Hugo Shi

@hhuuggoo

Entrepreneurship vs Quantitative Finance — an experiment

August 18th 2018

On June 14, 2018 I walked out of my hedge fund job for good. I had spent the past 3 years building technology as well as researching strategies for automated quantitative trading strategies. I had resigned 30 days prior to seek our other opportunities in the quantitative trading space, but I was hit a non-competition agreement.

In finance, non-competes are common. Companies are so worried about loosing intellectual property to competitors, that they will pay your entire salary for a period of time (in my case one year) to keep you from working for competitors.

A trading buddy of mine, Eric, has been arguing that entrepreneurship has better long term financial outcomes compared to finance. He argues that starting a business does not have to be risky if you pick conservative problems (vs trying to be the next Uber) and apply rationality and creativity to solving business problems. His view, and I am inclined to believe this or else I would not be testing this hypothesis, is that the 1–2 year outcome of working a high paying job (like finance) is better than starting out as an entrepreneur, however at the 5–10 year time horizon entrepreneurship outperforms.

Sounds pretty good to me, and I’ve got non-compete funding for the next year, so I’m going to give it a go. For the next year I’m going to try to bootstrap a profitable business, and I’ll document what I do here. My 1 year target is to have a monthly profit approaching $4000. My target 3year target is to have yearly profit around 500k with 100k annualized volatility. It’s important to me that my profit should not scale with my labor. Businesses where profits scale with labor (consulting for example) are not materially different to me from working in finance. I’m not shooting for a gigantic idea (so no ubers or unicorns) because those are riskier, and I am optimizing for low risk.

Ideas

The first thing I needed was an idea. I don’t think ideas are that important — I think good execution is important. There are plenty of profitable ideas which are ridiculous concepts. I’d like to single out poop senders and dicks by mail as 2 ridiculous companies that I’ve heard are quite profitable. I think having a non-insane idea is required, as is excellent execution. I was having some trouble so Eric recommended that I try to become an idea machine. That article is long but pretty entertaining, so you should read it. The short version is that we’re not used to thinking of new ideas, and practice helps, so every day you try to generate 10 new ideas. They don’t have to be good. Trying to come up with a really good idea is hard. Trying to come up with 10 terrible ideas is much easier, and often one of those ideas is non-terrible.

Here are some of my terrible ideas — mystery spice shaker (random spice comes out each time), bringing back squeeze-its, bicycle valet parking, and a baby themed bar, where drinks are served in sippy cups. I’ve also had some more reasonable ideas. For 2 weeks, I was convinced I was going to let hipsters send a tiny succulent plant in a beautiful box with a nice pot and a card to each other, and call it succulentgram. I even bought the domain (I decided against it because I don’t really have the space to bootstrap this kind of business, and I wasn’t ready to spend hours stuffing boxes full of tiny plants).

Sometimes I share my ideas with others, and people object becuase the idea already exists. I don’t think that’s relevant — you should try telling that to the second Italian restaurant ever created. The more relevant question is — how crowded is this market, and will I be able to get customers.

Cheap Chromebooks

My first idea was to build a comparison site for Chromebooks. Affiliate fees are around 4–8$ for a Chromebook, and Chromebooks are pretty popular these days. I also just bought one (I use it as my primary development machine, but I only use it as a web browser and an ssh client, I don’t try to do Linux-y things on my Chromebook). I purchased cheachromebooks.com (And a few other variants ) and got to work. I abandoned this for a few reasons. Amazon will ban your affiliate account if you try to use try to scrape amazon for prices. To show prices, you need access to the product advertising API. To get access to the product advertising API, you need to be an accepted affiliate. To be an accepted affiliate, you need to have some affiliate revenue already. It’s a chicken and the egg problem. They say you can ask for an exception, I did, and was refused.

The impression I got from this experience is that since affiliate marketing is super crowded, vendors are now being super strict and conservative about working with new people. Tiger Direct rejected me from their affiliate program without providing any reason. I think I could have made this work — I think if I built a successful blog around Chromebooks, and displayed pricing from the sites that would not ban me for scraping, I could have had a good enough site such that vendors would have been willing to look at me. I don’t have a good reason for turning this down, except that I no longer liked it, and the idea of scraping a bit of premium off of transactions reminded me too much of trading.

Something else I learned — don’t buy domains until you need them to test your idea. I’ve purchased 8 domains I’ll probably never use so far.

That’s the story of how I got started. I’m going to try to write a new post every week or so detailing the experiments I try, and the lessons learned. I’m 2 months behind right now, so they’ll be a bit more frequent as I catch up. If this interests you, then you should follow along. Wish me luck!

TL;DR;

I used to work in quantitative trading but quit and am serving a 1 year non compete. I believe finance payouts outperform entrepreneurship on short time scales but long term entrepreneurship wins. I’m testing that theory, and you can follow along to see what I learn. To come up with ideas, you should become an idea machine. There are many crazy ideas that work, like selling poop or gummy penises. Just because an idea exists, does not mean you shouldn’t do it. Or else there would only be one Italian restaurant in the east village. My first idea was to build a comparison site for Chromebooks, which I’ve abandoned because I don’t like affiliate marketing. Finally — you should not buy domains until you absolutely need them. That’s why I have 8 I’ll never use.

Thanks for reading — my past projects are about simple easy email groups, and getting help for open source projects. My current project provides cloud hosted Jupyter notebooks In the next article of this series, I write about building GroupMail.

More by Hugo Shi

More Related Stories