Lincoln B. Fiske Jr.
Founder, TradingVisions, Inc.
Program(s) Developed: Multiple Trading Services
Interview Date: June, 2014
Interviewed by John F. Gallwas, Founder of Striker Securities, Inc.
Since 1999, Lincoln Fiske, Founder of TradingVisions Systems, Inc., has been a system developer. He is located in Sedona, AZ a town of 10,000 located 115 miles North of Phoenix, AZ. Diversification is one of the most important keys to successful trading, and Mr. Fiske, a well known and prolific system developer, is one of the first to recognize the need to package various trading systems for customers interested in diversification. "Vista Portfolios", launched in November, 2004, offers (1) multiple trading systems, (2) trading multiple markets, and (3) trading in multiple time frames. He has also developed a disciplined system to annually review his trading using walk-forward analysis for updates.
John Gallwas: Although you are well known to many of our readers, tell us again how you became a system developer with current offerings of six different trading systems?
Lincoln Fiske: In the early-90s I became fascinated with the idea of developing a systematic way to trade or invest and spent many hours using System Writer, TradeStation's original program. The idea that we can observe patterns, translate that into trading rules, backtest to verify and modify, and then completely automate trading is a beguiling notion. I released my first system to the public in the late 90s, and since then have released several more, as I get new ideas.
John Gallwas: What is the trading philosophy behind each system?
Lincoln Fiske: I focus on the index futures markets because of their liquidity and leverage. Three of the systems are day trading and depend on time and price patterns. EarlyBird, as the name implies, gets an early reading of market strength, and can either go with the initial trend or fade it. Sentinel also looks for a type and degree of early session directionality and goes with the trend. Impetus looks for initial strength or weakness and when appropriate enters late session trades. Spectrum is also a day trader, and it has both a trend and countertrend component that key off extremes in volatility. Delphi and AXIOM are based on proprietary channels and buy or sell with the trend depending on breakouts of the channel. Delphi has both day and swing trade capabilities, while AXIOM is swing only. In general, I believe that over-optimization, overtrading, undercapitalization, and unrealistic expectations doom many traders/investors, so I work to address those issues in the development process by having systems with few optimizable parameters, trading 1-3 times per week, recommending account sizes relative to statistical drawdown norms, and deducting slippage, commission, and lease fees from published results.
John Gallwas: Describe Walk Forward Optimization ("WFO")?
Lincoln Fiske: Historically, the typical way to create a trading system is to test all available data for a market, run thousands of optimizations to find the "best" rules and parameters, and trade the system in real-time. This is how I did it for years.
I'm a firm believer that over-curve-fitting is the bane of system development, and to avoid it, in the past I was very reluctant to re-optimize systems once released. It is just too easy to find something that works on a single market for a few years in backtesting and re-jigger it every time results are poor.
Unfortunately, there are two critical shortcomings of traditional backtested optimization: 1.) traditional hypothetical performance records are idealized (optimized), "in-sample" results that usually vary widely from real-time results, and 2.) traditional optimization has no articulated way to adapt to different markets or to changes within a market.
But what if we had a way to find out how a system is likely to perform without waiting for years while risking our money? For me, this is the exciting advantage of walk-forward optimization, and it has led to my biggest change in how I see system development.
Walk-forward optimization ("WFO") optimizes a system over a set time period from the past--say five years--and applies the "best" parameters to a subsequent time period, say one year. The performance results of the five years are "in-sample," meaning that they are derived from data (the "study period" data) that was used to formulate the system rules or parameters; in other words, the in-sample data was used to find the optimal values. These results are curve-fitted, idealized performance that is rarely matched in actual trading. In contrast, the performance on that one year after the study period is "out-of-sample," meaning that it is the result of trades that occur outside the period from which the parameters were chosen. This is also called the "application period." The performance from the application period is much closer to what would have been achieved in actual trading, since the parameters were chosen prior to the trading, and we could theoretically have been trading during that time with that set of rules and parameters. In fact, if we use realistic slippage and commissions, we then have more realistic, relevant results, second only to actual trades in giving us a picture of how the system performs. These results are categorically more trustworthy than the backtested results from a system's study period, and this is the most important advantage of WFO.
The "walk-forward" aspect of this approach comes in the next step. If I used data from 1.1.2001 to 12.31.2005 as the first study period of the analysis and 2006 as my first application year, the 2006 trading results are my first year of the hypothetical performance record. I now walk-forward a year and use the data from 1.1.2001 to 12.31.06 as my next study period, and I apply the "best" parameters from this period to 2007. These 2007 out-of-sample results are added to my 2006 hypothetical performance record. I continue to walk-forward to 1.1.2001-12.31.07, apply the best parameters to 2008, and I now have a 3-year hypothetical, out-of-sample track record that should be reasonably close to what I could have achieved in actual trading. This methodology is called "anchored" walk-forward because it stays with one start date and does not drop the earlier "study" periods of data, thereby making each study period increasingly longer.
The result of WFO is that we get to see years of very realistic trades, and by comparing the out-of-sample trading period to the in-sample period, we are able to assess whether the system held up, whether it is robust enough to handle the curveballs that the markets throw at it. This then gives us a reasonable expectation of what the future is likely to bring. In other words, WFO validates the efficacy of a system. When only in-sample data is used, system validation is based upon a circular relationship between the study data and the rules/parameters adopted: the results justify the rules/parameters, but the rules/parameters have been justified by the results. WFO takes us out of this dangerous loop and puts our expectations on a more realistic foundation.
John Gallwas: How do you use WFO for system development and maintenance?
Lincoln Fiske: I use "anchored" WFO because I want to see that a system is able to handle a wide range of market changes. The parameters that "work" are the ones that yield an equity curve with the smoothest upward performance over many years. Since the optimizations are over an increasingly longer time, the parameters tend to stabilize: what worked over the last five years will most likely be close to what will work next year, though there will often be small parameter adaptations, with WFO acting like an in-flight guidance system. Along with this, I've chosen to re-optimize on an annual basis. For me, this represents enough time for market changes to be evident. I don't want WFO to react to short term market idiosyncrasies, because as soon as you adapt to those, their time is up. I use WFO to identify and adapt to larger market themes.
John Gallwas: As one of the first developers to recognize the need for a "portfolio" approach to systems trading you offer various combinations of trading systems through Vista Portfolios. We all know that diversification is important so tell us how this program works and what markets are involved?
Lincoln Fiske: Most every investor knows that diversification is essential to earning successful returns because it's impossible for most of us to pick the fastest horse in the stable. Diversification through using low correlation investments helps to smooth the equity curve. Unfortunately, in the futures arena there's more emphasis placed on finding the fastest horse and betting it all on that. That may work once in a while, but at some point the horse gets tired, sometimes right out of the gate, and then what do you do? Many times I've seen the hot system go cold, and the cold system get hot, so I think the best approach is to invest in a team of systems that have shown strength over time and that work well together. For the Vista WFO Portfolios, I've teamed the three low correlation WFO systems (Delphi, AXIOM, and Sentinel) into portfolios with account sizes ranging from $14,500 to $57,000, starting with two components and moving up to nine. I trade Vista VI myself at Striker, and the results for the individual components can be seen at your site. Eventually I believe you'll be able to report the combined portfolio results to see how the systems and markets complement each other.
John Gallwas: Any final thoughts?
Lincoln Fiske: Thanks for the opportunity to talk with you, and many thanks to Striker for providing the public with a simple way to access trading systems.
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.