A TradingView backtest and an MT5 backtest can produce different results because the platforms may use different price feeds, intrabar paths, spread and cost models, calculation times, contract specifications and strategy implementations. TradingView tests Pine Script through its broker emulator; MT5 tests native MQL5 Expert Advisors. Reconcile the first differing decision or fill before comparing totals. Live execution through alerts introduces additional stages.
TradingView versus MT5 backtests at a glance
Scroll horizontally to read every column.
| Layer | TradingView | MT5 Strategy Tester |
|---|---|---|
| Implementation | Pine Script strategy and broker emulator. | Native MQL5 Expert Advisor. |
| Price events | Chart data, with optional lower-timeframe historical detail. | Recorded ticks or a selected generated-tick mode. |
| Execution assumptions | Order timing, slippage, commission and limit-fill settings. | Tick mode, bid/ask data, delay and account settings. |
| Evidence produced | A list of simulated strategy trades. | A native-EA simulation and tester records. |
| External alert path | An alert can be based on a condition or simulated order fill. | A historical EA test does not establish webhook delivery or broker execution. |
TradingView documents its emulator and calculation settings; MetaQuotes documents EA testing and price-generation modes.[1][2][3] Place the comparison inside the backtesting process, with one saved rule specification for both implementations.
Are you comparing the same rules or only the same idea?
“Buy when the average rises” is not a complete shared specification. The two implementations must agree on the price input, lookback, initialisation, comparison boundary, evaluation time, entry type, exit rule, sizing and position state. A difference in any of these can change which orders are requested.
Write expected cases before reconciling reports. Include an equal-to boundary, a repeated signal, a session boundary and an existing-position case. Compare the intermediate indicator values and true-or-false decisions. Identical strategy names or parameter labels do not establish equivalent code.
Position handling deserves special attention. TradingView supports settings such as pyramiding and entry reversal behaviour. MT5 distinguishes netting, with one common position per symbol, from hedging, which permits multiple positions.[1][4] Two opposite instructions can therefore affect positions differently unless the intended handling is specified.
Why do price feeds, sessions and timezones matter?
PineConnector's price-difference explanation notes that TradingView and MetaTrader can use different feeds, bid/ask quotes, spreads and timestamps.[7] A matching ticker label does not establish an identical price series or tradable contract.
Compare raw bars around the first differing signal: opening time, closing time, open, high, low and close. Identify whether the relevant price represents bid, ask, last or another feed convention. Do not assume every TradingView symbol has the same quote basis.
As of 25 September 2026, TradingView documents that changing the chart's display timezone does not change Pine Script calculations. Calendar variables and time functions normally use the exchange timezone unless the function specifies another one.[5] MT5 displays trade times according to the broker server's timezone.[6]
Keep the original timestamps and convert copies to a common clock. Also compare the trading session, included weekdays, bar boundaries and daylight-saving treatment. Equal displayed times are insufficient if one record labels the bar opening and the other labels a signal at its close. See the broker-time and TradingView clock guide.
How do intrabar paths and Bar Magnifier change fills?
TradingView's default historical fill model infers a path from the chart bar's OHLC values. If the open is nearer the high, it assumes open → high → low → close; if nearer the low, it assumes open → low → high → close.[1]
The setting associated with use_bar_magnifier, commonly called Bar Magnifier, uses available lower-timeframe OHLC data to refine historical detail. Record that setting separately from script execution settings, which determine when the strategy recalculates. Lower-timeframe bars still do not establish the receiving broker's exact tick sequence.[1]
MT5 can use recorded broker ticks or generate events from minute bars. Real ticks can carry changing spreads within a minute; generated ticks use the minute bar's spread. Missing recorded tick minutes can be reconstructed from minute data.[2] The MT5 modelling-mode guide explains the remaining restrictions.
