Interview with Jon Tait, the Fickle Trader

October 18th, 2006

For the next interview in the StockTickr Interview Series (RSS feed), I spoke with Jon Tait who some of you probably recognize as the Fickle Trader. Jon’s site stood out to me because of his weekend watch list that he used to post. A lot of the stocks that he posted were also on the system that I traded at the time.

Jon and I would exchange emails occasionally and at some point Jon told me about his financial software project that he was apparently putting a lot of time into. Jon is big into computerized systems trading. In fact, he may be to blame for the current state of the markets. 😉

Read on for what I thought was a fascinating interview with Jon. He describes his approach to risk management, his take on discretionary versus systems trading, and more about his very interesting softare project. Feel free to ask Jon questions in the comments and I’ll see if he can answer them.

(Also check out today’s episode of WallStrip – I think it’s the best one yet.)

StockTickr: Tell us a little about yourself, Jon.

Jon: I come from a computer science background. I’m 25 years old and currently do java programming work for Asynchrony Solutions in St. Louis. I’ve been a programmer since I was in junior high school writing C code to do my math homework problems instead of working through them by hand. The idea of writing software to deal with the tedious parts of life has always appealed to me because it felt like having a huge edge over everyone else. It is nice realizing things like this when you’re 12 years old, I feel pretty lucky for it.

StockTickr: How did you get started trading stocks?

Jon: “Solving” the financial markets seemed like a natural extension of what I was already doing. During my later years in college, I was way into artificial intelligence research, but from a pragmatic perspective rather than an academic one. Because I already had spent a lot of time learning about the big picture things, confining my AI research to the financial markets seemed like the obvious first move because I knew I didn’t want to write papers to raise funds for my ideas; building the golden goose would mean funding for life and it seemed like it would be easy if I gave myself plenty of time to do it. As soon as I had a little money saved up from my first programming job out of college I opened a brokerage account and I’ve been building my stake up to critical mass ever since. I’ve been trading for over two and a half years, although I have been working on software relevant for exploiting financial markets off and on for about four years.

StockTickr: Most traders have a horror story about losing their shirt when they first started trading. What’s yours?

Jon: For me the hardest thing at the beginning was being under-capitalized and the brokerage houses with the best commission rates are only available when you’ve got over a year of online trading experience. I’ve never blown out my account or anything dramatic like that. The reason I began discretionary trading was to figure out how markets worked and get a feel for the best strategies so that I could automate them. That kept me out of a lot of trouble because I wasn’t looking for windfall gains or vacation money.

StockTickr: What single lesson did you learn along the way that has helped you the most in your trading?

Jon: Cut losses fast. Enter trades in such a way that you can put stops where they need to be, not where you would be selling when you should be buying.

StockTickr: Describe your style of trading. How long do you typically hold stocks?

Jon: I use two kinds of trading styles, discretionary trading and systems trading. When you hear a trader say something like, “I would have made money if I had only followed my rules,” it usually means that they can’t figure out how to separate the two. For me, discretionary trading means buying when I think price is going to get more expensive before it gets cheaper, and selling when I think price is going to get cheaper before it gets more expensive. In contrast, systems trading means that a set of complete, fixed rules describe what to buy and sell, how much to buy and sell, and when to buy and sell. If all people have to do is to simply follow a complete set of rules, this is a job for software programs, so I’d just automate it and be done with it. Here’s an analogy. It might look like you could get every last drop out of an orange by squeezing it by hand, but 99% of the time you’re better off using a juicer. People are scared of leaving anything on the table, and also of tackling all the uncertainties in their “strategy”.

