In our previous articles, we discussed what a moving average is, the different types of moving averages, and simple yet effective ways to use them in trading. In this article, we will explore more advanced or alternative methods of utilizing moving averages.
There are numerous approaches to using indicators, particularly moving averages, in technical analysis. However, the most effective approach is to develop your own system. In this article, we aim to provide various perspectives to help you generate your own ideas.
Moving Linear Regression – Moving Average Cross
Linear regression is typically applied within a confined range of bars alongside standard deviation lines to assess the prevailing trend and its deviation. Similar to averages, it can be adapted into a moving form referred to as moving linear regression by considering the most recent points of the period. Moving linear regression behaves akin to moving averages but tends to align more closely with the trend direction, albeit often with a delayed yet sharper response.
This characteristic of moving regression renders it well-suited for replacing the slow moving average in a crossover strategy. It helps mitigate premature false crossovers while exhibiting a more decisive response to shifts in trend direction. In the example below, a 21-day exponential moving average serves as the fast MA, while a 100-day moving linear regression acts as the slow MA. As illustrated, horizontal consolidation periods still pose challenges for the MA crossover strategy, but the utilization of moving linear regression significantly reduces the impact of false attempts at trend changes.
Z-Score – Returning to Average
Using moving averages in trends is easy. Many different methods work rather well. But what should traders do when the price is ranging with no clear trend? Can traders use moving averages in horizontal or even whipsaw charts? There is one easy way, and it is from the principle that prices tend to return to their average. Almost all sharp moves end up with the price returning to its moving average, and it might be a good idea to capture these extreme sharp moves and use them as buying or selling opportunities. Usually, the extent to which the distance deviates from its moving average differs from security to security, depending on factors such as the period of the moving average, volatility, volume (low volume securities often experience sharper moves), and type (currency, stock, crypto, etc.). It is up to the trader to discern if there is a possible pattern.
For traders to utilize the deviation, they should first decide how to calculate the distance from the moving average. The first and most obvious method is simply taking the difference between the price and the moving average. However, this might not be a good idea in the long term. For example, if we are looking at gold when the price was at $1000, and let’s say when the price differs from our average by a little more than $50, it tends to reach a top or bottom. But what should we do when the price reaches $2000? Shouldn’t we adjust the $50 rule? How many times should we adjust our rule on the road from$1000 to $2000? The second method solves this problem by using a percentage of the price to calculate deviation. But sometimes volatility drops for months, and sometimes it increases, creating big whipsaws. Should we disregard the volatility factor? That might result in no signals for a prolonged period or perhaps too many false signals when the volatility is high.
With the third option, both problems are solved, at least mostly. The third method is called the Z-Score. It uses standard deviations to calculate how much the price deviates from the moving average. In the example below, we used an hourly silver chart with a 200-hour moving average and used 2.5 standard deviations as our indicator. Whenever the Z-Score passed the 2.5 deviation mark in either direction, it often marked the current extreme and a local top or bottom. But unlike MA crosses, this strategy won’t work well during trends. As seen in late November, it marked two false tops.
Moving Average Slope
Moving average is a valuable tool for trend-following, but price-MA crosses often prove ineffective due to numerous false signals, particularly during horizontal ranging periods. This means that a significant portion, if not all, of the profits may be eroded by these false signals. One potential approach to mitigate this issue is by incorporating the slope of a moving average.
The rationale behind this is that when the slope of the moving average crosses the zero line, it indicates a local top or bottom. As illustrated in the example below, the tops and bottoms of the moving average occur later than those of the price. This helps prevent our strategy from falling for many local tops and bottoms, thereby reducing the occurrence of false buy or sell signals.
However, a drawback of using the moving average slope is the occurrence of small consolidation zones within trends. In the first rectangle, after the formation of a local top, the consolidation period before another wave extends too long. This leads to the moving average changing its slope around zero multiple times, resulting in numerous false signals before resuming an upward trend. Similarly, in the second rectangle, during a downtrend, a consolidation period lasting longer than a week generates many false signals, albeit fewer than those produced by the price-MA cross strategy.
For non-algorithmic use, experienced traders can easily identify these false signals of moving average slope. However, for algorithmic traders, incorporating two lines slightly below and above the zero line may significantly reduce these false signals.
Moving averages are indispensable in technical analysis, serving to identify trends to capture extreme price movements. Additionally, many oscillators can be enhanced by incorporating moving averages, such as the RSI and MACD. However, the most advantageous aspect of moving averages is their direct reflection of price data; thus, any price action is mirrored in the moving average.
Experimenting with different approaches to utilizing moving averages or integrating them with other tools can lead to the development of personalized trading strategies. The TradingView codes for moving linear regression and moving average slope will be made available upon request to FTD Limited traders.
