Fingers are always pointed at the Office of National Statistics but the Confederation of British Industry seems to live on a different planet.

Fingers are always pointed at the much-maligned Office of National Statistics (ONS), but it now seems to be the Confederation of British Industry (CBI) that lives on a different planet from the rest of us.

Today’s British Retail Consortium (BRC) retail sales survey for May reports that like-for-like sales were literally flat last month versus last year, and given that May 2014 was nothing to write home about (+0.5% like for like), that is a disappointing performance.

Volume versus value

But is this the same month that the CBI was talking about back on May 26 in its monthly distributive trades survey, when it trumpeted: “Retail sales ramp up as expectations hit 27 year high”?

Well, yes, it was, but given that the CBI survey was conducted between April 30 and May 15 and only 63 retailers took part, it is surprising that people gave it any credence at all, not least as it only measures perceived shifts in “volume”.

As a guide to what is really happening on the high street, focusing on “volume” is surely pretty meaningless at a time of price deflation, as retailers focus on the “value” of sales not the “volume”.

Volume throughput is important for supply and distribution networks and it is better to be positive than negative, but retailers want both the value and the volume of sales to be up on a like-for-like basis, to help offset the pressure on operating costs and gross margins.

CBI’s mid-month snapshot

However, the CBI survey is widely followed by City economists and, to be fair, the CBI Distributive Trades survey does have a surprisingly good long-term record over time of tracking the ups and downs of the official retail sales figures.

“The CBI survey is not quantitative, it is little more than a snapshot of mid-month retail sentiment”

Nick Bubb

But it is important to remember that the CBI survey is not a quantitative survey, notwithstanding some of the coverage it gets in the press every month. It simply measures the percentage balance between those CBI members who think that volume is up less those who think volume is down and as such, it is little more than a snapshot of mid-month retail sentiment.

Thus last month 60% of the CBI respondents said that “volumes were up” on a year ago and only 9% said that volumes were down, giving a “percentage balance” of +51%, a strong pick-up on the previous month’s outcome of +12% (which was distorted by Easter) and above the expectation of +40% a month before. And most retailers expect sales volumes to grow again this month (63% expect them to rise and only 4% to fall).

Bizarrely, despite the tiny sample sizes, the CBI even highlights sub-sector reported growth in sales volumes, with grocers reporting a +50% balance last month, non-specialised stores a balance of +98% and recreational goods a balance of +100%, although hardware and DIY reported a negative balance of -67%.

But as volume “should” be up on last year at present for most retailers, these sub-sector CBI figures are hardly worth the paper they are written on.

There are, of course, three monthly national measures of total retail sales, including food and non-food: the CBI survey, the BRC-KPMG survey and the ONS figures.

BRC’s unadjusted monthly monitor

The best thing about the monthly BRC retail sales monitor is that it is based on real figures, which are unadjusted and unrevised.

“The monthly BRC retail sales monitor is based on real figures, which are unadjusted and unrevised”

Nick Bubb

As well as receiving sales information from its members, the BRC also receives food and drink sales data from the The Institute of Grocery Distribution (IGD) and these are aggregated by the hard-working accountants KPMG, with a gross sales and like-for-like sales measure, split between food and non-food (which is itself split down into many non-food sub-sectors).

The problem with the BRC figures is that they are only for the big retailers that are members of the BRC – they don’t include small independents – and there are some big retailers who apparently still don’t give their figures to the BRC (Amazon and Arcadia to name but two, by all accounts).

The BRC is also unashamedly an industry lobby, so it is inevitably prone to moan about how difficult life is for retailers in order to influence government policy.

“Official” figures from the ONS?

So what of the “official” retail sales figures? In theory these are far more comprehensive than any other measure, as the ONS’s monthly survey of 5,000 businesses in Great Britain includes all the largest retailers (retailers employing more than 100 people or with annual turnover of greater than £60m) and a representative panel of smaller businesses (which employ fewer than 99 people).

The ONS tends to focus, irritatingly, on month-on-month movements in “seasonally adjusted sales volume”, but it also produces detailed sub-sector breakdowns for the year-on-year change in “sales value”, non-seasonally adjusted, so these are the figures to look at.

The glaring problem with the ONS figures is that they are based on samples, depending on how many retailers have returned the monthly forms. In April, the response rate, not untypically, was only 63% of the forms, although these retailers presented 86% of the total turnover of the industry.

The reason why the ONS monthly figures are often rather flaky is that the ONS makes estimates for the retailers that haven’t returned their forms (the smaller retailers) and then adjusts the overall figures subsequently for “late returns”. The fact that these adjustments nearly always seem to be downward reflects the old adage that “bad news comes late”.

It is also easy to mock some of the ONS sub-sector breakdowns, for example the notoriously unreliable series for carpet retailers’ sales.

“Whilst by no means perfect, the ONS retail sales figures are too comprehensive to be dismissed”

Nick Bubb

And yet the ONS has worked hard in recent years to improve its coverage of online retailers (many of which are too small to be BRC members) and it is possible that it picks up the growth of start-up businesses missed by other surveys.

That may explain why the ONS tends to say that retail sales are growing by a bit more than the BRC says: thus in April there was a 0.9% increase in ONS retail sales (sales value, non-seasonally adjusted, excluding petrol), versus the 1.3% decline (missing Easter) in gross terms (including new store space) in the BRC’s survey for April.

We will find out on June 18 what the ONS made of retail sales in May (the BRC said that sales were up 1.1% overall) and indeed whether it revises down the April figures.

Such is the reporting cycle that just a week later, on June 25, we get the CBI Distributive Trades survey for “June”.

In the past the ONS has been a figure of fun in the industry because of its unreliable retail sales figures. However, whilst by no means perfect, the ONS retail sales figures are too comprehensive to be dismissed.

Perhaps it is now the CBI that is living on a different planet from the rest of us, as it drools over the growth in “volume”, and its monthly distributive trades surveys will now be cruelly mocked.

  • Nick Bubb is an independent retail analyst