Systems trading the stock market is a lot tougher problem than it is for the futures and currencies markets because in stocks the domain of instruments you have to search is a lot larger and the price behavior you have to model is a lot more diverse. Another problem is that US stocks are all very highly correlated. Over a one year period, the monthly brokerage statements from my discretionary stock trading account were ~75% correlated to the nasdaq, which is 90%+ correlated to the S&P 500. This is a problem for systems traders because they are aiming to create consistency from non-correlation. On the other hand, when you are discretionary trading, the correlation helps because it gives you a lot less to worry about. You’ve only got to nail the big-picture direction over time to make money if you spread your bets conservatively. This is made easier by the US stock indices natural upward drift.

Since most system traders are doing futures and currencies, those markets are a lot tougher to trade discretionary style because it is nearly impossible to beat good software programs over time. My approach is to compete in those markets with my own software programs, where all of my work is done up front and the markets are mined over time without any intervention from me except for replacing systems as they meet retirement criteria, counting the money, and paying taxes.

My holding time can be anywhere from minutes to years. Hold time is a function of the particular strategy being deployed and the market price movement after my entry. Often times I add to or remove from an existing net position in an instrument because another completely independent strategy with a historically positive expectation kicks in. All time horizons for each trading instrument have their own unique characteristics, and they all have their place in my arsenal, especially in systems trading where each time horizon can be given the proper attention.

This is probably a good place to gripe about something that bugs the crap out of me because it has to do with my stock trading strategy. My IBD paper shows me all these stocks that have just broken out. MSN Money, CNBC, Yahoo finance, and all the brokerage firms have tools to look at the most actives, but where are are the “best inactives” lists? A lot of money can be made giving stock to all the people looking at these most actives lists featured everywhere. Even today looking at a “most actives” list gets me a little emotional. Chasing envy with money is like quicksand for your account, which is probably why these “most actives” lists are featured everywhere. So for a long time, I decided to feature a small-cap “best inactives” list each weekend on my blog. I wrote software to identify these stocks. Just looking through my computer-generated list each week to hand-pick the best ones for trading calmed me down and cancelled out any bad emotional effects from the IBD 100 and MSN’s top 10. I guess it helped other people too because it became more popular than I thought it would be.

StockTickr: What’s your exit strategy for winning and losing trades?

Jon: I pull the plug on losing stock trades very quickly. I have a large ratio of “number of realized losses” to “number of realized gains”, but that is just a function of my discretionary trading style where I’m pyramidding into winning positions, so I necessarily have fewer of them.

StockTickr: What 3 books do you recommend traders read?

Since many of your previous interviews already cover all the obvious ones, I thought it would be useful if I mention books that are outside the scope of trading, yet have influenced my trading probably more than any trading book. I consider these essential for any budding master of the universe.

Chaos by James Gleick

“Chaos” is an amazing rollercoaster ride of relevant scientific discovery. Interestingly enough, several of the scientists written about in this book are now money managers. Most importantly, for a lot of people this will be a first introduction to the “determinism” ideology that fits like a glove with systems trading. If you’ve seen Jurassic Park, then you may remember Jeff Goldblum’s character, Dr. Ian Malcolm was a chaotician. I’m pretty sure the Malcolm character was inspired by Gleick’s book.

Statistics for Dummies

Unless you’re already a stats guru, this book is a great refresher and a helpful reference for deciphering what quants describe in books and on internet forums. If any reader doesn’t understand what I just said, start googling and wikipedia searching.

The Singularity is Near by Ray Kurzweil

Almost everybody I come in contact with, including many highly intelligent and accomplished traders and software developers, have a lot of misconceptions about how technological changes shape the future. More specifically, they have ill-concieved notions about their own life expectancy and what software is ultimately capable of. Software is becoming exponentially better at managing risk, and after what I’ve seen in Ray’s book I believe that in our lifetime humans will be all but crowded out of the markets by fiercly competing software that can trade thousands of instruments simultaneously and more effectively than any human, 24 hours a day. This will happen well before the development of “strong AI” because managing financial risk is a much simpler problem. The people who control these software programs will control the world because they allocate resources more optimally than anything or anybody else. Others who don’t embrace this idea will find eventually see their fortune erode away because they won’t be able to get yield from their capital except by playing a dangerous game of chance. All the money is up for grabs sooner or later, which is why you don’t ever see estates with interest compounded over tens of centuries.

