Tick Data vs OHLC Backtests: When the Difference Matters

TL;DR: OHLC backtests are fast, good enough for swing and position systems, and fine for most indicator development work. Tick data backtests are slower and more complex, but they are the only reliable option when your strategy fires inside a single candle. Knowing which situation you are in will save you from discovering the gap between your backtest equity curve and your live account the hard way.**

What the Backtest Engine Is Actually Doing

Before the OHLC vs tick argument makes sense, you need to know what MetaTrader's strategy tester is doing under the hood when you pick each modeling method.

When you run a backtest on OHLC data, the tester receives one price record per candle: open, high, low, close. It has no idea which came first, the high or the low. It makes assumptions. By default, on a bullish candle it guesses open first, then low, then high, then close. On a bearish candle it reverses that sequence. This is called the OHLC on M1 modeling method when it reconstructs smaller timeframes, and it is a shortcut that works surprisingly well in many situations and breaks down completely in others.

When you run on tick data, the tester replays actual bid/ask prices as they arrived at the broker's feed, one tick at a time. Every stop, every limit, every trailing adjustment gets evaluated against a real price sequence rather than a reconstructed one. The simulation is slower by a significant margin, but the price path is real.

The gap between those two approaches is where traders get into trouble.

When OHLC Is Good Enough

Swing and Position Systems

If your strategy operates on daily, H4, or H1 candles and holds positions for hours to days, OHLC data is almost certainly sufficient. Here is why: when a candle has a range of 80 pips and your stop is 50 pips away from entry, the question of whether the high came before or low on a given bar has very little practical effect on whether you got stopped out. The candle either reached your stop or it did not.

For swing systems the relevant variables are:

• Entry signal: Generated at bar close, confirmed on bar open. OHLC captures this accurately.
• Stop placement: Usually outside recent structure, far enough from intra-bar noise to make bar-order irrelevant.
• Target placement: Reached over multiple candles, not within a single one.

In these conditions an OHLC backtest is not just acceptable, it is often preferable. It runs faster, the results are easier to iterate on, and you are not introducing tick-data quality problems (more on those shortly).

Multi-Timeframe Indicator Systems

If your strategy uses higher timeframe confirmation to filter entries on a lower timeframe, and both entry and exit happen at bar opens or closes, OHLC remains reliable. The entry is either valid or it is not at the moment the bar closes. Your backtest engine handles that correctly.

multi-timeframe indicator strategies

When You Cannot Trust an OHLC Backtest

This is the part most traders skip, and it costs them real money when they go live.

Scalping Strategies

A scalping system operating on M1 or M5 candles with a stop of 5-10 pips lives and dies by intra-bar price action. When the tester has to guess whether the high or low came first on a 12-pip range candle, it is guessing about the event that determines whether your position survived or got closed. That guess is wrong often enough to produce backtest results that look meaningfully better than live performance.

The classic symptom is a backtest that shows a smooth equity curve with a controlled drawdown and a live account that hits stops at a rate that feels impossible. The stops are not the problem. The simulated price path was.

Strategies with Intra-Bar Trailing Stops

Trailing stops that move inside a candle create the same problem even on longer timeframes. Say you are trailing on an H1 chart with a 15-pip trail. During a single H1 candle, price could move up 25 pips, trigger your trail adjustment twice, then reverse and hit your revised stop, all within the same candle. OHLC data gives the tester one open, one high, one low, one close. It has no record of that sequence. It will model it incorrectly in a meaningful percentage of cases.

If your trailing logic updates more frequently than once per candle close, you need tick data or at minimum M1 OHLC reconstruction to get a backtest you can trust.

Breakout Systems with Tight Stops

Breakout entries that trigger on price crossing a level, combined with stops that sit close to the entry, are vulnerable to the same modeling error. The tester may show a clean entry and a trade that ran to target, when in reality the price spiked through the entry level, reversed, took out the tight stop, then moved to the target without you.

This specific failure mode is sometimes called stop-hunting in reverse: the backtest misses the stop because it did not model the real price sequence inside the candle.

How Much Does the Difference Actually Matter?

This depends on three things: your holding period, your stop distance relative to bar range, and how often your strategy interacts with stops or targets within a single candle.

A rough way to think about it: calculate the average ratio of your stop distance to the average candle range on your trading timeframe. If your stop is typically larger than the candle range, OHLC is almost certainly fine. If your stop is smaller than half the candle range, you should be using tick data or M1 reconstruction at minimum. If your stop is a fraction of the candle range, tick data is not optional.

Tick Data Quality: The Problem Nobody Warns You About

Getting tick data is only half the problem. Getting reliable tick data is the real challenge.

MetaTrader's built-in data download gives you tick data from your broker. The issue is that this data is only as good as your broker's feed, and broker feeds vary considerably. Some brokers have gaps during low-liquidity periods. Some aggregate price differently. If you downloaded historical ticks from a broker with a wide spread policy and you are testing a strategy you plan to run at a broker with tighter spreads, your backtest results will not translate accurately.

