"Your WaveFin indicators, particularly the Morlet, are some of my favorite indicators. They provide great pattern recognition and generalization, as well as control of the neural net's trading pattern. For many nets, they are all one needs to generate startling returns."  --  Lawrence Weathers, Ph.D.


Who should use WaveFin?

WaveFin is ideal for traders using neural networks.  WaveFin can expose features and events in the underlying data series that neural nets can then be trained to detect and recognize.  Typically, the WaveFin filters are applied to the price series and the WaveFin outputs are then used as inputs to the neural network.  

To apply WaveFin to non-neural network based trading strategies will require a great deal of technical sophistication.  Unless you are already versed in filtering and wavelets, we encourage you to look at some of the neural network products that are available.  Our favorites are Ward Systems Neuroshell Trader and BioComp Profit.


Why Use WaveFin?

WaveFin enables traders to implement continuous Morlet wavelet filters for the purposes of noise elimination, information separation, and perhaps most importantly, feature detection. WaveFin provides users with a filter of superior consistency, accuracy and extremely high-resolution capabilites for the detection of patterns of varying frequencies and time scales.

Noise Elimination

Off-floor traders often find that removing very high frequency information enables their systems to trade more profitably. Once "noise" has been defined in the context of an application or trading system, it is relatively easy to design a filter to eliminate it: We simply define a filter that captures the noise, then subtract the captured noise from the original series. What remains is (presumably) the important information.

Information Separation

Trading signals are often hidden under worthless clutter, invisible to the naked eye as well as to most prediction models and trading systems. This situation is often best remedied by using multiple filters to separate the original series into two or more components, each of which can be examined separately without interference from other components. Most filters commonly used by traders suffer from information from one filter "leaking" into another, greatly limiting the value of the information captured by the filter. WaveFin accurately separates information with a minimum of leakage

Feature Detection

WaveFin excels at the accurate and consistent detection of events. For example, perhaps when very large market participants enter a market their presence causes some short-term price fluctuations that our model recognizes as indicative of future price changes. In these cases we may define, and then trade upon, the short to mid-term filtered information.

The Power of Local Filters

When a pattern becomes apparent, alert traders capitalize on it and thereby eliminate it. Using short-term filters helps us discover and act on short-lived periodic phenomena that appear in financial series. Event traders can use WaveFin to recognize events and profit on the subsequent price moves. Breakout traders can use WaveFin to better characterize the quality of breakouts. Cycle traders can use WaveFin to determine with accuracy and consistency the presence of market cycles.

Why is WaveFin Valuable?

The Morlet wavelet implemented in WaveFin is probably the best filter choice for feature detection in financial series:

· Morlet wavelets are naturally robust against shifting a feature in time. A feature will make itself known in the same way no matter when it occurs. Daubechies wavelets, and in fact all orthogonal wavelets present great challenges in ensuring consistency across time. WaveFin provides what matters most to real time trading - accurate and consistent information.

·A famous mathematical formula called the Heisenberg Uncertainty Principle decrees (roughly) that no filter can, with arbitrary accuracy, simultaneously locate a feature in terms of both its period and its time of appearance. In order to gain more precision in one, the other must be sacrificed. The laws of physics are quite firm here. This principle imposes a bound on how well a filter can detect a feature. WaveFin, for all practical purposes, achieves this bound. In other words, no other wavelet can do better at simultaneously locating a feature in terms of its period and when it appears. Most other wavelets do worse, and many wavelets and other filters do considerably worse. This is a very valuable property.

WaveFin and Neural Networks

There is a balance to be struck in successful neural net modeling which must provide enough model complexity to recognize complex patterns buried in mounds of noise and simultaneously strive to prevent overfitting. It can be a very tough task. Wavelet filtered variables provide an excellent form of preprocessing for a neural network model. Neural networkers and other systems modelers are highly encouraged to try WaveFin in their applications.