And a fall is just what we've seen. Official figures show that seasonally-adjusted sales volumes fell by 1.9 per cent in December, reversing months of decent growth with the result that sales are no higher now than they were in July.
But what should we make of this?
One possibility is that sales were strong after the vote to the leave the EU because consumers anticipated that sterling's fall would lead to higher import prices, and so they bought goods before those price rises - in effect, pulling spending from 2017 into 2016. Strong sales in the summer and autumn were, therefore, only temporary and December's drop shows that this process has ended. This implies that we can anticipate weaker spending this year.
This theory, however, is inconsistent with a big fact: shops that sell high-ticket items, such as electrical appliances shops, jewellers and computer shops, haven't enjoyed especially strong sales since the referendum. But it's these you'd expect to have benefited most from spending being pulled forward.
This suggests a more mundane possibility. That 1.9 per cent drop in December was after statisticians had seasonally adjusted the numbers; actual spending volumes in December were 13 per cent up from November. This seasonal adjustment, however, is based upon seasonal patterns in previous years, but this means that if spending was less seasonal than usual in December, statisticians would have adjusted too much and so reported a fall in adjusted sales.
If this is right, underlying spending is still strong, which suggests consumers are optimistic about the future.
But is it right? Again, there's a fact that speaks against this theory. The most obvious reason to think spending has become less seasonal in December is that we've imported 'Black Friday' from the US, and so begun Christmas shopping in November to a greater degree than usual. But this runs into a problem: seasonally adjusted spending actually fell slightly in November. (Its strength came in October.)
All this perhaps warns us of the dangers of over-interpreting one month's numbers: short-term fluctuations tell us little.
We should, therefore pan out and look at the broader data. This tells us that the ratio of retail sales to the All-Share index has been around its post-1996 average - above it in November, below it in December. This matters, as this ratio has been a handy predictor of equity returns in the past. It's therefore pointing to only average returns.
Worse still, the ratio of retail sales to house prices has been below its post-1996 average in the last few months. Historically, this has pointed to below-average growth in GDP, spending and house prices. In this sense, retail sales were weak even before December's drop - which is telling us something slightly worrying.
