Britain's Volatile Export Data - ONS Replies

The ONS has replied to my query about the extraordinary volatility of Britain's export data this year. The brief reply is worth retailing in full, in part because it discloses their own (muted) concern about the series.  Actually,  times of significant structural economic change inevitably do tend to throw seasonally-adjusted data out of whack for a while -  Japan's statisticians struggled with these problems for years after the Bubble imploded. . . .  

The nature of a monthly dataset does tend to highlight the volatility of the statistics - but it is true that the Goods series has become particularly volatile as of late.

Although we don't quantify the impact of special events, the anecdotal evidence indicates that a lot, but not all, of the impact comes from specific economic events over the past year (additional factory closures etc).

The issue is that because we have a series of genuine erratic movements, there is an element of negative/positive autocorrelation in the data (as expected). Although we don't doubt the quality of our dataset we're currently looking at our seasonal adjustment and time series compilation techniques to ensure that our treatment continues to be correct.