Trading Strategy Guide [How to Use Smoothing]

This RoninAi Smoothing Guide Includes:

AI Smoothing: How it works

RoninAi offers analytics to both short-term and longer-term cryptocurrency investors. Smoothing techniques are used in many industries to determine the underlying trend. For example, the majority of technical analysis indicators in trading apply some kind of a smoothing technique to represent a longer-term view of a given financial asset. Based on the feedback from its users RoninAi team decided to introduce the most effective smoothing techniques to add flexibility to the longevity of AI predictions. Three smoothing techniques of moving average, exponential moving average, and relative Mstrength index have been applied generating 9 unique trading strategies that dramatically outperformed the HODL in both bullish and bearish markets.

Moving Average Smoothing

Simple moving average smoothing is one of the most widespread techniques out there. Simple moving average takes the average of the previous terms in time series and consistently rolls them over. Similar to linear regressions moving averages provide linear data approximations. The nitty-gritty of artificial intelligence and its effectiveness to identify hidden patterns lies in its ability to deal with non-linear relationships. Linear approximations for such non-linear relationships on shorter-time frames work excessively well.

RoninAi’s Moving Average Smoothing computes short-term and long-term averages of the most recent AI predictions. The averages are then subtracted from one another and the difference is normalized into a prediction number between 0 and 100 percent. 100 percent indicates that AI is very confident that a given cryptocurrency will go up in the next timeframe, while 0 percent indicates that AI is very confident that a given cryptocurrency will not go up in the next timeframe (meaning AI is pretty confident cryptocurrency will go down).

It is fair to assume that predictions above 50 percent are more bullish, while predictions below 50 percent are more bearish. However, it makes sense to look for higher accuracy entry point of, say 70 percent, when considering purchasing a given cryptocurrency. Vice versa is true too. At the screenshot below RoninAi forecasts are quite bullish on ETH (81.7%), XRP (100.0%), LTC (94.4%), and BCH (100.0%), while BTC at 45.9 percent should be viewed as a bearish prediction.

RoninAi data analysts has also included three great trading strategies that have been dramatically outperforming the HODL. These strategies are called MAS-X, MAS-XS, and MAS-XX:

– MAS stands for moving average smoothing;

– X, XS, and XX stand for the frequency of trades and risk/reward ratio. X represents lower risk trading signals based on moving average smoothing and, therefore, generates lower number of trading signals. XX, on the other hand, typically generates trading signals more frequently yet each signal bears higher level of risk. XS is somewhere in between: moderate level of risk coupled with medium frequency of trading signals.

Thus, MAS-X, MAS-XS, and MAS-XX offer great flexibility for both active traders and risk-averse investors.

When it comes down to effectiveness of RoninAi predictions results speak for themselves. For example, MAS-X, MAS-XS, and MAS-XX all resulted in +35 percent return in a little over 3 months while BTC HODL finished nearly flat.

Similarly, despite severe bearish market in Ether since June 2018 RoninAi algorithms outperformed ETH HODL dramatically. While Ether dropped by more than 50 percent, moving average smoothing strategies resulted in +20 percent profit.

Litecoin is yet another example of how effective RoninAi algorithms are. When LTC dropped by 30 percent in less than 4 months RoninAi moving average smoothing strategies delivered returns of 18 percent and above.

How to trade MAS-X, MAS-XS, and MAS-XX? Tables below summarize entry and exit points as well as general analytics on every moving average smoothing strategy.

MAS-X Strategy: How to use it

Moving Average Smoothing strategies are very straight-forward. Tables presented in this section provide optimal entry and exit points. For example, MAS-X entry and exit points work in the following manner. When moving average smoothing AI prediction for BTC breaks above 99 percent it is time to purchase BTC. Once the position is established it should be held until smoothing AI forecast for BTC drops below 50 percent. Since end of June MAS-X strategy generated 61 signals with a total cumulative return of 41.1 percent and Sharpe ratio of 2.74. In a little over three months, MAS-X strategy applied on BTC outperformed BTC HODL by more than 36 percent. MAS-X strategy statistics for BTC and other 4 cryptocurrencies is presented in the table below.

