My lessons from a fintech startup

This week, LinkedIn reminded me that it was the anniversary I had become independent. Seven years ago this week I quit my job from Nomura, to work independently, both at Thalesians and latterly with my focus at Cuemacro. I’ve found every anniversary to be a good point to look back to think what has gone right and what has gone wrong (and very wrong) during these past few years. There’s clearly a lot I have still left to learn about startups, and for Cuemacro to grow. However, if you’ve gone out to work independently or you’re thinking of going it alone soon, I hope you find the article useful (and any feedback is welcome). 

Planning your independence
I somehow negated to do much planning before I quit my job. My plan was to simply continue doing what I was doing at Nomura. I would sell research subscriptions, and all the clients who would read my quant research at Nomura would automatically switch over. Let’s just say they didn’t switch over, because now I was asking to charge them for research they had previously got for free. No one of course owes you, and whilst I would love to think that my research was being read, because it was written me, clearly a part of the reason was that I was at a large bank. It was also a time well before MiFID II unbundled research (and even with MiFID II, it is not clear it has really helped independent research providers).


I eventually ended up pivoting away from quant research subscriptions and I’m glad to say it has turned out the right move, after learning from  the problems I encountered along the way. These days I still do a lot of quant research and I can leverage all the experience I have in this area. However, it’s paid for in different ways. For example I’ve done consulting research projects for data firms and funds. Also in terms of products, I’ve got a TCA Python library and data/trading signals as well as training courses for Python and alt data. Products are of course more scalable than consulting projects.


I would also argue that independent work and the startup life isn’t for everyone. I’m glad I took this path, but acknowledge there were many ups and downs. Whilst I’ve stuck to a very similar area that I was doing in banks (albeit with some embellishments), in many cases people might do startups in a totally new area. I must admit that must be even more daunting, but planning is perhaps even more crucial there.


Listen for feedback
What feedback I did get about my research subscription idea pretty much flagged the problem with it, that clients were used to get it for free. Whilst, it is possible to sell research subscriptions and many do, it tends to be more economic or macro research, rather than quant research.


I also got feedback, all those years ago, that maybe I should get involved in doing transaction cost analysis for FX. The timing for that would have been excellent, but I didn’t follow that advice either, as it just seemed to far outside what I did. In the end, I did get involved in TCA space a few years later, developing a Python based library for a large European asset manager and a large part of it has been open sourced as tcapy. Whilst, I might be somewhat late to the TCA party, I’m hoping my business model of open sourcing it, and charging for commercial technical support, will be somewhat more attractive than clients spending six figure sums on developing their own TCA or having vendor lock in with closed source solutions.


I’m not saying you should do everything everyone tells you, but at least take it on board and think about it. Indeed, some feedback I got I was probably better off not going with.


You’ll need some source of funding
For the first few years of working independently, nearly all my income was from systematic trading my own cash. It was enough to live off and basically funded working independently at the start. As my client work took off, I ended up focusing more on that, as ultimately, even though I achieved a good Sharpe ratio, trading was taking up a lot of time, and I could never build up my notional, as I was continually taking out P&L to pay for bills.


Ultimately though, the key point is that it takes time to get established as an independent, and there might be a period of time, where your business is not generating sufficient income, so you’ll need to somehow bridge that gap.


Getting external funding can make that runway time longer and also help in expanding your business quicker, given you it can be used to fund hires and investment more in the business.


Trading can help in other ways for a startup
I’m not suggesting everyone trades systematically to go independent (!) but there are other ways a trading mentality can help. One is in terms of diversification. Would I ever recommend someone put all these risk in one currency pair? Probably not! Diversification of what approaches you try for a startup can also help. If you stick to one thing, and can do it very well, that’s great! But if it doesn’t work out, then you’ve placed all your eggs in one basket.


It becomes particularly challenging if you haven’t gone down the route of securing external funding, where potentially you might have more time to try your “one big idea”. I certainly couldn’t spend many years doing this, I had a generate P&L (well, generating P&L is a trader thing too..). Of course P&L doesn’t just include your revenues, but also your costs. Keeping costs under control is key and being able to manage the risks of a downside.


Another thing that traders are well acquainted with, is that concept of luck. That can be pretty helpful! No one really wishes to attribute their success to luck, but let’s face it, it can be helpful. Obviously, we can try to maximise our chances at the same time. There were lots of episodes I can try to attribute to “bad luck”, but equally I had some good luck. When I started my career as a quant strategist developing models for trading and looking at funky datasets (I looked at Google Trends data in 2007!), it wasn’t particularly fashionable. However, fast forward to 2020, and everyone is talking about quant trading, machine learning, alternative data etc. I’ve managed to be in the right place, but probably more out of luck, because I happen to like maths and coding, rather than any particular foresight I had at the start of my career.

Sales is key
One thing that I’ve learnt over the years, is that sales is probably the most important part of any startup. True, you might have a great product, be able to deliver great results. However, unless other people know about it, and are willing to buy your product (or use you as a consultant), you aren’t going to get anywhere. No clients means no revenue.


Doing your own thing is never easy, and let’s just say that I’ve learnt a lot from the process (and there’s still a lot left to learn, after all, there’s still a lot left to grow). Ultimately, if you choose to work independently and develop a startup, you’ll always still have a boss, namely your clients! However being independent gives you more freedom to do what you want. This of course means you still need to make the right decisions.