What have I learnt after 9 years of independence?

Over 15 years ago, I had seen the original scroll of On The Road by Jack Kerouac at the New York Public Library. A few years later, I finally read On The Road as it had been written in the original scroll. The “stream of consciousness” style was very different to anything I read before, and I enjoyed it greatly. Sometimes I think about the way we scroll through social media these days, whether it is LinkedIn, Twitter or anything else. In a weird sort of way, these continually updating timelines are like a stream of consciousness . I’m not suggesting we should all go and frame our LinkedIn newsfeed as some work of literature.


I recently spotted something on my timeline. LinkedIn reminded me it’s been nine years since I quit my job at Nomura and began to work independently. In a sense, I feel I’ve come full circle, given that after nine years, as I’ve also recently returned to working in the City, a few minutes walk from Nomura, at Turnleaf Analytics which Alexander Denev and I cofounded recently (and yes, our office happens to be fairly close to a number of very good burger joints…!). I thought it would be a good idea to write about what I’ve learnt working independently all these years.


I’ve gone through various paths during my independence working in a startup environment. Initially, I was primarily focusing on writing research papers. Then I moved towards consulting and later teaching. I’m still doing some teaching with Thalesians/QDC and at Queen Mary University of London, and maintaining my various open sources projects (and of course continuing my blog here). My main focus, however, is on Turnleaf Analytics, which is a data firm, where we forecast inflation using machine learning and alternative data.

It’s all learning process, no matter what the result
All of these projects that I’ve done over the years, have got some similarities! Notably they all in the area of quant finance and with a focus on FX, macro and related fields. There are also important differences, for example between doing a consulting project and creating a data product. Of course, some projects haven’t gone as well as I had hoped, such as my initial foray into selling independent research. However, the experience proved useful when I moved onto consulting. All these past experiences, have also informed what I’m currently doing at Turnleaf Analytics, whether it is on the technical side or on the marketing side.


Getting things done needs a team
I very much enjoyed working in a team at Nomura. Over my independence, I’ve worked on many projects both individually and as part of a team. Working on your own doesn’t necessarily mean you can’t get things done. You can do! However, when the problem is sufficiently large there simply isn’t enough bandwidth. Furthermore, having complementary skillsets makes the team stronger as a whole. I enjoy teaching too, which comes in useful when training more junior members of the team. Training the team is a key part of my work at Turnleaf Analytics. It’s a two way street though, I’ve learnt a lot from the rest of the team too. Of course the caveat, is that it needs to be the right team!


Yes, data is important
Everything I’ve done my whole career, whether in banks or independently, has required data in order to model markets. There is so much focus these days on machine learning. Indeed, we use machine learning extensively as part of our inflation forecasting at Turnleaf Analytics. Not all relationships can be modelled using simpler linear models, given there are so many non-linear relationships in markets. Machine learning, however, does not negate the need to spend time on the data initially. We have spent a huge amount of time ingesting dataset, cleaning it, sorting etc. to provide an input for our inflation forecasting model. It’s also important to get a feel for the data, how it looks, doing an EDA, what it represents etc. Data shouldn’t be a “black box”, we need to know what it represents. That can help give us clues in terms of knowing how to clean it and model it from an economic standpoint.


I very much enjoyed working in banks over the years. Going independent has required a different approach compared to working in a large company. Over the past few years, working in startups has been great too. Of course, the path I’ve taken during this period has changed, and all these experiences have been helpful for what has followed. I’m happy where I’ve ended up at Turnleaf Analytics.