MAS-XS Strategy: How to use it

MAS-XS works very similarly to MAS-X. While exit points are the same, the entry points are lower, thus generating more signals and increasing the risk from low level to moderate. For example, when moving average smoothing forecast for ETH breaks above 95 percent it is time to purchase ETH. When the long position is established and the moving average smoothing forecast for ETH drops below 70 percent the signal is considered to be closed and it is time to sell ETH. In case of ETH, MAS-XS algorithm grossed 21.5 percent return since June 30, 2018 with a Sharpe ratio of 1.78 and 59 signals. On the other hand, ETH HODL dropped by 54.1 percent within the same timeframe. MAS-XS strategy statistics for ETH and other 4 cryptocurrencies is presented in the table below.

MAS-XX Strategy: How to use it

Entry and exit points as well as statistics of MAS-XX strategy effectiveness are presented in the table below. For example, when moving average smoothing forecast for LTC break above 90 percent it is time to purchase LTC. When the long position is established and the moving average smoothing forecast drops below 50 percent the signal is considered to be exhausted and it is time to sell LTC. Statistics for LTC and other 4 cryptocurrencies is presented in the table below as well. In case of LTC, MAS-XX strategy generated 63 buy/sell signals with total cumulative profit of 21 percent and Sharpe ratio of 1.56 when LTC HODL strategy showed a huge loss of 27.2 percent within the same timeframe of just a little bit over 3 months.

Exponential Moving Average Smoothing

Exponential moving average smoothing is another popular smoothing technique used in a number of technical analysis indicators. Exponential moving average (EMA) is a weighted moving average that gives more weighting, or importance, to recent AI forecasts than the simple moving average does. Non-linear nature of hidden patterns in cryptocurrency markets could be more flexibly caught by EMA smoothing strategies. EMA smoothing strategies could add value during trend reversals and above-average volatility periods when linearization of AI forecasts could hold certain prediction discrepancies.

RoninAi’s Exponential Moving Average Smoothing generates both short-term and long-term averages of the most recent AI predictions. Difference of two averages is normalized into a prediction range between 0 and 100 percent. 100 percent indicates that AI is very confident that a given cryptocurrency will go up in the next timeframe, while 0 percent indicates that AI is very confident that a given cryptocurrency will not go up in the next timeframe (meaning AI is pretty confident cryptocurrency will go down).

It is fair to assume that predictions above 50 percent are more bullish, while predictions below 50 percent are more bearish. However, it makes sense to look for higher accuracy entry point of, say 70 percent, when considering purchasing a given cryptocurrency. Vice versa is true too. At the screenshot below RoninAi forecasts are very bullish on LTC (100.0%), neutral on XRP at 44.7 percent, bearish on BTC at 18.1 percent and very bearish on ETH and BCH: 4.3 percent and 0.0% percent respectively.

RoninAi data analysts has introduced three comprehensive trading strategies that have been dramatically outperforming the HODL. These strategies are called EMAS-X, EMAS-XS, and EMAS-XX:

– EMAS stands for exponential moving average smoothing;

– X, XS, and XX stand for the frequency of trades and risk/reward ratio. X represents lower risk trading signals based on moving average smoothing and, therefore, generates lower number of trading signals. XX, on the other hand, typically generates trading signals more frequently yet each signal bears higher level of risk. XS is somewhere in between: moderate level of risk coupled with medium frequency of trading signals.

Thus, EMAS-X, EMAS-XS, and EMAS-XX offer great flexibility for both active traders and risk-averse investors.

EMAS could be reacting much better to market moves driven by real supply and demand. On the other hand, authentic market moves driven by unusual market behavior might be avoided, whether these moves are positive or negative. For example, BCH reacted to exponential moving average so well that despite severe bearish market EMAS strategies outperformed the HODL by almost 80 percent in under 4 months.

Conversely, XRP rally has not been detected by RoninAi algorithms. It happened either due to the market pattern that has not have happened prior or due to unusual market behavior. It is worth noting that before XRP rally took place EMAS strategies have been outperforming the HODL despite significant selloff.

