How have noted trend followers Bill Dunn, Jerry Parker, John W. Henry, Ed Seykota, etc. pulled millions from the market? They plan their trades and trade their plans. Think about it.
Keep It Simple
“We could still imagine that there is a set of laws that determines events completely for some supernatural being who could observe the present state of the universe without disturbing it. However such models of the universe are not of much interest to us mortals. It seems better to employ the principle known as Occam’s razor and cut out all the features of the theory which cannot be observed.”
Stephen Hawking
What is Hawking driving at? Trading rules must be easily explained. All great trading concepts avoid layers and layers of interconnected rules. Tradestation might invent invent hypothetical profits, but is this wise? Occam’s razor dictates that if there are two trading methods that achieve similar results, choose the simpler method. Period.
Change
“Another lesson I learned early is that there is nothing new in Wall Street. There can’t be because speculation is as old as the hills. Whatever happens in the stock market today has happened before and will happen again.”
Jesse Livermore
Doubt Livermore? You think 2003 is somehow different? Here are the hot tech stocks of 1968:
| Company | 1968 High | 1970 Low | % drop | P/E at High |
| Fairchild Camera | $102.00 | $18.00 | -82% | 443 |
| Teledyne | 72.00 | 13.00 | -82 | 42 |
| Control Data | 163.00 | 28.00 | -83 | 54 |
| Mohawk Data | 111.00 | 18.00 | -84 | 285 |
| Electronic Data | 162.00 | 24.00 | -85 | 352 |
| Optical Scanning | 146.00 | 16.00 | -89 | 200 |
| Itek | 172.00 | 17.00 | -90 | 71 |
| University Computing | 186.00 | 13.00 | -93 | 118 |
The names change. The year changes. The market might change. Trend following using price data never changes. More.
What the 1968 Table Proves
Livermore’s statement that there is nothing new in Wall Street is not nostalgia. It is a structural observation about human behavior in speculative markets. The 1968 technology stocks in the table above had P/E ratios of 42, 54, 118, 200, 285, 352, and 443 at their highs. The Nasdaq bubble of 1999-2000 produced comparable or higher valuations on comparable or newer technology themes. Both cycles ended with 80-93% declines. Both cycles were driven by the same combination of genuine technological advancement, extrapolated earnings expectations, and the momentum of capital chasing rising prices. Both cycles left investors who held through the peak with devastating losses.
A trend following system would have been long Fairchild Camera on the way up in 1968. It would have exited when the price broke below the defined exit level on the way down. The P/E ratio at 443 was not an input to the exit decision. The price was. When the price trend reversed, the exit signal fired. This is identical to how a systematic trend following approach would have handled the Nasdaq names in 2000. The story changes. The mechanism does not.
Hawking’s Occam’s razor argument applies to both the 1968 and the 2000 bubble. Analysts developed elaborate models for valuing technology companies at those P/E levels. The models became progressively more complex as the valuations became progressively more difficult to justify with straightforward analysis. The simpler model, price is going up so be long, and price is going down so exit or go short, required no elaborate theoretical framework. It required only price data and a rule. Occam’s razor selects the simpler model when the results are comparable. Trend following is that simpler model.
The Tradestation hypothetical profits reference is the practical application of Occam’s razor in system evaluation. A complex system that produces impressive backtested returns is not necessarily better than a simple system that produces similar backtested returns. The complex system has more parameters, which means more opportunities for curve-fitting to historical data. The simpler system has fewer parameters and is more likely to reflect genuine market structure rather than historical pattern-matching. When backtested results are comparable, the simpler system is the better choice.
Frequently Asked Questions
What does Occam’s razor mean for trading system design?
It means that when two trading approaches produce similar results, the simpler one should be preferred. Complex systems with many parameters have more opportunities to fit historical data without reflecting genuine market structure. Simple systems with fewer parameters are more robust to new data because they have fewer degrees of freedom to overfit. The great systematic traders, from Donchian’s weekly rule to the Turtle system, built approaches that could be explained in a paragraph rather than a manual.
Why does Livermore say there is nothing new in Wall Street?
Because the human psychology that drives speculation does not change between cycles. The 1968 technology stocks and the 2000 technology stocks were different companies in different decades, but the mechanism was identical: genuine technological advancement attracted investor attention, extrapolated earnings expectations drove valuations far above any justified fundamental level, price momentum attracted more capital, and eventually the prices collapsed 80-90%. The names and years changed. The pattern did not.
Why does trend following using price data “never change”?
Because it responds to price movement rather than to the specific story behind the price movement. The rule that buys when price breaks above a defined level and sells when it breaks below a defined level does not need to be updated between the 1968 technology bubble and the 2000 technology bubble. It responds to whatever price is doing in whatever market is trending. The stories that produce the trends are endlessly varied. The price movements they produce are structurally similar enough for the same rules to capture them.
Trend Following Systems
Want to learn more and start trading trend following systems? Start here.
