AT: What were your trading experiences in that period from 1973 to 1979? How long did it take you to develop your systematic approach?
PP: It took me from 1973 to 1977 that’s when my trading blossomed and I started to make oodles of money. It takes four years of tuition in any endeavor in life. If you want to learn golf, tennis, piano, a language — it always takes at least four years, doesn’t it? But before that I didn’t make money — I made it, I lost it. I survived. But that’s normal. Whatever I lost, I considered tuition.
AT: How did you initially approach the markets?
PP: I’ve been a contrarian since 1972. The figures published by Market Vane (Sibbet-Haddady) started in 1972, and ever since then I’ve been following them. But then to apply that information, that takes insight, too. So when I say it took me four years of learning, that includes learning contrarianism. All the money I made before I started system trading was through contrarian trades.
AT: What’s an example of the kind of trade you might have used in this regard?
PP: Take sugar, for example. After running up from 6 cents to more than 60 cents throughout the early 1970s, it had just collapsed back to 6 cents. Everyone considered it a bargain, of course.
Well, contrarian thinking taught me to go against “everyone.” I went short at 6 cents and everyone thought I was nuts — until I cashed in my profits at 4.5 cents. In relative terms, a move from 6 to 4.5 is huge. Nobody thought that way, but it’s a 25-percent drop, after all.
The following winter (1978) I holidayed in Mexico City, but, in true Jesse Livermore fashion, instead of visiting monuments I hung around brokerage houses and got to talk to a Mexican client who “confided” to me that his brokerage had lent him the required additional margin for his short cocoa contracts, which were then trading at 89 cents/pound. “We don’t want you to close out your position and let some gringo profit from your losses,” they had told him.
To me that sounded like a starting pistol. I immediately called my partner in Toronto and asked him to go long six contracts of cocoa in my account. It was a bumpy ride all the way up to above $2 when I took my profit.
AT: What ideas did you go through in that period that you ended up casting aside?
PP: Econometrics — what we might call one step ahead of fundamentals — didn’t work for me. For example, at our firm we had a curve that showed the average weekly movement of pork bellies in and out of warehouses. That has a seasonal pattern — it goes up around Easter, down in the summer and increases in the winter. You can average that over 20 years and get a smooth curve.
Then I compared the actual movements against the 20-year average to determine what that would mean in terms of a move in pork belly prices. That idea makes a lot of sense, but it doesn’t work. This is what led me to the numbers game.
AT: Is “the numbers game” what most people would consider technical, quantitative or statistical?
PP: Completely statistical. I do not believe in technical indicators per se, although some people use them as a tool — and there’s nothing wrong with that. Let’s say you have a hammer. If this hammer suits your hand particularly well and you can hit a nail right on, that’s a good hammer for you. For me, system trading is a good hammer.
AT: Is it fair to say that whatever you use has to be testable?
PP: Certainly. A lot of research goes into the stability of parameter spaces — which are ranges of system parameter values across which you want performance to be consistent as markets evolve.
But I think the most important thing I’ve learned in the past 25 years is that if you delegate trade execution to someone else, you’re more likely to succeed. I’m purely systematic. We now have what we call a “robot interface” with one broker. That account has no interference whatsoever. No resting orders are placed in advance, so the broker doesn’t know the robot’s “intentions.”
We prefer not to show our orders. I believe in that in principle. If a broker sees the order, there’s always the argument that brokers [execute] where the orders are bunched up, for example.
With the robot, the order sits on our computer server and the broker doesn’t see it. The moment our server sees the trade trigger, it executes at the market with the electronic broker platform.
AT: That sounds like a good idea for individual traders, too: Instead of placing limit orders, place the trade as a market order when you get a signal from your system.
PP: Yes, of course. All I can say is, in all likelihood, this is the best way to execute a system. Second guessing it is the biggest danger.
AT: What’s the TETRA program?
PP: It’s the program applied in most of the funds we manage. We started it in January 1998.
AT: Is it a trend-following approach? How does it work?
PP: It’s a hybrid system. A trend-following system normally has some kind of sliding protection something that keeps moving with the market; that’s how you follow the trend. We have a hybrid because we use both profit targets and sliding protection, in a systematic way.
We use four-hour (the primary time interval), hourly and daily data in the TETRA program. These intervals are not being changed. Some researchers, such as Richard Olsen (see “Richard Olsen: Tuned to high frequency,” Currency Trader, Oct. 2004 p. 42), have found you can use “tick time” that multiple hours of trading activity can be compressed into one hour, for example
A system consists of three things:
Entry mechanism, exit mechanism and then re-entry mechanism. If you don’t have a re-entry mechanism, you have no system.
AT: How did you determine these time intervals? How long a process was it?
PP: It was a process covering years of research and hard toil.
Subscribe to:
Post Comments (Atom)
Post a Comment