Examples of BCH and XRP indicate that EMAS strategies work excessively well for market moves driven by authentic forces of supply and demand, while questionable market behavior might not be caught by exponential moving average.

More established cryptocurrencies such as BTC and ETH should be reacting to EMAS with much higher level of predictability. Despite the fact that BTC was trading sideways, EMAS strategies resulted in 28 percent of profit and higher in under 4 months.

How to trade EMAS-X, MAS-XS, and MAS-XX? Tables below summarize entry and exit points as well as general analytics on every exponential moving average smoothing strategy.

EMAS-X Strategy: How to use it

Exponential Moving Average Smoothing strategies are sufficiently straight-forward. Tables presented in this section provide optimal entry and exit points. For example, EMAS-X entry and exit points work in the following manner. When exponential moving average smoothing AI prediction for BTC breaks above 99 percent it is time to purchase BTC. Once the position is established it should be held until smoothing AI forecast for BTC drops below 70 percent. Since end of June EMAS-X strategy generated 27 signals with a total cumulative return of 28.1 percent and Sharpe ratio of 1.99. In a little over three months, EMAS-X strategy applied on BTC outperformed BTC HODL by almost 25 percent. EMAS-X strategy statistics for BTC and other 4 cryptocurrencies is presented in the table below.

EMAS-XS Strategy: How to use it

EMAS-XS works quite similar to EMAS-X. While exit points are the same, the entry points are lower, thus generating more signals and increasing the risk from low level to moderate. For example, when exponential moving average smoothing forecast for ETH breaks above 95 percent it is time to purchase ETH. When the long position is established and the exponential moving average smoothing forecast for ETH drops below 50 percent the signal is considered to be closed and it is time to sell ETH. In case of ETH, EMAS-XS algorithm outperformed the HOLD by almost 50 percent. Despite the fact that EMAX-XS strategy produced a loss of 8.3 percent between June 30 and October 15 it is important to understand that EMAS strategies as a whole work much better in times of authentic trend reversal and above average volatility. It is fair to assume that once markets start gaining volatility and the trend starts reverting EMAS will outperform the HODL. EMAS-XS strategy statistics for ETH and other 4 cryptocurrencies is presented in the table below.

EMAS-XX Strategy: How to use it

Entry and exit points as well as statistics of EMAS-XX strategy effectiveness are presented in the table below. For example, when exponential moving average smoothing forecast for BCH break above 90 percent it is time to purchase BCH. When the long position is established and the exponential moving average smoothing forecast drops below 70 percent the signal is considered to be exhausted and it is time to sell BCH. Statistics for BCH and other 4 cryptocurrencies is presented in the table below as well. In case of BCH, EMAS-XX strategy generated 27 buy/sell signals with total cumulative profit of 40.9 percent and Sharpe ratio of 2.31 when BCH HODL strategy showed a huge loss of 40.6 percent within the same timeframe of just a little bit over 3 months. Thus, EMAS-XX outperformed the HODL by more than 80 percent in case of BCH due to significant authentic trend reversal that occurred between June 30, 2018 and October 15, 2018.

Relative Strength Index Smoothing

Relative strength index smoothing technique is used in one of the most popular technical analysis indicators of all time, RSI. RSI-based smoothing measures the speed and change of AI predictions and it is oscillating between 0 and 100 percent. RSI smoothing is great for detecting the general trend. Areas when RoninAi forecasts are pointing out to excessive buying or selling of cryptocurrencies is reflected in the RSI smoothing quite well. The beauty of RSI smoothing lies in the fact that it has non-linear nature of predictions which is perfect for non-linear patterns observed in cryptocurrency markets.

RoninAi’s Relative Strength Index Smoothing generates both short-term and long-term averages of the most recent AI predictions. Difference of two RSI averages is normalized into a prediction range between 0 and 100 percent. 100 percent indicates that AI is very confident that a given cryptocurrency will go up in the next timeframe, while 0 percent indicates that AI is very confident that a given cryptocurrency will not go up in the next timeframe (meaning AI is pretty confident cryptocurrency will go down).

