What 0.2 pip of worse fills does to a 0.5 pip edge

[Fairman]

What 0.2 pip of worse fills does to a 0.5 pip edge

Small cost changes matter a lot when the edge is small. Here's a worked case.

A scalping system takes 5 pip targets and 5 pip stops, wins 55%, and the figures already include 0.9 pip of average round trip cost. Those are net numbers, so a winner adds 5 pips and a loser takes away 5. Expectancy is 0.55 times 5 minus 0.45 times 5, which is 0.5 pip per trade.

Now suppose fills get 0.2 pip worse on average. Maybe the broker widened spreads, or you moved to a VPS further away, or you're trading a busier hour. The net win becomes 4.8 and the net loss 5.2.

New expectancy: 0.55 times 4.8 minus 0.45 times 5.2, which is 2.64 minus 2.34, so 0.3 pip.

A 0.2 pip change in fill quality cut the edge by 40%. Over 400 trades a month at two lots, that's 80 pips times $20, so about $1,600 a month less.

At 0.5 pip worse, the expectancy goes to zero.

This is why I think scalpers should track fill quality as closely as they track win rate. A system with a small edge can be killed by execution changes that are too small to notice trade by trade.

[FXS HongKong Scalper]

A clear example of how thin edges are. It shows why scalpers need to track their own costs closely. A change in broker or time of day can wipe out most of the profit without any change in the strategy itself. Logging the cost per trade and comparing it each month is a simple way to catch this early. If costs rise by a fraction of a pip, it may be worth trading only at times when spreads are lowest, which is often the London morning for EURUSD.

[Fairman]

The way to measure it is to compare the requested price with the actual fill price on every trade. For market orders the requested price is the bid or ask at the moment you clicked. MT5 records the price in the deals history, and if you log the price you saw, the difference is your slippage. Averaging it over 50 trades gives you a decent figure to plug into the expectancy formula. Logging the price you saw needs a habit or a script, since the platform won't record it for you. A hotkey that writes the current bid and ask to a file when you click is one way to do it.

[Fairman]

Exits count too, not only entries. A stop that fills 0.3 pip worse than its level and a take profit that fills exactly gives an asymmetric cost, which hurts more than equal slippage on both. Looking at slippage separately for entries, stops and targets often shows that the stops are where most of the cost is hiding, especially around news. If your stops are the expensive part, the fix might be placing them where spreads don't spike, or avoiding trading into news. Targets with limit orders are usually filled at the price or better, so they rarely add cost. Stop orders become market orders and take whatever price is there. That difference is structural, not the broker being unfair.

[Fairman]

Sometimes a worse fill is the price of a better trade. A limit order that fills at the level has zero slippage but misses the fast moves, while a market order fills every time with some slippage. Comparing systems only by average slippage can push you towards limits, which might have lower expectancy overall because of missed winners. Fill quality is one input, not the whole picture. Testing both order types on the same signals for a month, if you can, gives the real comparison. Half the trades with limits and half with market orders, alternating, is a simple design. It's slow but it answers the question properly.

[Fairman]

If your edge is this sensitive, look hard at the hours you trade. The same broker can have very different execution quality at 07:00 and at 09:00 London. If one hour has 0.3 pip worse average fills, dropping that hour might improve results more than any change to the entry rules. The data is in your trade history already, it just needs sorting by time. Sorting by hour and pair together might show that one combination is responsible for most of the slippage. Dropping that combination is often easier than trying to improve execution.

[Shadow Trader]

Most platforms let an order carry a deviation or slippage tolerance, often set in points. Many traders leave it at the default and never look again.

That number is a permission slip. It tells the broker how far from the requested price it's allowed to fill. A value of 10 points means anything within 1 pip is accepted. A value of 100 means 10 pips of tolerance.

Set it too wide, and bad fills are authorized in advance. Set it too tight, and orders get rejected during any fast move, which means missing the trade entirely.

The trade-off differs by strategy. A momentum scalper wants fills at almost any price, since missing the move is worse than a pip of slippage. A mean-reversion scalper, who is entering at a precise level, wants tight tolerance, since a bad fill ruins the whole premise.

Some brokers ignore the setting on market execution accounts, where the fill is simply whatever the market gives. The field is present but inert.

It's worth testing: place orders with different tolerances in fast conditions and see whether anything changes.

Does changing the deviation setting visibly change fills on a live account, or is the field just for show?