Renko chart backtests: brick size, missing wicks and the fills that lie

Contents

1. What is a Renko chart and why do its backtests look too good?
2. How does a Renko brick get built, and when is it really known?
3. Where does a Renko backtest see the future?
4. Brick-price fills
5. No wicks, no stop-outs
6. Phantom bricks
7. How does AlphaPrime backtest a Renko strategy?
8. How do you choose a Renko brick size for backtesting?
9. Which markets and strategies suit Renko bricks?
10. Frequently asked questions

What is a Renko chart and why do its backtests look too good?

A Renko chart is a price chart made of equal-sized bricks: a new brick is drawn only when price moves a set amount from the last brick, and time is ignored. Renko backtests look good because the brick's own price is used as the fill, and a brick has no wick to hit your stop.

The name comes from the Japanese word for brick. A brick size is the fixed distance, in points or ticks, that price must travel to add a brick. On most charts an up brick continues when price closes one brick size above the previous brick's top, and the direction flips only after price moves two brick sizes the other way: one to get back across the previous brick, one more to draw the reversal brick. That double requirement is why Renko trends look so clean. Small pullbacks never reach the chart.

Traders use Renko to strip noise from trend-following rules: a moving-average cross or a two-brick breakout reads cleanly on bricks and terribly on 1-minute candles. The chart is a good filter. The way most engines backtest it is not, and the difference between the two is the subject of this page. The common rules for every synthetic chart type are on the overview of why Renko and Heikin Ashi backtests cheat; this page is the Renko-specific part.

How does a Renko brick get built, and when is it really known?

A brick is known only when a real trade has crossed its completion level, and a backtest engine built from minute bars only learns that at the close of the minute. Everything before that is a forecast dressed up as a bar.

Take a 10-point brick. The last brick's top is at 100. Price trades up through 110 at 09:41:17. On a chart replaying history, a new brick appears with a close of 110 at that moment. In a backtest that fills at the brick close, you are long at 110 as if you had been standing at the exchange watching the exact tick. Live, your platform sees the brick complete, your rule fires, your order goes out, and it fills at whatever the next trade is. In a market moving fast enough to complete bricks, the next trade is rarely 110.

AlphaPrime (Valynox) builds Renko bricks from the underlying time bars you choose, usually 1-minute, and a brick becomes visible to the strategy only at the close of the underlying bar that completed it. If three bricks complete inside one minute, all three appear at the end of that minute, and every signal from them fills at the same next real price. This is slightly later than a tick-built chart would show the brick, and that delay is the point: it is the earliest moment the brick is certain.

Two more rules shape the bricks. A minute bar does not say whether its high or its low came first, so when both could matter the engine assumes the order that is worse for a trend follower: if the last brick was up, the minute's low is taken first, so a reversal can sweep you out before the continuation prints. That uses only the minute's own four prices and never borrows from the next one; its cost is a somewhat higher brick count than the real market would have shown. And the first price in the data decides where every brick boundary sits, so two charts of the same market that start on different days can disagree about brick positions for a while; test across a range of brick sizes rather than trusting one grid.

Where does a Renko backtest see the future?

In the fill, in the stop and in the gaps. All three come from treating the brick as if it were a traded price.

Brick-price fills

The brick close is a level, not a trade. Filling at it assumes you were filled at the exact tick that completed the brick, with no latency and no slippage, every time, in both directions. For a reversal brick the assumption is even kinder: the reversal brick's close is two brick sizes from the previous extreme, so a fill there means you caught the turn at the best possible level the chart can express.

No wicks, no stop-outs

A classic Renko brick records only its open and close. The real trades inside the brick may have run well past both. If your stop is 15 points below entry and price dipped 18 points before completing the next up brick, a brick-range stop check says you were never stopped. A real-price check says you were. Renko backtests with tight stops are the most flattered of all, because the chart erases exactly the excursions that would have hit the stop.

Phantom bricks

When price gaps, for example at the open or on a news print, a Renko builder can draw every brick between the old level and the new one in a single step. Those intermediate bricks, called phantom bricks, represent prices that never traded. A strategy that enters on the second of five phantom bricks is entering at a price nobody could have bought at. A gap from 100 to 150 draws five 10-point bricks at once; only the last one corresponds to anything you could trade. AlphaPrime lets you show or hide phantom bricks (Show Phantom Bars, on by default so the chart reads as one line), but the choice only changes the drawing: all bricks a gap implies appear at the same underlying close, and nothing can fill inside the gap.

Renko feature	What a brick-price backtest assumes	What AlphaPrime does
Brick completion	Known at the completing tick	Known at the close of the underlying bar
Entry fill	Brick close	Next real price plus doubled slippage
Stop and target check	Brick open and close	Underlying bar high and low, stop first
Order of high and low inside a minute	Whichever order the engine prefers	The order that hurts a trend follower: against the last brick first
Gap of several bricks	Each phantom brick tradeable	All bricks appear at once, drawn or hidden; fill at the real price after the gap

How does AlphaPrime backtest a Renko strategy?

Signals are read from the bricks, exactly as you wrote the rule. Fills, stops and targets use the underlying real bars. Those two sentences are the whole method; the rest is detail.

