Buying the rumour before the data

[Fairman]

Buying the rumour before the data

Markets often move in the hours before a scheduled event as traders position. The direction usually reflects the consensus expectation, and the move can be significant.

When the news arrives and matches expectations, price often reverses, as those who positioned take profit. The "buy the rumour, sell the news" pattern is a staple of trading folklore.

It is not reliable. Sometimes the news confirms the expectation and the market keeps going. Sometimes the expectation was wrong and the position unwinds violently.

For a scalper, the more interesting question is what pre-news positioning does to liquidity. A market that has drifted a long way in one direction before data is a market with a lot of crowded stops and trapped positions, and the news can trigger a fast unwind.

Watching how price behaves into the release, rather than just reading the forecast, gives clues. A steady one-way drift with narrow ranges suggests heavy positioning, and the reaction can be sharp in either direction.

A practical test: log the price drift in the six hours before each major release, then the move in the first ten minutes after. Over a year, the relationship between the two says whether fading pre-news drift was a good idea for the pair.

How has price tended to behave in the hours before major events, and does it predict anything?

[Amanhasnoname]

I kept a simple two-column note for about ten months, and the useful part was not the direction of the number. For thirty-eight red-folder dollar events I logged the drift from six hours before down to five minutes before, then the first ten minutes after the release. Twenty-two of those prints were preceded by a one-way drift of more than twelve pips. Fourteen of those twenty-two gave back at least half that drift in the first ten minutes, whether the figure matched the forecast or not. The other eight kept going.

When the pre-news tape was choppy, with less than eight pips of net drift, I could not find a relationship worth using. The pattern that did stand out was a steady grind with shrinking ranges: eleven of fifteen of those unwound after the release in my sample.

I do not fade every pre-news drift automatically, but I did stop adding to that drift in the last hour before the number. Those were the crowded positions, and after spread they paid me nothing.

[FXS London Scalper]

Buying the rumour before the data Markets often move in the hours before a scheduled event as traders position.

On EURUSD this shows up clearly before ECB meetings. If the euro has already run up into a widely expected hawkish decision, the actual announcement often marks the high. Measuring the move in the 24 hours before the event gives a sense of how much is already priced.