2014issue C0918-23
A two-average brick crossover is not yet a trading system
A two-average crossover is not yet a trading system. The testable object is the full bundle of price definitions, brick sampling, brick-color confirmation, trend agreement, conflict-zone abstention, and next-brick fill, written so entry, exit, and standing aside are one mechanical procedure.
- Both averages are eight-brick simple moving averages computed on completed modified-renko bricks, one on typical price and one on heikin-ashi recalculated price, not on fixed-time bars.
- Virtual bricks that fill gaps are not valid entry or exit prices. Regular-hours session-break gaps must stay visible if brick-based signals are to remain valid.
- A modeled comparison treats a value above 0.8 as a long or short crossover signal, but a flip is accepted only when the trigger brick closes on the same side of its open as the new direction.
- A 21-bar simple moving average of close must agree with the crossover state. Disagreement is a conflict zone. The fill opens only at the next brick’s open, and a signal blocker stops repeated emissions in an already-open trend.
The testable object is the full procedure
The archive describes a brick-sampled comparison of two eight-period simple moving averages, not a standalone indicator reading. TradersWeek editorial view: that comparison is not yet a trading system. The testable object is the full bundle of price definitions, brick sampling, brick-color confirmation, trend agreement, conflict-zone abstention, and next-brick fill, written so entry, exit, and standing aside are one mechanical procedure.
Typical price and heikin-ashi recalculated price
The illustrated procedure compares an eight-brick simple moving average of typical price with an eight-brick simple moving average of heikin-ashi recalculated prices, both computed on modified-renko bricks. Typical price is high plus low plus close divided by three. It is the input series for one of the two averages. The heikin-ashi series is haOpen plus haHigh plus haLow plus haClose divided by four. That heikin-ashi recalculated price is the input series for the comparison average.
Modified-renko bricks set the sample
The worked example is drawn on a 100-tick, one-point modified-renko series of the S&P 500, with both averages sampled on those bricks rather than on fixed-time bars. Each completed modified-renko brick is the sampling unit for both averages and for every signal.
Virtual bricks and session gaps
Virtual bricks used to fill gaps are not valid prices for entry or exit. They keep the chart continuous, but they are not usable as a trigger price or as a fill price. Regular-hours instruments typically show large session-break gaps that must remain visible if brick-based signals are to stay valid. A 24-hour series has fewer small gaps but more quiet, limited-range bricks.
Crossover, threshold, and brick direction
A modeled comparison of the two eight-period averages treats a value above 0.8 as a long or short signal. That comparison threshold is the cutoff for a crossover signal on completed bricks. A long-to-short flip is accepted only when the trigger brick closes below its open. A short-to-long flip is accepted only when the trigger brick closes above its open. A zero open-to-close range does not confirm direction. This brick-direction filter withholds a flip unless the trigger brick’s close sits on the same side of its open as the new side.
Trend agreement and the conflict zone
A 21-bar simple moving average of close is combined with the crossover state so a long or short requires both inputs to agree. That slower close-based average is the trend-agreement filter. Disagreement produces a flat conflict zone, including intervals where crossover state disagrees with brick direction or with the slower trend average. The conflict zone may be held through by repeating the prior state, or used as an early-exit condition.
Next-brick fill and the signal blocker
The mechanical fill opens a trade only at the open of the brick after the trigger. Virtual bricks remain invalid for that fill. A signal blocker is required so a running trend does not keep repeating long or short signals until the next reversal.
All readings on this track · 14 readings
- 1990Moving-average direction, range penetration, and a dollar stop
- 1990Rally and reaction day signals as a mechanical stock system
- 1991Constructing steppable currency games with stop alerts
- 1995Trading as a business: a case study of preparation and rule-based entry
- 1997Bond timing rules as one testable procedure
- 2000Rule-based moving average commentary as a testable filter
- 2004Constructing a fifteen-bar trend trigger as a next-open system
- 2005State-locked rules for mechanical trading
- 2007Same-session long entry from a moving-average filter
- 2012Liquidity, share count, and stop-trigger costs
- 2012Monthly eight-bar average entry and exit rules
- 2012Constructing a simple moving average filter with fixed crossing rules
- 2014A two-average brick crossover is not yet a trading system
- 2020Reject long-bar entries driven by volatility fear of missing out