Ichimoku on Renko charts: three system ideas to test

A forum member asked me whether Ichimoku works on Renko charts. Very few Indian platforms let you apply Ichimoku to Renko, and little has been written about the pairing, so I wrote up how I would build a system around it. The full post on the Definedge forum has more charts (Natural Gas, a stock and an option premium chart) and the ideas that came up in the discussion. This page is the short version. It gives ideas to test, not backtest results.

Goichi Hosoda, a Japanese journalist, built Ichimoku in the 1930s. It uses the middle of the recent high and low instead of the closing price. The Tenkan is the midpoint of the last 9 periods and the Kijun of the last 26. The cloud, made of Span A and Span B, is plotted 26 periods ahead. Price above the cloud is read as an uptrend, below it as a downtrend, and inside it as undecided.

On Renko a period is a brick, not a minute or a day. At a 0.05% brick, one Nifty brick is about 11 points, so the Tenkan becomes the middle of the last 9 bricks and the Kijun the middle of the last 26. In a clean run of up bricks the Kijun sits about 13 bricks below price. An exit on a close below the Kijun therefore needs a fall of 13 to 14 bricks from the top, about 0.7% or close to 150 Nifty points. Sideways minutes print no bricks, so they cannot move the lines.

1. The classic cross, on bricks. Buy when the Tenkan crosses above the Kijun and the brick closes above the cloud. Square off when a down brick closes below the Kijun. The short side is the mirror image. The Kijun works as a trailing stop on its own.
2. The cloud as the filter, Swing Breakout as the trigger. Buy when the brick is above both spans and a Swing Breakout Bullish signal prints. Square off when a brick closes below the cloud. The cloud picks the side and the breakout times the entry; inside the cloud you stay flat.
3. Long only. Nifty and Sensex have risen over long periods, so a system that shorts as often as it buys works against that drift and pays charges on both sides. Take only the bullish trades from either idea and sit in cash while the bricks are below the cloud.

Whichever you pick, test the long and short sides separately before you decide that you need both.

• Run the same rules on 5-minute and 15-minute candles. If a time chart does as well, Renko is one way to cut noise, and you have learnt that it is not the only one.
• Try a smaller and a larger brick, and other markets such as sectors, midcaps and commodities. A rule that works at only one brick size may be luck.
• Leave the Chikou out of the rules. It is today's close drawn 26 periods back, so a backtest rule that reads it can look into the future.
• Charge realistic costs. A fast Renko system trades often, and charges and slippage add up quickly.

You can also take the signal on the index and trade it through an option credit spread, the approach I describe in my Finding Edge talk.