It is fair to assume that predictions above 50 percent are more bullish, while predictions below 50 percent are more bearish. However, it makes sense to look for higher accuracy entry point of, say 70 percent, when considering purchasing a given cryptocurrency. Vice versa is true too. At the screenshot below RoninAi forecasts are very bullish on LTC (99.7%) and XRP (91.8%) and pretty bullish on BTC at 86.5 percent, ETH at 88.2 percent and BCH at 84.5 percent respectively.

RoninAi data analysts has introduced three comprehensive trading strategies that have been dramatically outperforming the HODL. These strategies are called RSIS-X, RSIS-XS, and RSIS-XX:

– RSIS stands for relative strength index smoothing;

– X, XS, and XX stand for the frequency of trades and risk/reward ratio. X represents lower risk trading signals based on moving average smoothing and, therefore, generates lower number of trading signals. XX, on the other hand, typically generates trading signals more frequently yet each signal bears higher level of risk. XS is somewhere in between: moderate level of risk coupled with medium frequency of trading signals.

Thus, RSIS-X, RSIS-XS, and RSIS-XX offer great flexibility for both active traders and risk-averse investors. Moreover, RSIS react greatly to trend reversals. The chart of historical performance shows how masterfully RSIS avoids bearish trend in BCH while capturing the upside.

Similarly, despite significant selloff followed by an even greater upside, RoninAi RSIS strategies avoid negative trend reversal while capturing parts of the bullish trend for XRP.

In case of ETH that entered into a severe bearish trend, RSIS strategies have been minimizing losses and maximizing gains grossing total profits at 25 percent and higher in just a little over 3 months.

RSIS strategies detect more signals compared to EMAS, for example. Thus, relative strength index smoothing should be considered as the one offering more flexible strategies, which are able to capture shorter-term moves and realize higher returns while minimizing risk. Tables provided in the sections below summarize entry and exit points as well as general analytics on every relative strength index smoothing strategy.

RSIS-X Strategy: How to use it

Relative Strength Index Smoothing strategies are also very straight-forward. Tables presented in this section provide optimal entry and exit points. When relative strength index smoothing AI prediction for ETH breaks above 99 percent it is time to purchase ETH. Once the position is established it should be held until smoothing AI forecast for ETH drops below 70 percent. Since end of June RSIS-X strategy generated 67 signals with a total cumulative return of 25.7 percent and Sharpe ratio of 2.20. In a little over three months, RSIS-X strategy applied on ETH outperformed ETH HODL by almost 80 percent. RSIS-X strategy statistics for ETH and other 4 cryptocurrencies is presented in the table below.

RSIS-XS Strategy: How to use it

RSIS-XS works quite similar to RSIS-X. While exit points are the same, the entry points are lower, thus generating more signals and increasing the risk from low level to moderate. For example, when relative strength index smoothing forecast for ETH breaks above 95 percent it is time to purchase ETH. When the long position is established and the relative strength index smoothing forecast for ETH drops below 70 percent the signal is considered to be closed and it is time to sell ETH. In case of ETH, RSIS-XS algorithm outperformed the HOLD by more than 90 percent with the total return of 36.5 percent and Sharpe ratio of 2.87. RSIS-XS strategy statistics for ETH and other 4 cryptocurrencies is presented in the table below.

RSIS-XX Strategy: How to use it

Entry and exit points as well as statistics of RSIS-XX strategy effectiveness are presented in the table below. For example, when relative strength index smoothing forecast for BCH break above 90 percent it is time to purchase BCH. When the long position is established and the relative strength index smoothing forecast drops below 70 percent the signal is considered to be exhausted and it is time to sell BCH. Statistics for BCH and other 4 cryptocurrencies is presented in the table below as well. In case of BCH, RSIS-XX strategy generated 91 buy/sell signals with total cumulative profit of 18.8 percent and Sharpe ratio of 1.63 when BCH HODL strategy showed a huge loss of 40.6 percent within the same timeframe of just a little bit over 3 months. Thus, RSIS-XX outperformed the HODL by almost 60 percent in case of BCH.

For more information see our In-depth Usage Guide > here  and Our Alerts, Signals and Notification Guide > here.