forex-scalping.com

Limit one pip in front of the level or right on it

I'm curious how others handle this one, because I go back and forth.

Say there's a clear support at 1.2650 on GBPUSD from the Asian low. I want to buy the first test in London. If I put the buy limit right at 1.2650, I need the ask to trade there, which means the bid has to drop to about 1.2649 with a 1 pip spread. Lots of tests stop a pip or two short of the line. Price reacts at 1.2652 and goes, and I'm left watching.

If I put it at 1.2652, one or two pips in front, I get filled more often. But on the tests that go straight through, my entry is worse, and my stop below the level is now 2 pips further away in loss terms, so I have to cut size to keep the dollar risk the same.

What I'd like to know is whether anyone has tracked the fill rate both ways on the same level types. Has front-running by a pip or two improved your results, or did the worse price on the failures eat the extra fills? I suspect it depends on the pair and the session, but I'd love some actual numbers.

We put limits one pip in front for the same reason. On GBPUSD the spread matters too: with a 1 pip spread, a buy limit at 1.2650 needs the ask at 1.2650, so the bid has to reach 1.2649. One pip in front means the bid at 1.2650 fills you.

From our logs, the fill rate rose a lot with a one-pip offset, and the average result per trade hardly changed. Going further, two or three pips, started to include more trades where the level never really held. So for us, one pip plus spread is the sweet spot.

One thing I tried was to keep the same stop distance and slide the whole trade forward. Entry at 1.2652, stop at 1.2644, target 9 pips above at 1.2661. That keeps the reward to risk the same, but the stop now sits only 6 pips under the level instead of a clean distance below it, so wicks through 1.2650 hit it more. In my notes that version did worse than either of the other two. The level is the reason for the stop location, and moving the stop to fit the entry undoes that. The cleanest fix was to accept the miss rate.

A middle route is splitting the order. Half the size at 1.2652 and half at 1.2650. You get some exposure on the shallow tests and full size on the deep ones. The average entry is a pip better than the front-run version when both fill. It's more orders and more commission per round trip on some brokers, since a few charge a minimum per ticket, so check the commission schedule before going this way on small sizes. One more detail on the split: if only the first half fills and price runs, you've got a winner at half size, which is fine. The annoying case is when only the shallow half fills and the trade fails. The loss is smaller, but the setup was still wrong.

Session matters a lot here, I think. During London open, levels from the Asian range get swept more often than they get respected exactly, so the front-run order fills and then gets stopped as price runs a few pips under. In the New York afternoon, when things are slower, shallow tests that stop short are more common. If I had to guess, front-running works better in quiet hours and waiting for a sweep works better around the opens. I'd love someone to check that with real data. Logging the session next to each fill would answer it.

Whatever you choose, write the rule down before the session and stick with it for at least 50 trades. Switching between front-running and exact placement depending on how you feel that morning makes the results impossible to judge. You'll remember the missed fills on the exact-level days and the bad fills on the front-run days, and both versions will look worse than they are. Fifty trades of each is a small sample, but it's a lot better than mixing them. Keep a tally of misses as well as fills, since the misses are the part people forget to count.

Two traders compare charts at the same moment. One shows a three-pip wick at 14:32. The other shows none. Who's right?

Both. Each broker's chart is built from that broker's own quotes. A single provider may have posted a brief outlier price that was filtered by the other broker's aggregation. Or one platform reported a spread spike that pushed the bid down momentarily.

On quiet pairs, these differences are common, small, and mostly irrelevant. At key levels, they matter. A stop placed a pip beyond a visible wick on one chart may sit right in the path of a wick that exists only on another broker's feed.

For strategies that use highs and lows as signals, such as breakouts of the prior bar's extreme, different feeds produce different signals. A breakout on broker A is a non-event on broker B.

Worth remembering when reading forum posts that say "price hit this level exactly." On whose feed? The trade was triggered on the trader's own server, and that's the only one that counted.

When wicks matter, comparing the same minute on two brokers' charts takes just a moment. A difference of more than half a pip at the high or low is common during active moments. A strategy that triggers on those highs and lows should be tested on the feed it will actually trade.

When the broker's chart shows a spike that nobody else sees, is it worth acting on or ignoring?