Regression-based High-probability Momentum Buy Strategy

A. Introduction

The strategy begins with a weekly support and resistance setup which establishes the buy zone of the week. When the market is trending up and the price falls into the predetermined trade zone, a set of trigger indicators will cast their votes to decide whether it is a favorable time to buy.

Once an order is active, indicator-based stops will stay on guard. Those indicators will trigger sell signals when price and volume converge and price volatility is low, warning signs of an impending exhaustion in the prevailing uptrend. In essence, the strategy aims to secure the best possible closing price before the prevailing uptrend is broken.

The strategy put forth uses moving averages to define the trend, Bollinger Band to determine price volatility, and divergence of Demand Index and price regression to gauge and forecast the strength of an uptrend.

B. Four indicators

1. Weekly Fibonacci Support and Resistance Setup

Every Monday, the strategy observes the market condition before establishing weekly support and resistance levels of the week. First, the indicator uses moving averages to divide the price development leading up to that point into uptrend and downtrend areas. Then, once a certain set of conditions has been fulfilled, major upper fractal of the latest uptrend and major lower fractal of the latest downtrend are used as reference points to draw Fibonacci 50%, 68.2% and 100% retracement levels on the chart.

The section between the 68.2% and 100% Fibonacci retracement levels is the weekly trade zone of the strategy. That is, the strategy will not pursue any trading opportunities with prices fall outside of the predetermined trade zone.

2. Bollinger Band Volatility

The indicator is derived from two Bollinger Bands to determine price volatility.

3. Moving Averages Divergence

The indicator is derived from two moving averages to signal an uptrend that has not yet fully formed, suggesting a high probability buy.

4. Price Volume Buy Momentum

The indicator is derived from Demand Index* and price regression*. When the two indices diverge, one suggesting neutral and the other suggesting uptrend, the indicator becomes positive and suggests a buy signal. On the contrary, when the two indices converge, implying the uptrend is in full force and may soon peaked out, the indicator turns negative and generates a sell signal in the hope that the strategy can close a sell order at the most favorable price.

*Demand Index:

*Price Regression: 3*Simple Moving Average (13) – 2*Linear-Weighted Moving Average (13)

C. Buy Signals

1. Market is trending up and price is within the buy zone of the week

2. Price Volatility is high

3. Moving averages diverge

4. Price and Volume diverge

D. Two Trading Examples

1. One Good Trade: AUD/JPY

The example below shows an AUD/JPY 15-minute candlestick chart with two buying signals (green arrows) and two selling signals (red arrows).

At 11-08-2011 15:45 (the vertical dotted line on the left), the price was in the weekly trade zone and indicators signaled that the price volatility was high [1] and moving averages diverged [2]. Notice that the price volume buy momentum was neutral [3]; therefore, the momentum indicator did not contradict the buy signals generated by the other two indicators. As such, it was still a favorable buying opportunity.

Before confirming a buy market order, the strategy calculated the Risk/Reward (R/R) ratio to filter less profitable trades (only trades with less than 0.34 R/R ratio will be considered) at the outset. In this example, expected R/R is 0.28 and therefore the strategy entered a buy order.

Here is the calculation:

Expected price: 78.7400 + Slippage

Stop loss: 78.5363 (60 Linear Weighted Moving Average)

Take Profit: 79.4700 (Weekly Fibonacci 100% retracement level)

Risk/Reward Ratio: 0.2790 ( (78.7400 -78.5363)/( 79.4700 - 78.7400) )

At 11-08-2011 19:00, three hours and 15 minutes after the order entered, the price volatility turned to neutral [4], moving averages converged [5] and price volume also converged [6]. Even though the price might still stand a chance to touch the Fibonacci 100% retracement level, the order was stopped out by the strategy since two of the three indicators suggested the uptrend was about to end.

The trade resulted in a profit of 60 pips. (79.34 - 78.74)

2. One Losing Trade: EUR/USD

Although the Moving Averages Divergence indicator is designed to filter noise in the market to reduce whipsaw losses, the strategy still falls into those traps.

Here is an example that resulted in a loss in a choppy market.

The example below shows a EUR/USD 15-minute candlestick chart with three buy signals (green arrows) and one stop loss signal (red arrow).

At 2012.07.19 8:00, when price was in the weekly trade zone with high price volatility [1], moving averages divergence [2], strong price volume buy momentum [3] and Risk/Reward ratio calculated as 0.26, the strategy then decided to go long.

Here is the calculation before confirming the long order:

Expected price: 1.2293 + Slippage

Stop loss: 1.22870 (60 Linear Weighted Moving Average)

Take Profit: 1.23160 (Weekly Fibonacci 100% retracement level)

Risk/Reward Ratio: 0.2609 ( (1.2293 -1.22870)/( 1.23160 - 1.2293) )

At 2012.07.19 8:30 [4], the order stopped out prematurely when the price plunged and crossed the 60 moving average stop loss (the red line). The order ended up a losing trade. (-6 pips = - 1.2293 + 1.22870)

E. Conclusion

The strategy discovers profitable trades by exploiting uptrend price movements that have not yet fully formed. To further improve the entry point of each trade, expected Risk/Reward ratio is calculated and trade signals with R/R ratio greater than 0.34 will be discarded by the strategy. Nevertheless, the strategy still cannot avoid being whipsawed in a choppy market.

To remedy this shortcoming, head-and-shoulder pattern recognition can be incorporated into the strategy to filter false outbreaks and thereby boosting the profitability of the strategy.

Translate to English Show original