I know this is cheating to mention a fourth book, but I thought this one deserved an honorary mention:

The Richest Man in Babylon by George S. Classon

Chaos: Making a New Science Statistics for Dummies The Singularity Is Near: When Humans Transcend Biology Richest Man in Babylon

StockTickr: Other than your own, what are your 3 favorite sites?

Jon: The ones I probably get the most out of are:

StockTickr: What is your typical R value per trade? i.e. what % of your portfolio do you risk with each trade?

Jon: I like to risk about 1% of equity per trade. Using much more than that when you’re fully margined and carrying overnight risk is tempting fate. This kind of position sizing technique is pretty simplistic and less than optimal for automated strategies, but it seems to work well for margin trading stocks by hand. This gives you 10% distance to stoploss for position sizes that are 1/10 of your capital. It is easy to figure in your head, and when you’re fully margined you can have 20 to 25 different positions at once, which is about as diversified as you want to be for an actively managed portfolio. When you get into longer term strategies with 4+ years expected holding time, it makes sense to diversify out even further though. One major catch with this kind of position sizing technique is that account equity can be a fast moving target during high volatility trending periods.

StockTickr: What technical indicators could you not live without?

Jon: I use moving averages and bollinger bands quite a bit. Most indicators are pretty good, but no indicator is going to look great if you’re trying to find a holy grail. Big suprise, I know. Doing some research on the internet, I found this interview with Chuck Le’Beau from 1996. I was shocked because Chuck had come to a different conclusion than I had. In the interview, he claims to have found only one indicator that works well in all markets: directional movement (DI). So he built a system around this holy grail and traded it exclusively. I couldn’t believe he was systems trading this way, using only one system and optimizing it for all markets. So I dug further. I then found a bulletin from Chuck on his Traderclub forum that appears to have been written in 1998. In it, he talked about the deployment and eventual failure of his DI holy grail system. He then went on to describe something very similar to the systems trading conclusions that I had come to:

“Consider now the merits of multiple systems: We would obviously start with a trend following system for the big trends. But we must also have a system that works well when there are no trends. We would trade systems where the profits are taken at target points and have systems that patiently let the profits run. We could design systems for long only and short only. We could have systems with tight stops and systems with wide stops. When we were trading we would have long term, intermediate term and short term systems all running at the same time. We could even use systems that day trade.”

I’m not looking for a holy grail. I don’t care if one of my models works in other markets or not. I model price characteristics, which are subject to change. This is curve fitting, but done very carefully. The reason I can do this is because I know how to detect when a model starts to lose its predictive value or can be replaced with something better, and I trade conservatively enough that I’m not bankrupted by the layover.

StockTickr: How do you think the market has changed over the last several years? How have you adapted?

Jon: The nasdaq has been very choppy since 2004. A lot of the characteristics that people were trading, volatility, and trendiness, seem to be all but gone as the bull market matured. However, there are distinguishable trends, they just happen to be shorter in duration than before. I’ve expanded my discretionary trading arsenal to include some range trading techniques to capture this index movement.

Over the past few months, when the nasdaq weekly chart broke its long bull market uptrend, I’ve become very uncomfortable with long positions, so I’ve been doing some pairs trading to hedge the directional exposure. This summer there was a flight to quality, which I wasn’t positioned for. Traditionally I’ve focused on highly volatile smallcaps. Now, it seems that big caps are having a run. I’m not shooting the lights out this year, but my first objective is capital preservation, and only after that is secured do I worry about generating return. In the markets, as in the rest of life, cash does not easily escape a tight fist. My discretionary trading account is outperforming the nasdaq in 2006, so I can’t complain. Right now I barely have any market exposure, I’m waiting for the next big trend to come around.

