Just to be clear, I'm not the author of the article. LOL.
Why I quit algorithmic trading
- Thread starter Maverick74
- Start date
Just to be clear, I'm not the author of the article. LOL.
I have an idea of what is going on and how I am being gamed by the market, but I haven't put together a good solution yet. Lot of intense research the past few weeks. I'm wondering if the guy who wrote the article just hates the low points, where the PnL starts drying up and the solution isn't an obvious fix.
Interesting story... thanks for the insightful read.
"Every time I teamed up with a trader in an automated trading venture, their trading strategies and ideas ended up not working despite the fact that they had been previously successful (often remarkably successful) as either floor or screen traders."
-----
Perhaps the strategies didn't lend themselves very well to being automated.
-----
Perhaps the strategies didn't lend themselves very well to being automated.
I've been writing code in C++ for about 15 years and I have to agree with just about everything you've said, especially point #6. It's been my experience that programming trading apps are better served as trading aids rather than trading bots. One of my most valuable apps is a tool that measures volatility that is based in part on ATR. It's highly reliable at signaling trend vs non-trend days. I've played with writing pure algo trading apps and I've never been able to come close to what the human brain can interpret. Trading is an art and writing code is a science although writing code does require some creativity so arguably there is an element of art in writing code.
Point #3. Beating the market shouldn't be a goal as much as trading profitably should. I've beat the market by a wide margin in some years and trailed the market in others. The key is to try and beat the market over a long period of time. Doing it every year consistently is extemely difficult.
Point #3. Beating the market shouldn't be a goal as much as trading profitably should. I've beat the market by a wide margin in some years and trailed the market in others. The key is to try and beat the market over a long period of time. Doing it every year consistently is extemely difficult.
Quote from Maverick74:
http://www.codusoperandi.com/posts/why-i-quit-algorithmic-trading-to-do-web-startups
Why I Quit Algorithmic Trading to Do Web Startups
The kernel of this post originated as an overly long answer to a question posed to me on Google+, "So how come you stopped doing that stuff [automated / algorithmic trading]? Sounds like you'd have enough experience to be a successful algorithmic trader."
I spent the better part of six years, from roughly 1999 to 2004 and again in 2008, pursuing the chimeric fortunes of automated / algorithmic trading and while I never actually reached the promised land of a wildly successful trading system, along the way I did learn a lot about myself, the world, and how to write robust, highly perfomant code. The journey itself deserves its own post, or possibly several, as it was quite a strange adventure, but to get this subject kickstarted, I'll just begin with why I ultimately moved on to the web and startups.
1. An automated trading operation takes too much trading and operational capital (not to mention extremely low clearing rates, which are extremely hard to negotiate) to easily bootstrap on your own. Not that it's impossible, but it's definitely much less resource-intensive to attempt to make money from building web or mobile apps.
2. Every time I teamed up with a trader in an automated trading venture, their trading strategies and ideas ended up not working despite the fact that they had been previously successful (often remarkably successful) as either floor or screen traders. This usually meant that I had spent a year or more of my time coding up a state-of-the-art trading platform that amounted to nothing.
3. I'm not entirely convinced that it's all that possible to beat the market consistently, much less using algorithms, much less using self-learning algorithms (AKA data mining). From my perspective, buiding something that some number of people would be willing to pay for seems to be a much more tractable problem to solve. That said, I do think it "might" be possible to beat the market algorithmically if you had all the right pieces in place, the trick is just figuring out what those pieces might be.
4. If you spend years working on an automated trading project (usually in total secrecy) and it doesn't work out in the end, it can be hard to leverage that expertise for anything else. On the other hand, if you work on a web / mobile project and blog / tweet / podcast about the whole experience (which everyone knows you should do), the side effect is that you'll have built a public reputation that can later lead to all kinds of unexpected opportunities. This is what I refer to as increasing your luck surface area.
5. By building web and mobile applications, you're at least "attempting" to create value for the world and not just yourself. This may seem like a minor point, but if your trading venture doesn't succeed financially, you can't even fall back to the - "well, at least I made a lot of people happy, more productive, etc.". Plus, the thrill of having a large number of people use your software is something you'll never experience within the confines of an automated trading operation.
6. Trading is extremely stressful even if it's the machine itself that's doing the trading. In fact, I read a scientific study a while back that found that a losing trade is twice as psychologically draining as the equivalent winning trade is psychologically reinforcing, which basically means that you're generally going to be emotionally drained and operating at a psychological deficit at the end of every work day. I know a lot of startup founders like to talk about how hard startups are and about the incredible roller-coaster ride that is the startup life (which is true), but I got news for you, it doesn't compare to the grinding, gut-wrenching stress of trading, and frankly that's no way to live.
7. I've done a lot of the trading thing and while it's a fun and addictive game for sure, I found that traders are generally not my kind of people. The reason for this I think is that for most traders it's pretty much all about the money, and whenever anything is all or mostly about the money, it ends up having no soul. This reality turns out to be kind of depressing if you reflect on it for too long and I guess that's probably why most traders aren't all that self-reflective. That said, I'm not so sure this applies to algorithmic traders because from the few I've met they tend to love the technical challenge as much or more that the pursuit of fortune (which is not unlike technical startup founders).
8. The algorithmic trading world is so secretive that you rarely get to meet anyone else doing it, much less have the opportunity to discuss techniques, algorithms or experiences. As a result, there's little to no community to engage with, and in case you haven't already discovered this truth, being part of a community is a big part of what makes life fun.
Disclaimer: If I ever make a personal fortune from web startups, there's still a part of me that would like to give algorithmic trading one last try. There, I said it.
Never give up and find a way to adapt to changing market conditions. The reason traders lose their edge is because the market conditions change and the trader sticks with the same strategy that will no longer work. The transition from 2008 to 2009 is a perfect example. We went from a high volatility, high intra-day range to a low volatility, low intra-day range. Scaling in and out in a volatile market is a viable strategy. Not so in a non-volatile market. Get caught scaling on the wrong side of a trend and you're up the creek.
What didn't I figure out? I posted the article you tool.