Tick data

Tick data is the finest record of quoted prices most retail traders can get, far finer than a 1-minute candle. If your strategy takes small profits or uses tight stops, a test on candles alone can show a profit that real ticks would remove.

What one tick records

A tick is one update of the price. In forex, each tick holds three things: a time stamp, the bid (the price you can sell at) and the ask (the price you can buy at).

The gap between bid and ask is the spread. So tick data also stores the real spread at every moment, including the moments when it widens.

Tick data stream: bid and ask at each tick, with the spread widening during a fast move

Forex has no single exchange, so there is no single official tick record. Each broker and data vendor publishes its own stream. Two tick files for the same pair and the same day will differ slightly, and so will backtests run on them.

Why a candle cannot tell you which level was hit first

Suppose you buy at the open of a 1-minute candle. Your stop loss and your take profit both sit inside that candle's range. From the candle alone, nobody can say which one was touched first.

The drawing below shows that case. The candle closes higher and looks like a winning trade. The ticks inside it show the price falling to the stop before it rallied to the target, so the trade was a loss.

One 1-minute candle compared with the tick path inside it

A tester without ticks has to guess the order. A tester that assumes the price went straight from open to high books the take profit, although the real ticks hit the stop first. That is how a losing scalper can look profitable on paper.

How tick data changes a backtest result, in numbers

The figures here are an illustrative example. An Expert Advisor (EA, an automated strategy) scalps EURUSD with 0.10 lots, so one pip is worth $1. It takes 1,000 trades.

A coarse test assumes a fixed spread of 1.0 pip and reports an average net gain of 1.0 pip per trade. The tick data shows something else: the EA enters during fast moves, when the real spread averaged 1.6 pips.

Backtest profit with a fixed spread versus the real spread stored in tick data

Pip value at 0.10 lots10,000 × 0.0001= $1

Profit in the coarse test1,000 × 1.0 × $1= $1,000

Extra spread per trade (pips)1.6 - 1.0= 0.6

Net gain per trade on ticks (pips)1.0 - 0.6= 0.4

Profit in the tick test1,000 × 0.4 × $1= $400

Profit that existed only in the coarse t$600

The strategy is the same and so are the trades, but more than half of the profit is gone. You can copy this check: multiply your number of trades by the extra cost per trade and by your pip value.

Compare your average target with the spread to decide whether a coarse test is good enough:

• If your average net profit per trade is only a few spreads or less, the coarse test is not trustworthy. Rerun it on real ticks.
• If your stop loss or take profit sits within a few spreads of the entry, a wider spread alone can trigger it. Rerun the test on real ticks.
• If your average target is many times the spread, a coarse test is usually close enough. A strategy that aims for hundreds of pips on daily candles barely notices 0.6 pip. A scalper lives or dies by it.

Where you see and choose tick data on your trading platform

In live trading, the ticks are the flickering bid and ask in your quote list.

In a strategy tester, you choose how prices are modelled. The names vary by platform, but the choices usually look like this:

Real ticks are large files, so long test periods are slow.

After a test, find the data-quality figure in your test report. The EA backtest report at the top of this page, a past XAUUSD test on the 1-minute chart, shows 100%. That says the price data was complete. It does not prove real ticks were used, so check the line that names the price model as well. With 2165 trades and an expected payoff of 1.41 per trade, it is the kind of many-small-trades strategy where the price model matters most.

Checks before you trust a tick-data backtest

• Use the broker you will trade with. Their ticks carry their spreads. Another source tests another market.
• Confirm the test used real ticks with a variable spread. Generated ticks or one fixed spread throw away most of the advantage shown above.
• Look for gaps around weekends and holidays. Tick files often have missing stretches near the weekly open and on thin holiday sessions. A gap can hide a price jump that would have hit your stop.
• Match the time zone of the tick file to the broker's server time. If they differ, daily candles, session filters and news-time rules in your strategy fire at the wrong hours.
• Make the test period long enough to include high-spread news days. A quiet stretch with no major releases or central bank decisions shows only normal spreads and flatters a scalper.
• Look at trades that open and close inside one candle. Many of these means the result depends heavily on tick order.
• Remember what ticks do not include. Slippage, requotes and execution delay are not in the file. Even a perfect tick test is still hypothetical performance.

FAQ