The scalper's relationship with the spread at the moment of exit

[Fairman]

The scalper's relationship with the spread at the moment of exit

Entry costs are widely discussed. Exit costs get less attention. At the exit, the spread can differ markedly from that at entry, especially when exiting during a fast move or after a news event. A profitable trade by the mid price can end up at a loss once the exit spread is applied.

This matters most for stop-outs. When a stop triggers in a volatile market, the spread is often wide, and the fill price reflects both slippage and the extra spread.

To measure, record the spread at both entry and exit for every trade. Compare the averages. Many find the exit spread higher than the entry spread, especially for losers.

If there is a consistent gap, the target and stop distances should be set with it in mind. Alternatively, adjust the strategy to avoid exits during spread spikes, using limit-based exits when possible.

A complete cost model covers both sides of the trade, with the specific conditions at each moment.

How different are entry and exit spreads in the data, split by winners and losers?

[Fairman]

Logged this for six weeks on EUR/USD. Average entry spread on all trades was 0.2 pips. Average exit spread on winners was 0.2, on losers 0.45. The losers mostly exited during quick moves against me, when the spread was briefly wider. So the real cost of a loss was 0.25 pips higher than the cost of a win, beyond any slippage. At my size and about 150 losing trades in that period, that's around 37 pips, or a few hundred dollars I hadn't accounted for.

[Amanhasnoname]

My own check on this found something a bit different from the average spread story. When I split exits by type, the targets filled with practically no extra cost, as you'd expect from limit orders, but market closes during fast moves were where the spread blew out. Most of my cases were manual exits, because I panicked after a news spike and hit the close button right when the spread was at its widest.

That gave me a rule I find easy to follow. If the spread is more than double its normal value at the moment I want out, I don't close at market unless the stop is about to be hit anyway. Instead I place a limit order a pip or so from the current price and give it a few seconds. Spikes in spread usually fade within a handful of seconds, so waiting often saves several tenths of a pip.

It won't work if you're in a trade that really is running away from you, so I only use it when the stop is still some distance away.