IndexTrader - Background
A short-term 100% mechanical index trading model.

SPI • Nikkei • Taiwan • Hang Seng • Dax • Ftse • Nasdaq • SP500
 

The Story Behind IndexTrader

IndexTrader was originally called SpiTrader.

I released SpiTrader to the public in April 2001 and officially renamed the model IndexTrader in April 2003.

The reason why I renamed the model IndexTrader is an interesting story, and one that Traders will be pleased to learn as it goes to support the robustness of the model's simple methodology.

In a nutshell, SpiTrader was designed to trade the Australian SPI index futures contract, using the SPI's daily prices as its test data sample. Following numerous emails over the first couple of years enquiring whether SpiTrader worked on other indices (generally people wanted to be assured that SpiTrader was robust and not too "curve fitted" to the SPI®) I decided to collect the futures data and run SpiTrader over the Hang Seng, FTSE, DAX and SP500 contracts. To my delight SpiTrader made money (hypothetically) on all these markets!

After I settled down following this discovery I decided to look at how each individual pattern performed on the international indices. What it revealed was that half the patterns performed well across all indices while the other half only worked on the SPI. This led me to believe the non-performing patterns were uniquely SPI orientated.

I then asked myself the question, if I had the choice between trading patterns that worked across multiple markets, or trading patterns that only worked on a unique market....which patterns would I prefer to trade?

It didn't take me long to realise I would prefer trading patterns that were robust across multiple markets, especially if those indices were "out-of-sample" contracts. 

Consequently I decided to remove four patterns, add four markets and rename SpiTrader IndexTrader to both reflect it's robustness over out-of-sample global indices and to give the model a wider appeal to international Index Traders.

SpiTrader's transformation into IndexTrader validates the robustness of its simple methodology.  

You see..... the majority of trading systems are developed with the benefit of observing the particular market/s they're designed to trade. Traders look backward, observe a particular behaviour and then design a system to "capture" that behaviour. Without "observing" a market a Trader would not be able to develop an idea to test. This is what people refer to as "curve fitting" a trading idea to past data. Consequently the majority of systems are "curve fitted" as they're developed with the benefit of hindsight. And of course what is important for Traders/System Developers is to avoid the trap of using too many variables with "tweaked" parameters to fit their systems to historical data.

(As an aside Traders considering IndexTrader will be pleased to know that IndexTrader DOES NOT contain any parameters that can be played with!)

In addition, what is extraordinary and therefore pleasing for Traders who are considering IndexTrader, is that the model was designed without any "curve fitting", or as I described above, without observing ALL the particular market/s it trades. This is because IndexTrader was originally designed to trade the Australian SPI futures contract without any observation or studying of global indices (Hang Seng, FTSE, DAX  and SP500) it now trades.

The fact that IndexTrader/SpiTrader had without design, objective or purpose ALSO identified high probability repetitive patterns in other indices, provided me with a world's best practice out-of-sample test and validation of the robustness of IndexTrader's/SpiTrader's simple patterns and trade plan! Phew...what a mouth full!

So as you can see the story behind IndexTrader is that it came about by accident and not by design and the benefit of this is that it provides an out-of-sample test and validation of the model's simple, yet robust methodology.