The Striker Report - gold standard for objective trading research

Pareto

Developer, Pareto Diversified (as shown on Striker's website)

Program(s) Developed: Pareto Diversified

Interviewed by John F. Gallwas, Founder of Striker Securities, Inc.

Pareto, is the name of an anonymous system developer group. The group is research driven with a focus on the design and implementation of technical trading strategies across multiple markets.

John F. Gallwas: Before we go any further, can you say why you choose not to release your identities.

Pareto: Sure, it is nothing sinister. We trade our own systems but we also design systems for various external organizations and it makes life simpler if we keep as much as possible on a "white label" basis.

Our systems with Striker are totally different to those elsewhere so inevitably, at different times, the Striker ones may be doing well whilst the others are not and vice versa. This way people just focus on which program they are in and whether they are comfortable with that, rather than trying to time continuously jumping between different systems ?which is a frustrating and generally unsuccessful endeavor anyway!

John F. Gallwas: Please tell us much as possible about the group's background and qualifications.

Pareto: There are three of us in total, two of us have worked at a variety of different hedge funds in what could broadly be described as "quant" roles. We did a quick tally and between the three of us, we have eight degrees, speak five different languages and total over 35 years of system design and other financial market experience.

We do all have experience of actual trading and the challenges that come from that, for instance I have been doing fully algorithmic trading for about a decade. In everything we do, we try to design systems to be as practical and robust as possible.

We have traded accounts at various organizations from the low thousands of dollars to the tens of millions.

John F. Gallwas: What futures markets do your systems trade and which risk statistics do you use?

Pareto: We concentrate on products that have a certain level of liquidity so that followers are subjected to as little slippage as possible. Currently there are individual systems in Euro, Gold, NatGas, Crude Oil and a couple of other commodities. The actual products and systems vary over time as we continually develop and try to stay ahead of the "degradation curve" that most trading systems tend to suffer from eventually.

We don't like some of the standard risk metrics that are so popular with the investing community, e.g. Sharpe Ratio and standard deviation. For one thing, these penalize upside volatility as much as down ?and relatively few investors complain about unexpected upside movements in their performance. These metrics also do not appear in people's account statements; it is account equity that is the ultimate performance metric.

We look at the characteristics of the results of our systems, so rather than just examining largest historical drawdown, we will run Monte Carlo simulations and see what can reasonably be expected going forward, especially if we notice some reason for past test data looking "too good".

We also are very aware of the risk at any point in time being run by the portfolio which has various terms, we call it drawdown at risk. Other than that, metrics for correlation are probably the ones we pay closest attention to.

John F. Gallwas: Describe your basic market philosophy.

Pareto: Our basic market philosophy is one that markets are mostly efficient most of the time and that there is no universal market truth to be discovered and to seek one is just the world's biggest rabbit hole.

All that being said, there are many small inefficiencies at one time or another and the trick is to keep finding as many as possible of these and see if they can be practically automated.

The key thing for us is that we have to understand the underlying driver to any rule or strategy, we don't engage in data mining.

We design our trading systems by focusing on elements that must make sense, be robust, be practical in terms of liquidity and other market factors and have a reasonable expectation of similar performance going forward as they did in the past.

John F. Gallwas: Describe how you develop and monitor your systems?

Pareto: Whilst we may use a lot of the same development processes that others do, we believe that we take greater care in our system design than is the norm for the industry. Just one example of this would be that we test all our systems through two completely separate sets of "blind" data that were not part of the initial development process, whereas one (or none) is far more common for other systems designers.

We examine our results and take a step back and try to see that they are not too correlated i.e. even with completely different code they could depend on the same market phenomena such as volatility expansion. We then look to trade as many independently to get as much practical diversification as possible for a given level of risk.

Typically, dozens of systems are rejected at some stage of the process for every one that makes it through to the point that it is live traded.

Monitoring trading systems now is the relatively simple day to day checking that the systems are switched on, positions match theoretical and the like. There are rarely any issues there now as the technology has improved greatly over time and we use a variety of fail safes and the systems sit on cloud based servers with industrial level redundancy.

John F. Gallwas: What are your expectations with any trading system Pareto develops and offers to other traders?

Pareto: It is our intention that systems we develop will be simple, practical and robust ?words that we have mentioned several times in this piece. This is important for our reputation but also as our money is trading almost all the systems we develop for others, right alongside those other people's funds.

We far prefer systems that look less exciting when developed and thus have a reasonable expectation of performing at that level or even a little better, when traded with real money.

The alternative is systems that look great back-tested and have practically no chance of that level of performance going forward ?the dreaded over optimization that is the bane of the life of system designers.

John F. Gallwas: Is there anything in "What's New" you would like to share with us now?

Pareto: We are working on a completely new program; we will keep you updated when it is ready for release!

This interview is for informational purposes only and is not intended to be a solicitation of any kind. Trade only with risk capital. The risk of trading can be substantial and each investor and/or trader must consider whether trading systems are a suitable investment.