World Cup Trading Champion*®*, Exclusive Interview with Kevin Davey
As an award winning full time trader, and best-selling/award winning author, Kevin Davey has been an expert in the algorithmic trading world for several decades. Between 2005 and 2007, Kevin competed in the World Cup Championship of Futures Trading, where he finished first once and second twice, achieving returns in excess of 100%! Kevin develops, analyzes, and tests trading strategies in every futures market from the e-mini S&P to crude oil to corn to cocoa. He currently trades full time on his personal account. He also helps small groups of traders significantly increase their trading prowess via his award winning algorithmic trading course, “Strategy Factory“. Beyond his course Kevin also helps educate the trading community via his award winning book, “Building Winning Algorithmic Trading Systems“, published by Wiley.
Kevin discussed a variety of topics with us including how to identify when a system is no longer worth pursuing, the misuse of technical analysis, the emotional hurdles you may face when developing and running a trading system, and much more. Be sure to check out his twitter account at @kjtrading and his website at kjtradingsystems.com.
Kurt: If you could, tell us how you first got into trading and what that experience was like.
Kevin: I got into trading basically from a direct mail ad back when you used to get a lot of junk mail. It was by this guy wearing a cowboy hat and he had a real simple method of trading commodities. It was basically using a head and shoulders pattern and he showed chart after chart where it worked. He showed one example where, if you were trading sugar you just keep adding on contracts and wow your profits were enormous and it was just ridiculous, but I was hooked. So I bought his course only because they had a money back guarantee and it turns out he actually honored it because it didn’t make money in the sixty or ninety days that I evaluated it. By that point I was kind of hooked and I said trading is kind of cool, trading futures especially, because you can get more leverage and you don’t need as huge of an account as you would with stock trading for example. It just looked very appealing and so it just kind of went from there and you know I spent years going down the wrong paths and starting over and blowing out accounts and then eventually I just kind of settled on a system based kind of algorithmic trading that I do now.
Kurt: Do you think that retail traders are using technical analysis right for the most part or do you see some issues?
Kevin: I think most people use it completely incorrectly. The reason is because they just assume that because it’s a canned indicator or some software that it’s going to work. So what they’ll do is also throw down a chart and find a couple cases and, yeah it looks like it works or no it doesn’t maybe I should optimize it because it’s not “tuned to the current market”. Then they’ll optimize it and get better parameters and now it’s working great and then they’ll go on with that and, of course, you know the market ten days from now, ten weeks from now, ten minutes from now will behave completely different. Whatever parameters they picked probably won’t work going forward and then they’ll go back and try it again. That’s just with the canned indicators. People also tend to think everyone is using the canned indicators, so there’s got to be some super indicator out there that nobody knows about. So they either go try to develop it themselves or search the internet to find the greatest indicator of all time, which works until the day they buy it, then it falls apart. I think most people overuse or misuse or whatever you want to call it, they just don’t do it right.
So I think most people use technical analysis completely incorrectly
Kurt: What other criteria do you use to determine when you should add or remove a strategy?
Kevin: As far as adding strategies what I’ll do is pretty much once a month I’ll look at any new strategies that I’ve developed that meet my criteria, and then I’ll put it in the portfolio and run it through some sizing algorithms. It will tell me, yeah you should trade this or no just keep it on the sidelines. As far as quitting strategies, for the most part what I do is I track each individual strategy every month and look at its market to market performance over the past month. That includes closed trades and open trades to see how it has been doing. If you do that over a number of months you get an idea of how the strategy is actually performing so you can see the ones that kind of fall off the cliff, so to speak, where something broke, either the market changed, or maybe your strategy wasn’t that good to begin with and it was just curve fit or something. So I do that and then that might tell me it’s time to turn off these strategies and then usually what I’ll do is I’ll turn them off but I’ll still track them. If they come back maybe I’ll consider trading them again. A lot of times, at least to my criteria, when a strategy is “broken” it usually stays that way and it’s always hard to turn back on a strategy that you have already said is broken because chances are you were trading it live and you lost money with it. You know now you don’t have as much confidence in turning it back on.
Kurt: When it comes to emotional and psychological pitfalls, what kind of issues did you run into when it comes to maintaining and building trading systems?