StockTickr: Do you backtest? If so, what makes your approach unique? If not, how do you instill belief in your system?

Jon: My systems trading approach makes heavy use of backtesting, but is not unique. I’ve spent a lot of my spare time over the course of about 18 months writing java software to reproduce what has been described by Acrary on the forums. I’ve never spoken with him about it, but I’d like to someday because he has been an inspiration for me and probably shaved years off of my research and development time. You should try to interview him if you can get in contact with him, I’d love to read it.

StockTickr: What advice can you offer traders who are just starting out?

Jon: Realize that a profitable trade isn’t what is important. If you cannot quantify the conditions that gave you the win, then it is just a windfall. It is ignorant to think that you can build and maintain a fortune by simply moving from one windfall to the next for decades.

Be aware of whether or not your actions are moving you towards your goals. Always evaluate and re-evaluate what you are doing and what you have done. Remember that goals are subject to change, sometimes subconsciously, so be wary of shooting for moving targets.

Probably the most prevalent and plagueing problem for beginners is that they get an idea about what they want to do from looking at the broad price movement of hundreds or thousands of different instruments and then picking a few of the most favorable opportunies to enter trades. Probably more often than not they are correct in their assessment of what will happen to the broad price movement, but they sabatoge their trades because they start watching the tiny details of their trading instrument, which they would never be looking at if they weren’t intimitely affected by it. Out of 5000+ stocks, they watch every little zig and zag of the < 50 stocks that they own and none of the other ones because those are the ones that affect their P&L. In other words, decisions about entry are based on broad price movement, and decisions about exit are made on what would otherwise be irrelevent intra-day price movement. This can turn into an endless cycle where the trader can’t see that the damage is coming from a time horizon discontinuity in their strategy. I think this is a major reason why longer time horizon disciplines work well for a lot of traders, because they force their exit decisions to be made on the same time horizon as their entry decisions.

StockTickr: What do you like best about trading?

Jon: It sure isn’t the grind, so I try to minimize that part. I guess the coolest part is that when you’re trading, you’re taking part in a resource allocation optimization problem that is bigger than just human society. You’re playing a role similar to plants extracting energy from the sunlight in their surrounding environment. It affects the entire ecosystem into the future. The idea that I can tap into this process and potentially use it to finance anything I want is pretty cool.

StockTickr: Tell us about your software project you’re working on.

Jon: I’m coding a complete backtesting and trading strategies automation engine in java that is custom built to my interpretation of Acrary’s specifications. This is an ambitious project to tackle, but I’m proud to say it is my masterpiece. I even threw out months and months of code to start again from scratch when I learned about the new modeling capabilities built into the newest java version, 1.5. It took a few months to get back to where I was before, but the investment was worth it due to the modeling power we gained.

I have one collaborator, Brandon Kranz working with me on the project. It would be a weak shadow of itself today if it weren’t for Brandon. Most of the code was written by me, but Brandon has helped tremendously with research, creating and testing systems, gathering data, server administration, and most importantly for me, keeping me motivated. He just took two weeks off to go out of town and get his MCSE certification, which is a Microsoft systems administration certification. Needless to say, Brandon is the hardware and operating systems guru between the two of us.

I’m having the time of my life working on the project. Programmers usually have what is the equivalent of writer’s block, preventing them from reaching their full potential of creativite expression (not that most programmers I’ve met are creative), but this hasn’t happened to me in probably a year. Everything I’ve learned has led up to this, and it is empowering to see it all coming together so nicely.

StockTickr: Thanks, Jon!

Jon: Sure, Dave. Thanks for the opportunity!

Stay tuned – there are several interviews on the way. You can subscribe to these interviews via RSS feed.

Previous interviews in the StockTickr Interview Series (RSS feed):

Do you have suggestions for other traders you’d like to see an interview with? Let us know!

Comments are closed.