|
 |
|
|
THE TETRA SYSTEM TETRA (the Greek word for FOUR) uses a series of time intervals for signal generation, of which the basic 4-hourly interval provides the backbone in a time spectrum that runs form 1-minute signal generation to 12-hourly filters. Within these time frames PAC's position follow trends by way of sliding stop-loss orders. This defines the exit mechanism. For every exit, which puts the strategy "flat", a re-entry algorithm exists, on similar premises as the exit algorithm. In principle, stop orders should ideally be located where they are unlikely to be hit (e.g. far away), yet they should ideally also associate the result of being stopped-out with optimal preservation of equity. PAC uses standard deviation measures from a smoothed mean plus a randomized buffer to determine stop levels. This assumes that currency prices are random and are following a normal distribution. We know this not to be the case, since currency prices generally display a leptokurtic behaviour, with narrow peaks and fat tails. On the other hand they do behave quasi-normal most of the time, and the TETRA strategy exploits this premise. Trading signals occur according to strictly defined rules, which are programmable as well as teachable. The algorithms used in the TETRA system have been devised by Peter Panholzer and are the result of the observations and experience gained during his 25-years in currency portfolio management. The algorithm is proprietary and cannot be discussed in further detail than above. |
|
PAC trades only two major currency pairs, EURUSD and USDJPY, and this for a very good reason: they are the most liquid markets. In an environment that requires new trend-following techniques and more frequent trading, the lower execution cost offered by the extremely high liquidity of EURUSD and USDJPY becomes a crucial performance factor. No doubt currency prices have become more random with the proliferation of electronic trading. This has caused a deteroriation of returns after a new performance high was made on March 31st, 2004. Five negative months ensued causing us to re-evaluate the Tetra system parameters. The first improvement introduced a profit target in both currency pairs (EURUSD and USDJPY). Testing this measure over 15 years proved satisfactory enough to use this approach from September 1st, 2004 onward producing three winning months. See TETRA2 PERFORMANCE GRAPH. Yet one other concern, voiced by some of our clients, involved the distance of stop-loss orders. Being based on two standard deviations plus a buffer, they resulted on occasion in losses of 250 points or more such as one loss causing the severe setback of December 2004 when the EURUSD underwent a huge whipsaw. Introducing a lower fixed loss limit on all trades resulted in the current TETRA3 SYSTEM, our solution for 2005 and hopefully a good number of years ahead. The recent setback provides an opportune entry point. |
|
|
|
|
|