Kevin: I mean the first thing is there’s a lot of people out there that say hey, if your emotions are getting to you just trade automated or trade these algorithms or do system trading because that will take all the emotion out. I always laugh when I hear people say that because to me those are people who if they trade, it would amaze me, because it’s just not true. What it comes down to is anytime you have money involved you’re going to have emotions, simple as that. Automated trading, what I do, is still certainly emotional but the interesting thing is the emotions are different. So for example I don’t have to think during the day should I take this trade there’s a Fed announcement, should I just turn everything off or should I walk away? Or you know, it looks like I’ve lost a few thousand dollars today should I just quit everything? The kind of questions that more discretionary traders have, you know the inability to pull the trigger kind of thing. Those kind of emotions are gone but they get replaced with, wow my automated strategy has been running and in the past week it’s lost three trades in a row. Should I turn it off or should I keep it on? So the emotions are still there because you still feel like “Am I doing the right thing?”, but it’s just on a different format and different scale. The big thing again is to come up with a process that you’ve proven works and that’s what I do. So I develop systems, but I’ve been developing long enough that I know if I follow these rules for developing my strategies chances are they’ll work going forward. They will have acceptable performance and once you know that and have the confidence that what you’re doing up front is working, it will probably work in the future and that makes the emotions a lot easier to handle. Day to day drawdowns don’t bother me because I know in the long run I’m doing the right thing. It’s almost impossible for a new trader because every time they start trading something they expect to make money and when they don’t usually they start to freak out a little bit. You’ve got to have a long term view and that definitely helps having a long term view of where you’re going with all this.
I’ve been developing long enough to know if I follow these rules for developing strategies, chances are they’ll work going forward.
Grace: So I know there’s a lot of platforms out there with a lot of information. Is there a platform that you recommend or use most often?
Kevin: Well since I trade futures I use Trade Station and I’m pretty happy with that. I’ve also done some testing in forex with Metatrader and that seems to produce pretty good results. I didn’t mention ones like NinjaTrader and MultiCharts and those are good too. The key with any of those is you really need to learn the software well enough to know when it’s giving you good results and bad results. So what I always tell people is you know your trading software well enough when you can create a system or a strategy that gives a super good looking equity curve and you know that you faked the test engine. You know that you tricked it basically to produce that curve and if you know all or at least some of the tricks to do that, it means you know your software pretty well and then you can avoid a lot of things. There’s people with every software out there that will not flip a certain switch or check a certain box and they’ll get these ridiculously good results and they’ll actually believe them. Knowing what to believe and not to believe is just all part of learning the trading software. That’s probably the biggest thing I see people doing, they kind of misuse it and think they have something when they really don’t.
Kurt: I know you mentioned keeping it simple. What do you see as an appropriate amount of variables to optimize in order to retain robustness?
Kevin: Personally I keep about anywhere from two to five variables that I’ll optimize in a system. Actually one to five systems, I optimize only with one. Once you start getting above that things start to get a little hairy. I remember once talking to somebody who was wondering why their system was not performing live the same as in back tests. So they sent me their back-test report and it looked great for profit and everything but it turned out they had almost fifty parameters they were optimizing. I said well I think that’s your problem and they didn’t want to believe me because they said well this makes a back-test look great. Fifty is way too many. What I teach people is you really want five or less, so if you start getting above that I think you’re too complicated.
What I teach people is you really want five [variables] or less, so if you start getting above that I think you’re too complicated.
Kurt: How do you know when the system is no longer worth pursuing?
Kevin: I have some criteria I use that’s just based on years of doing this. For example, if you do some simple optimization of a system over a small period of time and no matter what you do, or what parameter values you use, you’re always losing money. Chances are that system probably will not be a good system when you look at all the data for all the years. So you can kind of focus in on a piece of data and you get a good indication if things are going to work or not. Now obviously when you do that kind of thing you run the risk of having a sample where it looks bad and everywhere else in history was great and you just happened to pick the bad part of history and then you threw away the idea because you said this ideas junk when it really wasn’t. That’s the drawback to doing that kind of thing and it’s really no way around that. But what I found is the benefit of running simple tests outweighs the fact that you’re going to throw away some potentially good ideas and let you just sift through all your ideas a lot quicker.
What I found is the benefit of running simple tests outweighs the fact that you’re going to throw away some potentially good ideas and let you sift through all your ideas a lot quicker.
Risk Warning: The FXCM Group does not guarantee accuracy and will not accept liability for any loss or damage which arise directly or indirectly from use of or reliance on information contained within the webinars. The FXCM Group may provide general commentary which is not intended as investment advice and must not be construed as such. FX/CFD trading carries a risk of losses in excess of your deposited funds and may not be suitable for all investors. Please ensure that you fully understand the risks involved.