Renko: the chart that draws distance travelled, not time

A renko chart is built not on a schedule but on actual price movement: a new brick is added only when the price has travelled a preset distance from the close of the previous brick. If the market stands still, the chart stands still too, even for the whole session. If the market moves, the bricks are laid one after another for as long as the move lasts. This is the only fundamental difference between renko and the familiar chart types, and everything else follows from it.

How a brick is laid

Renko has a setting that the whole exercise revolves around: the brick size. It is set in points, as a percentage of the price, or through volatility (usually ATR). From there the mechanics are simple: until the price has moved away from the close of the last brick by the brick size, nothing happens on the chart. Once it has, the brick is fixed, and its close becomes the new reference point.

An important detail follows from this: the bricks are always the same height and always aligned to a grid. Renko has no shadows, no long tails, no "small candle with a big wick". All you see is colour and sequence. Compare this with the guide to reading a candlestick chart: there, the shape of the body and the shadows carries half the information, while here that information does not exist at all.

Why a reversal costs more than a continuation

Continuing the trend costs one brick size: the price moves a brick size in the direction of the current colour, and a brick of the same colour is laid. To change colour, however, the price must retrace the body of the current brick and then gain a further brick size in the opposite direction. In other words, the reversal threshold is higher than the continuation threshold by construction, not because the trader decided so, but because that is how the grid works.

This is renko's built-in filter. Small fluctuations inside the brick size never reach the chart at all, and the picture looks smooth: long runs of a single colour and rare reversals. The price of that smoothness is lag. A reversal on renko is always confirmed later than it happened in the price, and the size of that delay equals that very threshold.

What renko throws away

Renko removes time from the chart. The X axis is no longer uniform: neighbouring bricks may be minutes apart, or several sessions apart. You cannot read from the chart exactly when an impulse occurred, how long it lasted, or whether it coincided with the market open. Volume in its usual frame of reference is gone as well: volume accumulates across bricks of differing duration, and comparing them with each other is not valid.

Gaps disappear along with time. Renko simply fills the gap between the close and the next day's open with bricks, as if the price had covered that distance continuously. For the Russian market this matters: a significant share of price movement here comes from the reaction to news and from dates in the events calendar, and renko shows such a gap as an even staircase, that is, as a move that never took place and that nobody could have traded.

Building on closes or on extremes

This is the subtlety that makes two traders with identical settings see different charts. If the bricks are calculated from the closing prices of the underlying timeframe, part of the intrabar move is lost, and a reversal is registered later. If they are calculated from highs and lows, there are more bricks, more reversals, and the backtest acquires a pleasing but unattainable precision: the movement inside the bar is reconstructed from an assumption about the order of events, an order that the source data does not contain. Before comparing results, it is worth finding out exactly how your terminal assembles a brick. This is part of the definition of the chart type, not a minor display setting.

Where renko behaves honestly

Renko makes sense where there is a continuous flow of trades and the brick size is noticeably larger than the spread: liquid names from the list of stocks, index futures. On illiquid instruments the picture breaks down: infrequent trades at random intervals produce either a dead chart or a run of bricks on one order. For the same reason renko sits poorly on bonds, where price movement is slow and is driven largely by the coupon rather than by order flow.

A related approach is the tick chart: there, too, an element appears on an event, but the event is a trade rather than a distance travelled. The difference is worth understanding: a tick chart compresses quiet periods and stretches activity, while renko compresses quiet periods and stretches movement, and these are not the same thing on days of high activity with no direction.

What renko does not do

It does not remove noise, it hides it: the fluctuations inside the brick size did happen, they simply did not make it onto the screen, and a stop order will feel them. It does not improve the entry point: the filter is always paid for with lag. And it does not provide a statistical edge by itself: the platform has no data on the performance of renko strategies on Russian instruments, so any claim that "signals are cleaner on renko" remains here an untested opinion, not a measured fact.

Definitions of related concepts are in the glossary.