• Brick confirmation: a brick exists for the strategy from the close of the underlying bar that completed it. Bricks that complete together appear together.
• Market orders: filled at the open of the next underlying bar after the signal, plus slippage.
• Stop and limit orders: rest inside the underlying bars until a real trade reaches the price.
• Stops and targets: judged on each underlying bar's high and low. If one bar covers both the stop and the target, the stop is taken first.
• Path inside a minute: the order of high and low is unknown, so the engine takes the order that goes against the last brick first.
• Gaps: all bricks the gap implies appear at once at the underlying close, whether or not phantom bricks are shown; no order can fill inside the gap.
• Slippage: multiplied by two, and you can only raise it, because bricks complete when the market is moving fastest.

The engine's own measurement on NQ 1-minute data from late 2018 to 2025 with a 10-point brick shows what the fast-market rule is for. About one minute in nine that produced a brick produced three or more. For a minute that produced one brick, the distance from the brick's close to the next real open averaged about 14 ticks; for three or four bricks it was three times that; for five or more, almost seven times, with the worst one percent of cases running to hundreds of ticks. Two honest footnotes: in the single-brick case the real fill was on average slightly better than the brick price, because bricks tend to complete at a minute's extreme and price pulls back before the close, and all of these distances are already inside the backtest, because the fill is the next real open. The multiplier is extra, because a minute bar cannot measure what the book looked like during those bursts.

What you lose is the illusion. A Renko breakout rule that showed a smooth equity curve with brick fills usually keeps the shape of its trades but loses the brick-to-fill distance, twice (entry and exit), and gains the uncertainty the chart hid. Whether what is left is worth trading is the question the backtest should have been answering all along. The same slippage logic applies to every multi-brick chart; see the general treatment of commission and slippage in a backtest.

How do you choose a Renko brick size for backtesting?

Pick a brick size that is several times your expected slippage and that produces a brick count your underlying resolution can resolve. A brick smaller than the typical 1-minute range is being built from data that cannot see inside it, and its backtest will be a story about the minute bars, not about the bricks.

Three workable approaches, in order of how often they hold up:

1. Fixed brick in points or ticks, chosen so the average brick takes several underlying bars to form. On NQ a 10-point brick gave about 210 bricks a day on the measured data and a 5-point brick four times as many; on ES a 4-point brick gave about 74 a day. The right number is the one where your underlying bars can resolve the brick, not a magic value, and the chart-type panel shows the bars-per-day figure for whatever you type.
2. Brick as a multiple of average true range, recomputed from completed bars only. This adapts to volatility, but the ATR must be frozen at the time each brick starts or you have built a brick size that knows the future. FlexRenko and ATR-based bricks have their own page: FlexRenko bars and offsets.
3. A brick that matches your stop, so a one-brick reversal is roughly your risk per trade. Simple to reason about, and it keeps the stop outside the noise the brick was meant to filter.

Whichever you choose, keep it fixed across the in-sample and out-of-sample periods. Tuning brick size on the test data is parameter fitting with a prettier name.

Which markets and strategies suit Renko bricks?

Renko suits trending, liquid markets and slow rules; it punishes mean-reversion ideas and anything that needs a precise exit. The brick reversal rule means you always give back at least two brick sizes before the chart admits the trend is over, which is fine for a swing rule and ruinous for a scalp.

• Good fit: breakout and continuation rules on index futures, major currency pairs and large-cap crypto during active sessions, where bricks form from genuine movement rather than from spread noise.
• Poor fit: thin markets where one trade can complete a brick, overnight sessions with wide spreads, and any rule whose edge depends on exiting within a few ticks.
• Watch for: strategies whose trades cluster into a handful of fast sessions. Their good results came from the bars where fills are worst, which is why the doubled slippage is the starting point, not a worst case.

If a Renko rule survives real-price fills, doubled slippage and the half-brick test, it has earned a walk-forward run. The method for that is on the page about walk-forward analysis, and it is the same for a Renko strategy as for any other.

What this means if you use AlphaPrime: choose Renko as the chart type in strategy generation, set the brick size and the underlying resolution (1-minute for intraday bricks), and let the search run. Every candidate is tested with confirmed bricks, next-real-price fills, doubled slippage and real-bar stop checks, in Classic and in Reserve. The strategies that come out the other side are the ones whose rule, not whose chart, did the work.

Frequently asked questions

Do Renko charts repaint?

Completed bricks do not change. The forming brick is redrawn with every tick, and on some charts its open shifts depending on which direction completes it. A backtest that trades the forming brick is trading something that was not yet true. AlphaPrime only shows a brick to the strategy after the underlying bar that completed it has closed.

Why is my Renko backtest much worse in AlphaPrime than on my charting platform?

Because the fill moved from the brick price to the next real price, the stop is now checked against real highs and lows, and slippage is doubled. Each of those takes back edge that was never available live. If the strategy is still profitable after all three, that is a result you can take seriously.

What brick size should I use for ES or NQ Renko backtests?

There is no single right number. Start with a brick that takes several 1-minute bars to form on an average day and is at least four times your slippage assumption. Then test half that size. If the strategy only works at one size, it is fitted to the brick, not to the market.

Can Renko be used for forex and crypto strategies?

Yes, with the same rules. Forex bricks are usually set in pips, crypto bricks in price units or as a percentage. The fast-market slippage doubling matters more in crypto, where bricks complete in bursts around news and liquidations.

Does AlphaPrime draw Renko wicks?

You can turn them on (Show Wick) to see how far price travelled before each brick completed, but the strategy does not depend on them: stops and targets are judged on the underlying real bars, which already contain every high and low a wick would show. The brick is for the signal; the real bar is for the money.