The Market Doesn't Trade in Candles. We Invented Those.

What happens when you shift a candlestick chart by just a few seconds? Our research suggests that even a small change can completely alter what traders see.

ApolloAlpha Research | 3 November 2026

The myth: A five-minute candle tells you what happened in those five minutes.

Open almost any trading chart and you'll see candlesticks.

Green means price went up. Red means price went down. The wicks show how far price travelled before the candle closed.

Traders use these candles to identify momentum, spot reversals, find breakouts and make trading decisions.

But there's something we rarely question:

Who decided when a candle should start?

A five-minute candle might begin at 10:00:00 and close at 10:04:59. The next begins at 10:05:00.

But what if we shifted everything forward by just three seconds?

Would the candles tell the same story?

We decided to find out.

The experiment: What if we changed the clock?

We analysed S&P 500 futures price data and reconstructed five-minute candles using 60 different starting points.

Instead of always starting at the beginning of the five-minute interval, we shifted the starting point in five-second increments.

That meant comparing candles built from the same underlying market activity, but grouped differently.

We wanted to answer three simple questions:

• Would the closing prices change?
• Would bullish candles become bearish, or vice versa?
• Would the differences disappear when we used longer timeframes?

The results were surprising.

1. The same market activity produced different closing prices.

Across our tested five-minute periods, the average standard deviation of closing prices across the 60 alignments was approximately 0.98 index points.

More strikingly, the average difference between the highest and lowest closing prices across those alignments was approximately 3.77 points.

For context, the average five-minute candle range was approximately 4.03 points.

In other words, changing where we started the candle could shift its closing price by an amount comparable to the entire average range of the candle.

2. The direction of the candle could change.

This was perhaps the most interesting finding.

Approximately 42% of the tested five-minute periods produced mixed bullish and bearish readings across different alignments.^*

Think about what that means.

You could be looking at a bullish candle while another trader, using the exact same underlying price data but a different starting point, sees a bearish one.

Neither trader has different market data. They've simply grouped it differently.

3. Longer timeframes didn't completely solve the problem.

We also tested timeframes from one minute to fifteen minutes.

Naturally, larger candles showed less relative variation. But the effect didn't disappear.

Even at fifteen minutes, the standard deviation of closing prices across alignments was still approximately 23.5% of the average candle range.

The problem wasn't exclusive to short-term charts.

See it for yourself

The chart below uses the same data. Drag Bar start offset to move the start of every candle by seconds and watch the candles change. Press Play to sweep through every start time; switch on Persistence first to keep each one as a faint layer, and Time-weighted median to add a line that weights each second of price equally.

Best viewed on a larger screen. Scroll to zoom, drag to pan.

So, what does this actually tell us?

Here's the bigger idea.

The market doesn't trade in candles. We invented those.

Price moves continuously as buyers and sellers interact. Trades don't wait for the next five-minute candle to begin.

A candle is simply a way of grouping that activity into convenient intervals.

And while certain moments genuinely matter to the market — such as exchange openings, closing auctions and major economic announcements — an arbitrary five-minute boundary doesn't automatically have any special significance.

Yet traders often build strategies around these boundaries.

We look for bullish engulfing candles, bearish reversals, breakout closes and candle patterns as though they are objective descriptions of market behaviour.

But what happens when the signal changes just because we move the starting point by a few seconds?

Perhaps we're not always measuring market behaviour. Sometimes, we're measuring the way our chart happens to display it.

Does this mean candlestick analysis is useless?

Not necessarily.

Candlesticks remain useful for simplifying price information and identifying potential patterns.

But this research highlights an important question that traders should ask:

Would my trading signal still exist if I changed the candle alignment?

If a strategy only works with one specific starting point, it may be relying more heavily on chart construction than on a robust market behaviour.

That doesn't automatically make the strategy unprofitable. It does mean its sensitivity to candle alignment deserves to be tested.

The takeaway

Most traders spend hours searching for the perfect indicator, the perfect candle pattern or the perfect timeframe.

But perhaps we should spend a little more time questioning the assumptions behind the charts themselves.

The market doesn't know when your five-minute candle starts.

Why should your trading strategy depend on it?

Research note: This study used S&P 500 futures price data from 5–30 October 2026, comparing 60 starting-time alignments across multiple intraday timeframes. Results are specific to the sample and methodology tested; they do not establish that all candlestick strategies are unreliable or unprofitable. This is an exploratory study, not a trading recommendation.

^* "Mixed" means that between 20% and 80% of the 60 alignments of a five-minute period closed up.

ApolloAlpha Research
Questioning assumptions. Testing what the market actually does.