forex-scalping.com

One-way slippage

Slippage should be random. Some fills come in better than requested, some worse, and the average sits near nothing.

A pattern shows up in some accounts though. Negative slippage appears on nearly every winning-direction entry and positive slippage almost never appears at all. Over a month the average cost per trade is more than the spread itself.

Possible reasons are not always sinister. A broker routing orders through a single slow bridge will naturally lose some price on fast moves, but those losses would still go both ways if the market were choppy. One-directional slippage points toward either a feed that is always a step behind or a filter that only passes trades that are good for the house.

The test is easy enough. Pull the platform's trade history, add a column for the difference between order price and fill price, and split it by direction and by time of day. Patterns appear quickly.

Then run the same exercise on a second broker with a small account. Same strategy, same hours, different venue. If the lopsidedness disappears, the first broker's routing is the culprit and not the strategy.

Is there a minimum sample size where a slippage pattern becomes convincing enough to switch brokers over?

This one is controversial, and plenty of people dismiss it as paranoia. Still, enough accounts report the same pattern to deserve a mention.

An account trades well on the first few weeks with fills close to the quote. Then, as profits build, execution seems to degrade: slightly more slippage, slightly more rejections, wider effective spreads. Nothing dramatic, just a drift.

There are innocent explanations. Position size may have grown, which naturally affects fills. The strategy may have started trading in thinner conditions. Market volatility might have changed.

There are less innocent ones too. Some brokers classify clients by profitability, and in a B-book model a consistently winning client is a cost. Routing, delay, or price feed can differ across client groups.

Proving it is hard. The way to test it is to open a fresh small account with the same broker, run an identical strategy during the same hours, and compare fills over a few weeks. Or open the same strategy at a second broker.

If the numbers match, the worry goes away. If they diverge, that's valuable information.

One more complication: scaling up makes the same problem look different. A trader who starts with 0.05 lots and moves to 2 lots a few months later will naturally see slippage change. The new size touches deeper layers of the book, and fills fragment. That is a size effect, not a punishment, and it needs separate treatment.

To separate the two, a fixed-size "probe" trade is useful. Every day, at the same quiet moment, a tiny order of unchanged size goes through and its slippage is recorded. If the probe's fills stay stable while the main account's worsen, size is the cause. If the probe's fills also worsen over time, something about the account classification may be changing.

Records beat suspicion. A month of probe data on a spreadsheet is far more persuasive to a broker's support team than a general complaint.

Would a side-by-side test with a fresh small account settle it, or is that wishful thinking?

One way slippage describes a pattern where fills are consistently worse than requested but never better. In a fair execution model, slippage should occur in both directions, since the market can move for or against the trader between the order request and the fill. If only negative slippage appears over many orders, it suggests the broker or its systems are passing on unfavourable price changes while keeping favourable ones. Some regulators have taken action against firms for this kind of asymmetric practice. Measuring slippage on your own trades, with enough data to draw conclusions, is the only way to know whether it is happening at your broker.

I went through this with a small account at my broker a couple of years back. Looking at about 400 trades, negative slippage was visible on nearly everything and positive slippage showed up on maybe one in twelve entries. The average was a bit over half a pip, more than the spread on that pair.

I did the test you describe by opening a small account elsewhere and running the same entries at the same times. The second broker's slippage was much more balanced, with positive fills about a third of the time, which told me it was not just my strategy trading into fast moves.

I split the data by hour and found the worst of it was around the London open and news minutes, so part of it was normal. Even after removing those, a gap remained. I ended up moving most of my volume to the second broker, and the numbers have been steadier since. Whatever you find, keep the spreadsheet as evidence.