A bar that touches both an entry and an exit level does not, by itself, establish their order. Likewise, a touched limit price does not establish a live fill. TradingView provides a limit-fill assumption requiring price to move beyond the limit by a specified number of ticks.[1]
Check the chart type too. TradingView warns that non-standard charts can use synthetic prices for simulated fills. Document any special fill setting and the prices used by the signal itself; changing fill prices does not automatically make every signal input a standard candle value.[1]
Worked example: one buy instruction, two price models
TradingView's slippage setting is a fixed number of ticks. For the illustrative market buy below, the adjustment is added in the adverse direction.[1]
Simulated buy fill = base fill price + slippage ticks × tick size
Illustrative prices and assumptions, not a recommendation or an executed trade. Assume equivalent buy decisions, a common evaluation interval and an instrument with a tick size of 0.25. Commission and other charges are excluded from this entry-price comparison.
| Component | TradingView illustration | MT5 illustration |
|---|---|---|
| Input at the chosen fill event | Base emulator fill price 100.50. | Bid 100.50; ask 100.75. |
| Additional assumption | Two ticks of adverse slippage. | No quote movement before the simulated buy fills at ask. |
| Calculated entry price | 100.50 + 2 × 0.25 = 101.00. | 100.75. |
The MT5 quote spread is 100.75 − 100.50 = 0.25. The illustrative TradingView adjustment is 2 × 0.25 = 0.50. The resulting entry-price difference is 101.00 − 100.75 = 0.25, even with equivalent buy decisions.
MetaQuotes documents that buys generally execute at ask and sells at bid.[4] The example supplies its assumed quotes and fills; it does not claim every MT5 order receives the displayed ask. The point is to expose different assumptions before interpreting a report difference as different signal quality.
How should costs, quantity and contract specifications be matched?
Record spread, commission, slippage and overnight charges separately. TradingView applies no commission when none is specified, and its default slippage input is zero.[1] A zero input is a simulation assumption, not evidence of zero real-world cost.
State whether any fixed adjustment is intended to approximate spread, slippage or both. If a price stream already uses executable bid and ask prices, adding the same spread cost again counts it twice. Check entry and exit conventions, not just a single round-trip total.
Commission inputs need matching units and charging events: per order, per unit or percentage of transaction value. MT5 permits custom commissions and account settings. Its profit-calculation-in-pips option excludes commission, swap and margin control, so that report cannot be compared directly with a full account-currency model.[1][3]
Compare economic quantity before comparing currency totals. TradingView's report records symbol point value, currency, tick size and strategy order-size settings.[1] PineConnector's syntax reference asks users to obtain the broker's contract size, tick value, price increment and volume limits.[9]
A quantity labelled “1” is not a complete unit specification. Record what one strategy unit and one broker lot represent, the conversion currency and permitted volume step. PineConnector's documented sizing methods are separate instructions; do not assume they inherit TradingView's simulated position size.
Why can live alert execution differ again?
TradingView alerts can follow a script condition or a strategy's simulated order fill. These represent different trigger points. Alerts operate on realtime bars, and TradingView saves the script, inputs, symbol and timeframe when an alert is created.[10] Editing the chart does not update that saved alert context.
With default next-tick order processing, an order created at a bar's close cannot fill until the next available tick. Enabling same-close processing changes the simulation, but an external alert still follows the relevant event. At session end, an actual order may wait until the market reopens.[1]
PineConnector's converter documentation states that supported strategy alerts follow TradingView's simulated fills. The integration does not reproduce TradingView's sizing or place its simulated stops and targets as protective broker orders. Separately configured broker exits can occur earlier, and the adapter does not resynchronise positions.[8]
A broker-side stop and a strategy-only exit therefore need separate records. MetaQuotes documents that broker SL/TP are stored and executed on the broker server.[6] A line or exit order inside TradingView's emulator is not evidence that the broker holds that protection.
Trace condition true → alert triggered → webhook delivery → PineConnector processing → EA request → broker acceptance → executed deal → resulting position. Use PineConnector's setup test and the backtest-to-verified-demo procedure to reconcile actual records.[11]
What is the reconciliation checklist?
- Freeze both runs: save versions, parameters, feeds, dates, chart types, execution settings and account assumptions.
- Normalise identifiers: map symbols, units and timestamps without discarding their original values.
