I’d think so too. But try to tell that to those retail quants/traders. Here on ET many people are concerned about desktop software and brokers stealing their strategies, before they even create a strategy.
Last edited:
Quantopian’s Community Services are Closing
- Thread starter greejan
- Start date
I’d think so too. But try to tell that to those retail quants/traders. Here on ET many people are concerned about desktop software and brokers stealing their strategies, before they even create a strategy.
a billion monkeys typing, one might write Shakespeare
They will steal their money and take it to Bolivia, long before they will feel the need to steal their pitiful strats
Thats a fallacy.
The truth is probably not nearly as exciting as the theories from most of the posters on this thread have come up with. Quantopian was venture funded. The way venture funding works is you get a bunch of cash, which you are expected to spend quickly to build an otherwise unsustainable team and succeed or fail quickly. If you're looking like you're on a path to being able sell the company or do an IPO for a couple hundred million or more you'll get another round. If not, you won't. The end result is that you almost always just go away even if it could be a decent $25M company, because at the point you're spending the last dollars of your last round you still have a salary base that's unsustainable with your current revenue. And you need that team to provide the level of service you need to stay in business. And the risk adjusted returns just aren't there for financing these kinds of entities for the next 3-5 years to the point they do generate enough revenue to break even. Quantopian did the whole venture funded thing, complete with an attempt at a pivot at the beginning of this year. It just didn't work out, the revenue wasn't there. If you were watching, they laid off most of their team over the past couple months, so this wasn't a total surprise.
They basically had a hypothesis that there were these hidden quants out there that they could find AND monetize. I'm not sure that they were even able to demonstrate the first part, but definitely not the second. I won one of the first monthly contests they ran (so I ended up helping part their VCs from some of their money!) and was pretty quickly contacted by several hedge funds that clearly already had a "farm team" concept set up to find the same elusive hidden quants that Quantopian was looking for. Since we use our real names on Quantopian it wasn't hard to find us, several of us have kept up with one another informally since then and we all had the same experience. All the people who wanted to be quants for a hedge fund went to work as quants for a hedge fund, and Quantopian didn't get so much as a finders fee for identifying them. The rest of us just do it for fun, and were never going to be interested in doing it as a job at any price let alone what Quantopian was able to offer. I really liked the platform back when it came out, and although I haven't used it in some time I'm sorry to see it go. At least it sounds like Fawce was able to negotiate to keep the IP and run a lifestyle business out of it going forward in a way that doesn't require much staff; good on him.
There is a saying: "if you don't pay for the product, you are the product." The community just couldn't find enough scalable alpha to make it worthwhile to institutional investors. They have been in trouble for a while now: their CIO was ousted after a lackluster start to their fund. While you can chalk it up to bad luck, it just underscores the fact that past-performance (even well-backtested) just doesn't guarantee future results.
Add to that the costs of not only running the infrastructure, but also the ongoing cost of licensing data and providing it for free to the community. Plus, the frontier of quant strategies has moved into the machine learning era, which can take serious compute resources (and dollars). If these strategies remain off-limits to Quantopian users, they fall further and further behind other market participants.
I'm sad to see Quantopian ramp down (or at least try to pivot). They offered exceptional educational material and produced an enjoyable quant finance conference as well as some useful open source projects: worth checking out Pyfolio and Alphalens for portfolio analytics if you have not done so.
Add to that the costs of not only running the infrastructure, but also the ongoing cost of licensing data and providing it for free to the community. Plus, the frontier of quant strategies has moved into the machine learning era, which can take serious compute resources (and dollars). If these strategies remain off-limits to Quantopian users, they fall further and further behind other market participants.
I'm sad to see Quantopian ramp down (or at least try to pivot). They offered exceptional educational material and produced an enjoyable quant finance conference as well as some useful open source projects: worth checking out Pyfolio and Alphalens for portfolio analytics if you have not done so.
