Many traders analyse on TradingView and execute on MT5. The two may show slightly different prices, which matters for tight setups.
Sources of differences:
Data feeds: TradingView charts may use a different provider than your broker.
Spreads: TradingView often shows bid prices only.
Time zones and session boundaries: daily candles may differ.
Ways to stay in sync:
Use your broker's own feed on TradingView, if available.
Set the same time zone on both platforms.
Place levels with a small buffer rather than exact prices.
Some brokers allow trading directly from TradingView, removing the problem altogether.
For traders who rely on exact levels from TradingView, recording the price difference between the two platforms at a few fixed times over a week gives a realistic buffer value for your setup.
Do you use two platforms, and have price differences ever caused problems?
I measure the TradingView versus MT5 gap at 08:00, 12:00 and 15:00 London for a week and keep the median as my buffer. On EURUSD it sat near 0.4 pips most days and jumped above 1.2 during the first minute after US data.
That buffer goes on every horizontal I draw on TradingView before I send the order on MT5. Exact price copies without the buffer are what caused two early stops that looked perfect on the analysis screen.
Broker feed on TradingView removed most of the drift for me, but session boundaries still disagreed until I forced both platforms to the same UTC offset. Daily levels were the loudest casualty before that fix.
I still will not trade a 4 pip stop off a TradingView level without checking the MT5 print first. The execution platform is the truth for fills.
Have you seen the gap widen mainly around news, or is it noisy all session for your broker pair? Mine spikes at releases and stays quiet otherwise.
One mismatch that a price gap log will not catch is which side of the quote each platform is drawing. Most TradingView feeds plot the bid, and MT5 by default does too, but a buy stop or a buy limit triggers when the ask reaches your level.
So a short with its stop one pip above a TradingView high can be taken out while the bid never prints there. Around the New York rollover the EURUSD spread on many feeds widens to two or three pips briefly, which is enough to tag stops sitting near the overnight range.
I keep the ask line switched on in MT5 and add the typical spread for that hour to any stop placed on the buy side. Sell stops below a low I leave alone, since those trigger on the bid that I can see.
Alerts are the other trap. A TradingView alert fires on the TradingView feed, so a level that just touched there may still be half a pip away on the execution side when I click.
Do you keep alerts on the analysis platform, or move them to MT5 so the trigger and the fill come from one source? I moved mine and the late clicks mostly stopped.
MetaTrader 5's strategy tester offers several modes. Real ticks mode replays actual recorded ticks from the broker's server, not interpolated ones. For scalping strategies, that's the only mode worth trusting.
What it does: the tester downloads tick history, including bid, ask and often last prices, and feeds them to the strategy in sequence, with spreads as recorded. It models order execution against those prices, including variable spreads. Latency can be simulated with a delay setting.
What it doesn't do: it doesn't replicate slippage as experienced live, last look rejections, partial fills, or liquidity limits. Results remain optimistic compared with live trading. The quality of the tick history depends on the broker, and gaps can distort tests.
Good practice includes checking history quality percentage reported by the tester, running the test on several brokers' data if possible, adding artificial slippage and delays, and treating results as an upper bound.
Comparing a few weeks of tester output against live or demo results over the same period reveals the typical gap.
A quick validation: run the same strategy over the same two weeks in tester and on a demo account, then compare the list of trades one by one. Differences in entry times, prices and rejected orders show where the tester's assumptions diverge from reality. Repeating this a few times gives a reliable sense of the tester's optimism for a particular strategy and broker.
Another useful trick is stress testing: increase the simulated latency to 200 or 500 milliseconds and widen spreads by 20%. If the strategy still earns a positive result, it has some margin of safety. If it collapses, the edge was a product of ideal conditions.
A deeper check involves comparing tester output with live results on three dimensions. Trade count: does the live system take the same number of trades as the tester predicts? If not, signal timing or filters differ. Average slippage: how far are live fills from the tester's assumed fills? Rejection rate: how often do live orders fail, and does the tester model that?
Gaps in any dimension point to specific fixes. Fewer live trades than tested may mean the strategy depends on prices that weren't actually available. Larger slippage suggests widening the tester's assumptions. High rejection rates call for adjustments to deviation settings or order types.
Over several rounds, the tester's configuration can be tuned until its results sit close to live results, with a conservative margin. A calibrated tester is a research instrument. An uncalibrated one is a source of false confidence.
Keeping a calibration log, with date, settings and observed gaps, makes the process systematic, and shows whether execution quality drifts over time.
What gap has been observed between real-ticks test results and live trading for the same dates?
