I can say for certain that I’ve never been kicked out of my apartment by the landlord for consuming all the power to the entire building, or been flown half way across the world by a billionaire for a favour. But these and other tales are just some of the fascinating insights in our next guest's journey from graduate to hedge fund founder to trading over a billion dollars a day.
Christina Qi started HFT hedge fund Domeyard with two others out of her dorm room. With hurdle after hurdle and a lot of ups and downs, Domeyard has become a successful outfit and was ranked in a recent list of the top 50 hedge funds.
Not content with this and wanting to give back, Christina has now gone on to start another company, but this time tackling the problem of high cost financial data, which is sometimes a barrier to entering the industry. Her company Databento aims to provide institutional quality data at retail prices, enabling teams to launch faster and scale seamlessly throughout their growth.
Without further ado, let’s get into Christina’s interview below...
Don’t forget your moral compass, and don’t bend to anything that compromises it.
1. Can you remember when, why and how you got into trading?
My unusual career arose from a string of failures. I had interned on the trading floor of a bulge bracket bank, and we were required to come up with a stock pitch for our final project. Given my technical education at MIT, I decided to go full-out, with a statistical arbitrage strategy. I was naive at the time, and didn’t realize that a discretionary trading floor would be taken aback by my quant strategy. I was devastated to not receive a return offer. Since I didn’t have a PhD, I didn’t qualify for jobs in high frequency trading (HFT) at the time. So my only choice was to dabble in trading at night, during my senior year, trading the German session at 3AM and going to class during the day.
My hedge fund, Domeyard LP, was born out of this experience. I had no idea at the time how difficult it would be to launch the fund, how great it would feel to trade over a billion dollars in a day, and the magnitude of the “first-time founder” mistakes that we made.
Recently, I started a data company, Databento, to lower the barrier of entry for hedge funds and traders. We’re the only firm that provides HFT-quality data in very customized, a-la-carte batches, with the goal of scaling with a team throughout the lifecycle of a business. Our order book data is up to the nanosecond resolution, but users can choose whatever resolutions and symbols they wish. This is completely separate from my hedge fund – you could say it’s the more meaningful thing I’ve been working on.
2. What was your trading style when you first started and what did you trade? Has this evolved since?
We’ve been in HFT for the past eight years, since inception. It was natural for us, given that our only edge was a technological one.
There are millions of orders coming in at every hour of the day. In order to truly understand what’s happening, we look at data through a microscopic lens and capture the opportunities that other folks may miss. We may be in and out of the market before an old-school platform notices our presence. It may sound daunting, but I would argue that it’s scarier to place giant, long-term trades based on uncertain economic forecasts. Our models are actually very conservative and data-driven. The law of large numbers also factors in - when you make thousands of trades per day, your results are based on skill, rather than luck. We don’t require a high AUM, as our portfolio is turnover dozens if not hundreds of times per day.
Contrary to popular belief about HFT, Domeyard has never bought order flow before. We only trade intra-market, and use market data in addition to basic alternative data, to fuel our strategies. Our colocated servers are assembled and then placed in exchanges’ data centers, whose racks are leased by popular vendors. Outside of order flow, HFT is surprisingly a level playing field, and the only times I experienced unfair treatment was from fund administrators, lawyers, recruiters, etc.
3. You set up a HFT hedge fund with two others a little over eight years ago back in 2012. We know that you can’t say too much on the details and performance of Domeyard but can you briefly provide an overview of the fund, how it came about and the journey that took you to now trading over $7.1b a day across thousands of trades?
Well, it took is almost three years to launch the fund. That shows how difficult it truly is, to launch a full-fledged hedge fund in this day and age. It also shows how inexperienced we were as college-aged founders at the time.
I always tell people that you can do better than me. You gotta know your strengths and weaknesses, have a solid team of employees, advisors, service providers, and people to hold you accountable, be a good leader, and learn lessons from the rest of us.
It took us forever to find those day one investors. I mean, who wants to be a day one investor in a hedge fund with 0 track record, young founders, and in a controversial field like HFT? Surprisingly, there are allocators who specifically invest in early-stage fund launches. There are also industry veterans who want to invest in the next big thing. But the reality is that it’s still tough to find those folks. Once we launched the fund, everyone wanted to be our day two investor, so we ended up with a giant waitlist over time. It’s crazy how fast things change once your fund launches and you build a live track record.
In terms of launching, I would advise early-stage funds not to launch too early. This isn’t like Silicon Valley, where you can launch a crappy overnight MVP and incrementally improve upon it as customers give you feedback. Think of this more like Theranos – they’re dealing with human lives, and can’t afford a single false positive or negative. As a hedge fund, you’ll be dealing with real returns, which will be tattooed on your track record for the rest of your trading career. Sometimes, Wall Street only gives you one shot. So, as tough as it is to see your bank account dwindling, it’s important to take your time and launch when you feel truly ready.
4. You actually started Domeyard out of a flat and had your servers stacked into a cupboard. There’s an entertaining story around this and the reason you had to move out to an office. Can you elaborate on this?
This is a summary of a chapter in my book, tentatively called, “The Finance Snake: A memoir about my billion-dollar hedge fund.”
We rented an apartment near MIT before we launched the fund. It was 3x cheaper than office spaces of the same size. We rigged the electrical system in the apartment and installed our servers in a closet. Since it was a residential space, I lived in the master bedroom, which was super luxurious for a new grad.
One night, the police knocked on my door and raided the apartment. They thought that we were running a marijuana farm, and were absolutely shocked to see the roaring servers in our closet. It turns out that we were using up more power than the rest of the building combined. That night, they served me an eviction notice, and we moved into WeWork the next day. Our rent shot up threefold, and our WeWork space was literally a tiny cell surrounded by plexiglass. One investor thought it was a dorm room. To be fair, it definitely smelled like a dorm room, since the seven of us were crammed into a poorly-ventilated prison. I had to move into a friend’s apartment, and slept on her couch for a while.
