A decade once seemed like an insurmountable stretch of time to me. Yet as the years have raced by, a decade has become far more tangible. It was with a certain sense of shock, that I realised it has been exactly a decade since I quit my last “proper” job, at Nomura. So I thought I’d write a bit about what I’ve learnt in the past decade and in particular over the past year since cofounding Turnleaf Analytics, where we forecast inflation using machine learning and alternative data.
I still remember sitting across from my manager at the time, explaining that I wanted to leave. I had actually really enjoyed time there, but it felt as though each year was somehow melding into the other. This isn’t to say that an element of routine is bad, it’s simply that when routine becomes too persistent, challenge seems to vanishes and it’s time to try something new. Sometimes you just have to do it!
In retrospect, I quit my job somewhat prematurely. I knew I wanted to work independently, but my initial idea really didn’t work out. The goal had been to replicate what I had done successfully at the bank, publishing quant research notes but as an independent. The difficulty was that before the advent of MiFID II, clients were still used to getting free research from banks, and it wasn’t a barrier I was able to overcome at the time.
Ultimately, I hadn’t done sufficient research into the viability of the business before starting, and this proved an important lesson for the future. Understanding the market’s appetite for a product, as well as its pricing is critical. Of course, not everyone will necessarily want to buy your product or invest in your firm, but there needs to be a critical mass of clients and funding to create and grow a business. There is always the challenge with a startup that you no longer have that big bank brand next to your time. You need to elevate the brand yourself.
As a result, I ended up funding the research idea from my own independent trading of intraday FX models. This hadn’t originally been the idea, but it ended up being an exceptionally good learning experience, that if I put my mind to it, I could live off my own trading (and it also makes it easy to answer the question why I haven’t traded my own models before). The downside was that I realised that to do this long term, it was necessary to have a large amount of capital to start! The end result was any P&L I generated was swallowed up by bills, rather than being continually reinvested, to grow my notional.
I then pivoted towards consulting, and over the years this became very successful. I did research projects for many firms ranging from data firms like Bloomberg to hedge funds and asset managers. I also managed to collect together a lot of the ideas around alternative data which I had in The Book of Alternative Data, alongside Alexander Denev. I began teaching on topics including Python and alternative data. Today, I still do some teaching, through Thalesians which I cofounded (on the QDC), and Cuemacro which I also founded. Furthermore, I teach students at Queen Mary University of London, where I am a visiting lecturer. My open source libraries such as finmarketpy, findatapy, chartpy and tcapy have also attracted a large following over the past decade, collectively notching up well over 100k downloads. Working independently gave me an opportunity to build my network and learn many new skills.
Whilst consulting was fun and I managed to work on a wide range of projects, ultimately, consulting is not as scalable as other businesses. With my fellow coauthor, Alexander Denev, we began to explore the idea of forecasting macroeconomic variables in the spring of 2021, building on ideas from our book. We founded our firm Turnleaf Analytics in December 2021, and completed our first round of funding in May 2022, when we also started publishing to our inflation forecasts since clients. It’s been over a year since then, and we have fully developed our forecasting product, and have signed up a large number of paying clients. We’ve built up a decent track record of forecast performance in our live publishing too. We’re currently in the midst of our second round. We have grown our team since our founding, and built up a decent array of paying clients. It’s definitely been great to go back to working in a team at Turnleaf Analytics.
The business model of Turnleaf Analytics is product based, distributing inflation forecasts. This is obviously significantly different to a consultancy. Often the image of founders can be of twentysomethings, perhaps skewed by the massive success of folks like Mark Zuckerberg. In practice, I’ve found having experience of working in a multitude of different environments, whether in banks or startups, has been invaluable at Turnleaf Analytics, to understand what has worked, and even more importantly what hasn’t. Also, I am probably more aware now, that there is a huge amount I really don’t know – and in any successful firm, you have a team, a cofounder, a board etc. Collectively, you are stronger, particularly when you understand how your strengths and weaknesses can be balanced within a team.
I’m looking forward to my second decade of independence. The first one has certainly been an experience, and it has provided lessons for the future. I’m excited for what the future holds for Turnleaf Analytics.