Market Data Feeds EA: Tick Data, Quality and Cost

An expert advisor is only as good as the data it reads. Signal logic, stops and risk management all operate on the price stream that reaches the terminal, and a stream that is delayed, incomplete or wrong infects every decision built on it. The market data feeds EA builders choose — tick data from the broker or an external provider, and historical archives for backtesting — determine whether the robot trades the real market or a plausible imitation of one. It covers what data an EA needs, where it comes from and what it costs.

Market data feeds EA: what an expert advisor actually needs

An EA consumes two kinds of data: the live stream it trades on and the history it was tested on. The first decision is granularity. An EA built on indicators calculated from one-minute candles needs M1 history and current M1 bars; its decisions change slowly and it never sees individual trades. A scalping or news EA reacts to what is happening right now — a burst of volume, a spread spike, a price jump — and needs tick data: every trade with time, bid, ask and volume, because a single tick can carry the entire signal.

The choice shapes everything downstream. Tick-driven EAs need faster data, faster execution and more care with the tester; M1-driven EAs are more forgiving. Decide the granularity from the strategy’s reaction time, not from whatever data happens to be free.

Market data feeds EA: broker feed versus external feeds

The default feed is your broker’s, delivered through the terminal with the account. Its virtues are price and consistency: it is free, and it is the same stream your orders fill against, so signals and execution agree by construction — the EA never sees a price it cannot trade. Its faults are the broker’s faults: reconnects during maintenance, quiet-period gaps, and spreads that widen beyond what the feed’s mid-price suggests.

External feeds are the alternative, for a price: independent tick streams from dedicated data vendors, uniform across symbols and brokers, and often cleaner historically. The trade-off is a second source of truth. If the external price and the broker’s price disagree — they will, at times — the EA can signal on one and fill on the other. External feeds suit testing and analysis; broker feeds suit live execution. For most retail EAs, the live data needed to trade is already included in the account.

Market data feeds EA: tick quality and backtest accuracy

Backtest accuracy is capped by data quality, and this is where most systems quietly lose their edge. The tester can model bars from M1 data and approximate what ticks would have looked like, but a modelled tick is a guess: it assumes smooth progression between bar endpoints, uniform spread and no reaction to liquidity. The backtesting guide shows how far those assumptions travel into your results.

Real tick history records every tick with bid, ask and volume, letting the tester reproduce fills, slippage and spread behaviour instead of assuming them. The gap between a backtest on real ticks and live results is usually small; the gap between a backtest on modelled M1 and live results is where the “backtest was beautiful, live was awful” story comes from. Buy or build real tick history before you trust any optimisation.

Market data feeds EA: historical data sources

History comes from three places. The broker’s archive is free and convenient, delivered through the terminal’s history centre — but it is shallow, often a year or two of M1, and it is cleaned and filtered in ways the vendor does not document. External archives sell decades of true tick data per symbol, in consistent formats, at one-off prices. And exchange or market data providers publish official archives for the instruments they list — authoritative, though aimed at institutions.

Whatever the source, plan the data pipeline before you need it. Configure the terminal to download and store history for the symbols and periods your EA uses, set the correct timezone alignment, and keep a separate, untouched copy of the raw data for validation. The MT5 installation guide covers the settings that control what the terminal keeps.

Market data feeds EA: what data costs

The honest summary: the data most retail EAs need is free. Live broker feeds are included with the account, and M1 history is bundled with the platform. Costs appear at three points. External live feeds for professionals cost tens to hundreds of pounds per month, aimed at traders whose strategy needs multi-venue or millisecond data. Exchange-traded instruments carry exchange real-time fees — futures data from the major exchanges is not free. And historical tick archives are typically a one-off purchase per symbol.

Spend accordingly: paid history improves backtesting realism, which protects live money; paid live feeds improve latency, which matters only to fast strategies. For most EAs, the first purchase should be history.

Market data feeds EA: common data problems

Even good feeds arrive with structural faults. Gaps are the big one: weekend gaps, news gaps and broker maintenance gaps leave the EA with a jump it cannot trade through, and the tester handles them differently from the live feed. Weekend spikes are stale quotes printed after the close — ticks that represent no tradeable market, yet they corrupt indicators and trailing stops. Rollover gaps on futures and CFD expiries shift the price series discontinuously, and an EA unaware of the roll will see a phantom loss or gain.

Adjusted versus unadjusted history is the quiet problem: dividend and roll adjustments make long histories look continuous but break the price levels the strategy trades. The CFTC and FCA describe the environment these quirks live in. Treat data cleaning as part of the build: verify each feed’s behaviour at known gap points before the EA trades on it.

Frequently asked questions about market data feeds EA

Does my EA need tick data or is M1 data enough?

It depends on the strategy. Trend and swing EAs built on M1 indicators work fine with M1 history and candles, because their decisions change slowly. Scalping and news EAs react to individual trades and need real tick data with bid and ask prices, because the tick stream is the market they trade.

Is my broker’s free data feed good enough?

For live trading, usually yes — the broker’s feed is the same stream your orders fill against, so signals and execution agree. For backtesting it is weaker: free history is often shallow, cleaned or filtered, and modelling M1 approximations instead of real ticks is a major source of backtest error.

Why do my backtest results differ so much from live results?

Mostly because of data. Free history is often missing ticks, spreads are modelled rather than measured, and weekend or news gaps are handled differently in the tester. Using true tick history and correct spread and slippage settings closes most of the gap; the backtesting guide covers the details.

How much does good market data cost?

Live broker feeds are free with an account. External live feeds typically cost tens to hundreds of pounds per month, and exchange-traded futures require exchange real-time fees. Historical tick archives are usually a one-off purchase. For most retail EAs, paid history improves backtesting far more than paid live data improves trading.

Feed the system properly. Re-run your strategy on real tick history with the backtesting guide, set up history storage with the MT5 installation guide, and for deployment-ready automation start with the automation playbook.

Risk disclosure: Trading foreign exchange, commodities, CFDs and cryptocurrencies carries a high level of risk and may not be suitable for all investors. Past performance is not indicative of future results. AlgoTM provides trading tools and technology only and does not provide investment advice, portfolio management or guaranteed returns.