Home › FX Trading School › Advanced

Modelling Broker Execution: Pepperstone vs IG

Advanced Updated 14 July 2026 · 9 min read · PipTax education

Modelling broker execution means building a repeatable way to measure how your orders actually behave in the market — not how you assume they behave. This is Module 18 of PipTax FX Trading School, and it builds directly on the total-cost thinking from Module 9 (spread, commission and swap) and the slippage basics from Module 12. Here we go further: turning execution into something you can log, test and compare across brokers such as Pepperstone and IG.

Why Execution Modelling Matters at This Level

By now you should already know that the quoted spread is not the full cost of a trade. Execution modelling is the next layer: it asks what happens *between* the moment you click and the moment your order fills.

Three things determine that gap:

• Latency — the time between your click and the broker's server receiving it
• Requotes or rejections — whether the broker fills at your price, a new price, or not at all
• Slippage direction — whether the difference between requested and filled price tends to help or hurt you

For a scalper or news trader, a consistent 0.3-pip negative slippage bias can erode an edge that looked profitable on paper. For a swing trader holding for days, it barely matters. Knowing which category you're in decides how much effort this module deserves.

This is advanced work because it requires data collection, not opinion. You cannot model execution from a broker's marketing page — you need your own trade logs, ideally hundreds of fills, before you draw conclusions.

Execution Models: Market Maker vs STP/ECN

Understanding the underlying model helps explain *why* execution differs, not just *that* it differs.

• Dealing desk / market maker — the broker may take the other side of your trade internally before hedging. Fills can be very fast but the broker has more discretion over pricing.
• STP (straight-through processing) — your order passes to a liquidity provider automatically. Speed depends on the LP and the broker's routing tech.
• ECN — orders sit in a shared order book with other participants; you typically pay a separate commission and see raw spreads.

Both Pepperstone and IG offer routes that sit across these categories depending on the account type you choose — Pepperstone's Razor account and its MetaTrader server infrastructure is aimed at traders who want ECN-style pricing, while IG's own platform is built to handle its own order flow with its own execution stack. Neither should be assumed superior without testing — the right question is which model fits *your* strategy and instrument, not which brand sounds better.

Building a Simple Execution Log

You don't need institutional tools to start modelling execution. A spreadsheet with these columns will do:

| Field | What to record | |---|---| | Timestamp | To the second, ideally millisecond if your platform shows it | | Requested price | What you clicked | | Filled price | What the confirmation shows | | Slippage (pips) | Filled minus requested, signed | | Order type | Market, limit, stop | | Instrument | Pair and session (London, NY, Asia) | | Volatility note | Was there news within 5 minutes? |

After 100+ entries, calculate:

1. Average signed slippage — is it centred near zero, or skewed? 2. Slippage during news vs quiet periods — most brokers widen spreads and slippage risk around releases 3. Rejection/requote rate — how often did the fill simply fail?

This is the same discipline PipTax's own cost-impact modelling uses at a broader level — see /cost-impact.html for how small per-trade frictions compound over a year of trading.

Comparing Execution Across Brokers Honestly

When comparing Pepperstone and IG — or any two brokers — resist the urge to run five trades and declare a winner. Execution varies by:

• Time of day — liquidity thins outside London/NY overlap
• Instrument — majors typically execute more predictably than exotics
• Account type — a standard account and a raw-spread account on the same broker can behave differently
• Platform — MetaTrader 4/5 versus a broker's proprietary platform can route orders differently even at the same broker

A fair test holds as many variables constant as possible: same pair, same rough time window, same order type, run over multiple weeks on a demo or small live account with each broker. Even then, you're sampling — not proving — typical behaviour, because live market conditions never repeat exactly.

For live, verifiable spread and commission figures to pair with your own slippage data, always check /audit.html rather than relying on memory or old screenshots — pricing changes.

Reading MetaTrader and Platform Execution Reports

Both MetaTrader (common on Pepperstone) and proprietary platforms (as used by IG) log execution data if you know where to look.

• MT4/MT5 trade history shows open/close time and price to the second; exporting to CSV lets you calculate slippage against your own requested-price notes
• Deal properties in MT5 sometimes expose requote counts if your broker's server logs them
• Platform-native journals on brokers' own platforms often show a "confirmed price" versus "requested price" field directly in the deal ticket

The key habit: capture the requested price *before* you see the confirmation. Reconstructing it afterward from memory defeats the purpose of the exercise.

Turning the Model Into a Decision

Once you have a few weeks of logged data, ask three practical questions:

• Does slippage direction correlate with my strategy's edge? If you're a breakout trader and slippage tends to favour fills in your direction, that's a small tailwind; if it works against you, it's quietly taxing every trade.
• Is the rejection rate acceptable for my style? A 2% rejection rate on a strategy taking one trade a week is irrelevant; the same rate on a 50-trade-a-day scalping system is not.
• Does account type change the picture? Compare a standard account against a raw/ECN account on the same broker before assuming the "cheaper-looking" one is actually cheaper once execution is included.

Modelling broker execution is ultimately about replacing assumption with evidence. It won't tell you which broker is "best" — that claim doesn't really exist — but it will tell you, with your own numbers, whether Pepperstone or IG (or any broker) suits how you actually trade. Pair this module's log with PipTax's broker comparison pages at /brokers/index.html and the methodology behind our figures at /methodology.html, then revisit your log every quarter — execution quality can and does drift over time.

Key Risks and Limits of This Approach

Before you over-rely on a small sample, keep these limits in mind:

• Small samples mislead — 20 trades tell you almost nothing statistically valid
• Market conditions change — a broker's execution during calm ranges is not evidence of behaviour during high-impact news
• Demo vs live can differ — some brokers route demo and live orders differently; test on live where possible, with small size
• This is not a guarantee tool — modelling reduces uncertainty, it does not eliminate the real risk that trading carries, including the risk of losing money

Trading with any broker, however well you've modelled its execution, remains risky, and past fill quality is no promise of future fill quality.

Keep going: Audit Index Cost Impact Methodology