Why it happens
When you click buy, the order travels to the broker's server, the broker checks the price and fills it. Between the click and the fill, a few milliseconds or a few hundred, the price may have moved. If it moved against you the fill is worse than the price you saw; if it moved in your favour the fill is better. Market orders, stops that have triggered and orders placed in a moving market are all exposed to it. Limit orders are not, because a limit order is an instruction to fill at the stated price or better, and is simply not filled if the price is not available.
When it is largest
Slippage is largest when prices are moving fastest and liquidity is thinnest, and those often coincide. The seconds around a scheduled news release, when spreads widen and prices jump, produce the worst fills of the trading week. The Sunday open, when the market reacts to the weekend's news with few participants, is another. The rollover period around 5pm New York, when many liquidity providers pull their quotes, and the last hour on Friday are quieter versions of the same thing. A trader who avoids market orders at those times avoids most slippage.
Positive and negative
In a fair execution system slippage is symmetrical: over many orders the market moves for you about as often as against you, and a broker's execution statistics, where published, show roughly equal positive and negative slippage in normal conditions. Asymmetric slippage, where orders are filled at the worse price when the market moves against you but at the requested price rather than the better one when it moves for you, is a server setting rather than a market effect, and regulators in the United States, Europe and Australia have fined brokers for it. It is invisible on any single trade and obvious across a hundred.
How to measure your own
Every platform records the requested price and the filled price for each order. Over a month, list the difference for each market order and stop, separate normal hours from news, and look at the distribution. Slippage that is small and goes both ways is the market. Slippage that is only ever negative, or that appears only on winning exits, is worth raising with the broker and, if it persists, with its regulator. The execution quality paper in the research section explains what the published statistics look like at brokers that disclose them, and the broker execution models article explains why a broker's model matters to this.
Educational content is general information and does not consider your objectives, financial situation or needs. Forex and CFD trading involves significant risk.