard in the supermarket, flicking across radio channels till you find that infectious tune.
The key point I suppose is that sometimes, learning is not some linear exercise, and instead it’s the random gems we hear that can get you to think. In finance, reading books is important, but it’s also key to meet folks and hear what they have to say about markets, using Twitter to keep track of what’s happening and so on. Whilst some areas of finance are relatively mature, others areas are still very much in development. One of these newer areas is alternative data, something which I’ve been involved with for a number of years. What’s a good way to learn about alternative data?
Well, getting my plug out of the way, Alex Denev and I are coauthoring The Book of Alternative Data, which will be published on Wiley (pre-order on Amazon here!). We are now well over half way through the writing process, and we’ll be tackling a number of areas, including defining what alternative is, thinking of ways to value it, as well as tacking techniques for structuring it. Later, we’ll also be going through many investing use cases for alternative data ranging from satellite imagery to location data to news data. I’ll be teaching a course based on the book in New York on July 25-26 (get tickets here). I hope to be teaching my alt data course on an in-house basis too at interested firms (drop me a message if your firm would be interested in that).
Ok.. plug is now out of the way! What else? There are a number of conferences which I’ve found pretty good for getting an overview about alternative data. I’ve attended Neudata’s alt data conferences in New York and London, and found them a great opportunity to network and hear the latest about the alt data market. There are also other alt data conferences, which I’m hoping to attend in the future, organised by Quandl, BattleFin and Eagle Alpha, which have a similar focus. They bring together alt data vendors and clients. Bloomberg also host events on alternative data for buy side. Refinitiv have recently hosted a few webinars on alternative data (I presented on one of their recent webinars about natural language processing). Quandl also have a weekly newsletter, which aggregated news and views in this space (I find it super useful!). Alternativedata.org is also a decent resource, and I’d also recommend subscribing to their newsletter. My firm, Cuemacro does a monthly newsletter, which covers a number of quant topics, including alternative data (drop me a mail if you’d like to subscribe).
Another great way to learn about alternative data, is through Twitter. Obviously, it can be pretty difficult to keep track of every account which tweets about alternative data, but here are a few, which I follow to get you started including @robpas, @jbaksht and @RobinWigg (and if you have any specific recommendations let me know!). Also obviously, it’s worth thinking about following the various Twitter accounts of many alt data vendors to get an idea of what datasets they have.
Once you’ve got a good idea about the alternative data landscape, it’s then a good time to start getting access to some alternative data and seeing what signals you can generate. After all, the only way to truly understand a dataset is to spend time analysing it. Of course, there are many ways of learning more about alternative data, but hopefully, some of the suggestions here will give you a good start. If you have any suggestions about other resources for alternative data let me know too, as I realise this article hasn’t covered everything.