SX Wealth | Crypto: Model Descriptions

The following is a description of SX Wealth Crypto’s technical trend-following models.  

VaMA Two Moving Average Crossover Daily

The VaMA Two Moving Average Crossover model is designed to perform in a way similar to a traditional two moving average crossover approach that uses simple moving averages, but with fewer false signals due to the use of moving averages that adjust to the market’s volatility. This is an intermediate-term trend-following model, meaning that it will generate several trades in a calendar year, rather than several trades per month.  

The VaMA Two Moving Average Crossover model performs similarly to a traditional two moving average crossover approach that uses simple moving averages, but with fewer false signals due to moving averages that adjust to the market’s volatility. This is an intermediate-term trend-following model, meaning that it will generate several trades in a calendar year rather than several trades per month.

When market volatility decreases and the market is not trending, our VaMA tends to drift sideways and generate fewer false signals. In contrast, a simple moving average continuously cycles up and down, generating a series of fake-outs from false indications of new trends. Because of this, our two moving average crossover system using VaMAs produces fewer false signals by crossing less often in a trading range yet exhibits beneficial sensitivity by crossing quickly when a new trend starts, and volatility remerges in the market.

VaMA Two Moving Average Crossover Weekly

This model functions like the Daily version previously described, except it operates on weekly rather than daily data.

VaMA “Turn” Daily

The beauty of the VaMA is that it is smooth and doesn’t often turn (reduced false signals). However, when it does make a turn, it usually means something.  This model is designed to generate buy signals when the VaMA turns up.  This intermediate-term model also signals several trades per year and will generate more trades than the VaMA Two Moving Average Crossover model.

It’s consistent at picking up the initial impulse of an emerging trend, and less likely to get faked out when the price is drifting up and down in a more meaningless fashion.

VaMA “Turn” Weekly

This model functions like the Daily version previously described, except it operates on weekly rather than daily data.

Volatility Breakout Daily

The Volatility Breakout model signals trades when the market cycles from periods of quiet inactivity (low volatility) to periods of renewed explosive momentum that is often the initial launch of a new trend.

The trick to building a model like this is to determine the threshold beyond which the market is exhibiting enough volatility to warrant an entry signal (and avoid false signals) yet is not too hard to overcome, resulting in missed opportunities or entries that are too late. 

Sovereign X has a unique approach to modeling volatility for trading model purposes. 

Traditional, standard volatility approaches, such as using periodic average true ranges or measuring the standard deviation of market prices, are based on the assumption that the market in question behaves in a “normal” way.  (We’ll not get too deep into statistics here). But market prices for many markets, especially those that exhibit high volatility (like crypto assets), don’t fit neatly or nicely into statistical models that assume normality.  

Yes, statistical theory says that data exhibiting non-normal behavior over a large sample size will show normal behavior with a small sample size. But in our measurements of historical data for crypto assets such as Bitcoin, we see non-normal behavior even in small sample sizes (e.g., for the statistically minded, Bitcoin exhibits a positive skew even in small samples sizes). This means that we are better off making adjustments to models to account for the non-normal nature of volatility trading vehicles like crypto assets.

Volatility Breakout Weekly

This model functions like the daily version previously described, except it operates on weekly rather than daily data.

Channel Breakout Shorter-Term

The Channel Breakout model generates signals when the market breaks out from a previous range. This is a historically time-proven approach for volatile markets that exhibit significant and sustained trends.

The most famous application of channel breakouts is probably the system introduced by Richard Donchian many decades ago, which was then adapted and taught to the world-famous Turtles by Richard Dennis and William Eckhardt. We won’t take up space here by delving into the backgrounds of these legendary traders and their acolytes (the Turtles), but we strongly recommend you do so.

On a side note, one of the core principles of trend-following for both Donchian and the Turtles is market diversification.  Follow this link for our thoughts on diversification and trading a single market. 

Our model does not precisely replicate the Donchian or Turtles approach.  Instead, we performed our own testing to “tune” our channel breakout system to the character and personality of crypto assets.  This model is a shorter-term version that generates several trades per year.  We also employ a longer-term version.

Channel Breakout Longer-Term

This Channel Breakout model operates in the same manner as the shorter-term version. Still, it uses a different parameter set with wider channels and thus will generate fewer trades per year.

Channel Oscillator Daily

Ironically, the name of this model aptly describes its original purpose. It is an oscillator designed to indicate the market’s current position within the short-to-intermediate term price channel. The verb “oscillate” means to “swing back and forth” and to “vary in magnitude in a regular manner around a central point.”

As you may have observed, tradable markets, including crypto assets, don’t move in a straight line. Even if prices generally move in one direction, such as higher (an uptrend), they swing up and down or, as some would say, they “zig-zag” up and down while prices on average move higher. In this case, the average is the central point (moving upward in this example) around which the market price is swinging back and forth.

This model was initially designed to indicate the high and low points of these “swings” or “zig-zags” within the overall high-to-low channel of price movement.

Note, however, that high points are positive values and low points are negative values. After a few years of observing this model in real-time, we noticed that when prices are in an uptrend, the values of the Channel Oscillator are usually positive. Conversely, when prices are in a downtrend, the values are generally negative. This observation spurred us to test an idea of using this model for an additional purpose – as a trend-following model.

And after some experimentation and testing, we were able to add a factor to the underlying algorithm so that the model generates valuable uptrend and downtrend signals. These are now plotted as +1 (uptrend or “buy”) and -1 (downtrend or “sell”) signals on the chart. The trend signal from the daily Channel Oscillator has been added to our portfolio of trend-following models that feed our Buy and Sell Gauges.