- Find the first different input: compare bars, indicator values, warm-up history and eligible sessions.
- Compare the first different decision: identify the condition, evaluation event and prior position state.
- Compare matching order requests: inspect side, type, quantity, entry, stop and target. A missing trade belongs in the comparison too.
- Explain different fills: inspect quote side, intrabar sequence, limit assumptions, delay and slippage inputs.
- Reconcile accounting: align partial fills and exits, fees, currencies and the definition of a completed trade.
- Then inspect totals: aggregate only after the differences in individual events have an explanation.
Changing one fill can change later signals. If the rule permits entry only while flat, an earlier exit changes which later candidates are eligible. Once position states diverge, later trade-count differences may follow mechanically from the first mismatch.
Does agreement mean either backtest is accurate?
Agreement only establishes that the compared implementations and assumptions produced similar observations in the tested sample. Both can share look-ahead bias, omitted costs or research choices fitted to that history. Disagreement is also not proof that either platform is defective.
Define accuracy against a specific question: faithful rule calculation, plausible historical fills, or observed broker execution. Each requires different evidence. No single report can settle all three, and matching historical totals does not establish future results.
Frequently asked questions
Why do TradingView and MT5 backtest results differ?
TradingView and MT5 can differ in source data, strategy implementation, calculation timing, intrabar price models, quote sides, transaction costs and contract units. Compare the first differing input, decision or fill using saved settings and trade records. Aggregate totals alone cannot identify which layer caused the difference.
Is MT5 more accurate than TradingView for backtesting?
Accuracy depends on the implementation, data and assumptions being tested. MT5 can use recorded broker ticks, subject to coverage and fallback rules. TradingView can use lower-timeframe historical detail for its emulator. Neither feature establishes that the strategy code is correct or that a future broker request will match a simulated fill.
Does Bar Magnifier make TradingView match MT5 real ticks?
Bar Magnifier uses available lower-timeframe OHLC data to refine TradingView's historical price model. MT5 real-tick testing uses ticks accumulated by the broker, with documented fallback when tick data is missing. Different feeds, spreads, calculation settings and execution assumptions remain, so enabling additional detail does not establish identical tests.
Why do PineConnector demo trades differ from a TradingView backtest?
A TradingView backtest contains simulated decisions and fills. PineConnector demo execution also involves alert triggering, webhook delivery, signal processing, the EA request and broker handling. Quotes, costs, volume and broker-side exits can differ. Match each alert to its processing record and actual broker deals before diagnosing the discrepancy.
Reviewed 25 September 2026. Facts were checked against the linked sources on that date. Nothing in this article was tested on a trading account and no code was compiled.
Related reading
- How to backtest a trading strategy
- MT5 Strategy Tester modelling modes
- Look-ahead and survivorship bias
- Why TradingView prices differ from a broker's
- From a TradingView backtest to a verified demo trade
Sources
- TradingView – Strategies: broker emulator, historical detail, costs, properties and order timing, accessed 25 September 2026.
- MetaQuotes – Real and Generated Ticks, accessed 25 September 2026.
- MetaQuotes – Strategy Testing: account, symbol and execution settings, accessed 25 September 2026.
- MetaQuotes – Basic Principles: quote sides, netting, hedging, orders and deals, accessed 25 September 2026.
- TradingView – Time: chart and exchange timezones, accessed 25 September 2026.
- MetaQuotes – Executing Trades: timestamps, volume and broker-held stop levels, accessed 25 September 2026.
- PineConnector – Frequently asked questions: data, timing and price differences, accessed 25 September 2026.
- PineConnector – Pine Script converter: supported behaviour and separate broker exits, accessed 25 September 2026.
- PineConnector – Syntax: broker specifications and explicit volume methods, accessed 25 September 2026.
- TradingView – Alerts: event types and saved script context, accessed 25 September 2026.
- PineConnector – Test your setup, accessed 25 September 2026.
PineConnector executes the instructions you send it. It does not select trades, manage money, or hold funds. Trading carries risk, and past performance of any strategy does not indicate future results.