A further complication: MetaTrader 4's tick data quality has historically been inconsistent. The platform's own quality score (visible in the tester report) gives you a rough indicator, but a high quality score does not guarantee accurate tick sequences. It only measures data completeness, not fidelity.

Where to Get Reliable Tick Data

For serious tick data backtesting, traders generally use one of these approaches:

Dukascopy Historical Data: Dukascopy offers free tick data for major pairs going back well over a decade. Their data is sourced from their own ECN and is generally considered reliable for major pairs. The format needs to be converted for use in MetaTrader, but tools exist to handle this.

Tickstory: A desktop application that pulls data from Dukascopy and packages it for MetaTrader import. It simplifies the conversion process and lets you apply spread adjustments to simulate different broker conditions.

Your live broker's data: If you are testing a strategy you will run at a specific broker, their own historical ticks are the most relevant, assuming the data is complete. Download and inspect it for gaps before trusting the results.

Third-party data providers: Several paid services offer institutional-grade tick data for a wider range of instruments. These make sense if you are trading less common pairs or need very long history.

setting up MetaTrader strategy tester for accurate results

Whatever source you use, check the data for gaps at known high-volatility events. News spikes and session opens are where lower-quality data sources are most likely to have problems, which is also where your strategy is most likely to be active.

Is Tick Data Backtesting Worth the Effort?

For most swing and position traders, the answer is no. The added complexity, slower test speeds, and data quality management overhead do not produce meaningfully better information than a well-structured OHLC backtest.

For scalpers, the answer is unambiguously yes. The only alternative is forward testing on a demo account for long enough to build statistical confidence, which takes months. A proper tick data backtest compresses that time considerably, even accounting for the imperfections in the simulation.

For intra-bar trailing strategies, the answer is yes if you are trying to get accurate drawdown and win rate figures. If you only want to confirm the general logic works, M1 OHLC reconstruction is often a reasonable middle ground.

intra-bar trailing stop indicators for MetaTrader

People Also Ask: Is Tick Data Better Than OHLC for Backtesting?

Tick data produces a more accurate simulation of how price moved inside a candle. That makes it better in the sense of being more representative of real market conditions for strategies that depend on intra-bar price behavior. It is not better in every situation. For strategies that only act on bar closes or opens, OHLC backtesting is accurate enough and meaningfully faster. The better question is whether your specific strategy's performance is sensitive to intra-bar price sequence. If it is, use tick data. If it is not, OHLC is fine.

FAQ

Q: Can I use MetaTrader's "Every tick" option instead of real tick data?

A: MetaTrader's "Every tick" model uses M1 OHLC data and interpolates synthetic ticks between bars. It is better than "Open prices only" for strategies with intra-bar logic, but it is not the same as real tick data. The synthetic ticks follow the same high/low ordering assumption as OHLC modeling. For scalping strategies, it can still misrepresent performance materially.

Q: How do I know if my backtest results are being distorted by OHLC modeling?

A: Run the same strategy on both OHLC and tick data for a sample period and compare the trade lists. Look specifically at trades that were stopped out. If a meaningful percentage of OHLC stop-outs do not appear in the tick run, or vice versa, you have a modeling sensitivity problem. Also check for trades that show markedly different entry or exit prices between the two runs.

Q: Does tick data backtest accuracy matter less on higher timeframes?

A: Yes, and quite significantly. On daily and weekly charts the gap between OHLC modeling and tick data is small enough to be practically irrelevant for most strategies. The candle ranges on higher timeframes are large relative to typical stop distances, which reduces the probability that bar-order assumptions affect trade outcomes.

Q: How much historical tick data do I need for a reliable backtest?

A: Enough to cover multiple market conditions: trending periods, ranging periods, and high-volatility events. For most strategies, two to three years of tick data gives you a reasonable sample. More is better, but data quality tends to decline the further back you go with some providers. A smaller sample of high-quality data is worth more than a large sample of incomplete data.

Q: Does spread affect tick data backtest results?

A: Yes, and it is one of the most common sources of discrepancy between tick data backtest results and live performance. If you test with a fixed 1-pip spread and your broker widens to 3 pips during news, every trade that fires near news will have a worse execution in live conditions. Good tick data testing tools let you apply variable spread models. Use them if your strategy is active during volatile periods.

The Bottom Line

The OHLC vs tick data decision is not about which method is more sophisticated. It is about whether your strategy's results are sensitive to what happens inside a single candle.

Swing traders can build robust systems and trust their results using OHLC data. Scalpers and anyone whose strategy fires stops or adjusts trailing logic within a bar needs tick data, and they need to source it carefully rather than assuming the broker's download is complete and accurate.

Run both if you are unsure. The comparison will tell you quickly whether the modeling method matters for your specific setup, and that is more useful information than any general rule.