5. You built your hedge fund systems from scratch, which I can imagine is very time consuming and has its pros and cons, but it includes everything from the trading pipework / infrastructure connected directly to the exchange, to the data collection and storage for your modelling and testing. Was this a choice to build it out rather than pay for software?
We didn’t really have a choice back then. In terms of “buy vs. build” we had to build, because there weren’t any good HFT trading platforms at the time. In addition, we didn’t have a lot of money, so we decided to build everything from scratch. The only downside to this is that it takes an unknown amount of time, and delays happened more frequently than not. So it took way longer than I expected to get to launch, and it was a frustrating journey in that sense.
6. Your experience with establishing a successful hedge fund from your flat has led to you setting up a financial data services company called Databento. Can you please explain briefly what Databento is and how it can help people in the industry or even outside the industry? (I feel like there is a good story behind the name too, can you elaborate on how you settled with Databento?)
In the process of launching my hedge fund, we wasted over a million dollars in data. There were a few startups that offered cheaper data, but the quality was horrendous and meant for retail traders. We’re launching Databento to provide institutional-quality data at retail prices, enabling teams to launch faster and scale seamlessly throughout their company's lifecycle. Instead of talking to a salesperson, teams can instantly purchase, utilize, and share multiple data sets within a company's existing infrastructure.
The name Databento derives from the idea of purchasing smaller chunks of various datasets. Instead of buying everything in bulk from traditional platforms, we allow users to only pay for what they need. It’s kind of like filling up a bento box of your favourite foods.
7. What was the main motivation of setting up Databento? What are you hoping to achieve with the service?
A lot of startup hedge funds have asked us for advice over the years, and we realized that everyone faced a universal challenge of finding better data and infrastructure. I’m very fortunate to have grown my hedge fund over the years, and I want to make a meaningful impact on future generations, by addressing one of their biggest challenges.
We’ve also attracted a variety of larger institutions, who have various use-cases for us, whether it’s launching a new strategy, or providing backup data, or helping them transition to remote work while saving on costs.
8. What sets you apart from other data providers? Is it cost, your technology and/or something else?
This is actually similar to how we differentiated our hedge fund. We aimed to be solid in more ways than one:
- Product offering is first and foremost. We can simultaneously offer the highest-quality data, with a very small set of atomic and orthogonal API functions, and offer it in smaller slices to save money.
- Then it comes down to pricing, ease of use and delivery, and the overall customer experience.
9. You’re running an online ‘Black Mirror – Bandersnatch’ style session where the users will get to interact and make collective decisions on the direction your theoretical start up hedge fund should take, and in turn this will affect the future direction and choices. Can you expand a bit why you decided to set this up for free and what you hope people will learn from it?
I was inspired by Netflix’s Bandersnatch, and wanted to create a hedge fund startup version so that people can experience, as realistically as possible, the rollercoaster of starting a hedge fund and the consequences of their collective decisions.
It’s been the highlight of my quarantine, and over 1000 people have attended thus far! Each session results in a different ending – it’s surprising how people vote differently from one session to the next. We’ve experienced the worst ending possible twice already, but have yet to unveil the best ending possible.
What’s fun about the session is that everyone gets to meet each other. People will share their Linkedin and Twitter profiles, and continue discussions well after the event is over. Through Bandersnatch, I’ve introduced co-founders to one-another and met people from all around the globe!
I also host a fundraising and hedge fund technology version of Bandersnatch. It’s free because it’s incredibly fun, and oftentimes funny and exciting!
This is the universal sign-up form. I promise to keep emails to a bare minimum, and will email you when the next session is scheduled so you can reserve a seat here.
10. What one bit of advice would you give your younger self 8 years ago that you’ve learnt from the incredible journey you’ve been through?
Don’t forget your moral compass, and don’t bend to anything that compromises it.
11. You’ve got a book coming out in early 2021, covering your experience of being in the HFT industry. It sounds very exciting. What drove you to write it all down and the decision to publish it?
Writing this book has been a huge weight off my shoulders. For the longest time, I was embarrassed about the many mistakes we made on our hedge fund journey. Through hilarious stories, my memoir details the well-intentioned principles that we set up early on, and the fallout of those exact principles that we thought were “genius” back then.
One of the biggest mistakes I made was calling everyone a Partner. (Google us and you’ll see articles about this.) Back then, it sounded rational – I was young and needed to hire people 3x my age. So giving everyone the Partner title was supposed to make them feel more valued. But boy, did this backfire on us. We had an office manager, whose job was literally to clean the kitchen, schedule meetings, and print papers. We felt bad and gave them the Partner title, just like everyone else. So what did they do? Demand a salary of half a million dollar per year and the title of Managing Partner. People need a career ladder in this industry, and they need proper titles that reflect their level of experience.
My book is also a tell-all about the lavish world of billionaires. For example, one of the richest men in Asia invested in my fund. I found out later on, that he only invested so that I could help his daughter get into Harvard. (Spoiler: She did not get into Harvard.)
Do your reverse due diligence before accepting any investor. Don’t make my mistakes.
13. We see you’ve achieved a number of first places in competitions including piano and Mahjong. Do you have any other hobbies, and how do you like to relax outside of work?
I watch anime and play board games in my free time. Writing my book has also been a form of relaxation for me. I hope you’ll enjoy my book, once it’s published in 2021.
If you’re interested in starting a fund or expanding your trading footprint, I hope you’ll sign up for Databento and meet me in Bandersnatch!