The great like-for-like debate

Like-for-like sales are the usual barometer for the health of a business. But are they valid? By Mark Stretton.

It was a classic triumph of spin over reality. Whitbread published figures for its pub-restaurant brand Beefeater that at face value looked fairly respectable. It had turned the corner. The concept once famous for prawn cocktails, steak with chips and Black Forest gateau was back.

The renewed optimism was driven by like-for-like sales growth of four per cent, which for a mature brand wasn't half bad.

Like-for-like sales are of course a standard measure in retailing, comparing sales on a pub-by-pub basis across an entire estate. Pubs used in this gathering of comparable information will have been owned for at least two trading periods.

So if the Dog and Duck shows sales of £100,000 last year and sales of £104,000 this year, this equates to like-for-like sales growth of four per cent. Do that analysis across an entire estate and you get a pretty good measure of how the management is organically growing its business. It is the barometer for the health of the business, or at least that's the theory.

Underpinning Whitbread's apparent revival of Beefeater was the fact that 50 of the worst performers had been sold from an estate of just over 200, which was not mentioned in the statement. Removing the bottom quarter from annual comparisons of the estate should have lifted like-for-likes dramatically.

The goalposts had moved and despite this massive difference between yearly comparisons the company had only put like-for-like sales into positive territory to the tune of four per cent.

Safe to say the results probably won't stand out as a defining moment on the CVs of directors at Whitbread.

Anybody who follows the industry would know that the leisure group sold a raft of Beefeaters but the lack of disclosure is an example of the spin that is applied to figures released to the market.

Whitbread could keep removing 50 pubs a year to bolster like-for-likes but they might have a problem in three years' time.

The industry has a problem with like-for-like sales figures. The criteria for how they are measured differs from company to company. It could be argued that the way some companies report like-for-like sales is less than robust. Some would say it is a joke.

It is just one of a number of key performance indicators and the different approaches to measuring like-for-like sales make it impossible to compare peer groups and management teams. "The big concern is that the equity market is highly sensitive to like-for-like sales reporting," says Douglas Jack, an analyst at Panmure Gordon. "Yet the approach from company to company is hugely inconsistent."

Whitbread had also spent a significant chunk of cash revamping 24 Beefeater sites out of 150 to Beefeater II, a new improved format for the chain. These were stripped out from comparisons because they had not been trading for more than a full-year in their current form. The ground rules always used to state that like-for-like sales figures should strictly exclude pubs that had seen significant investment of typically more than £20,000 (this varies from company to company but analysts believe it should never equate to more than five per cent of a unit's yearly turnover).

Pubs that see more cash than that will naturally attract more customers and consequently see an uplift in sales as a result of the re-invigoration. Therefore such a pub should be left out of like-for-like sales figures because it would unfairly lift the numbers.

But this rule gives rise to a technique one prominent nightclub boss affectionately calls the "lampshade" phenomenon. You take a badly performing site and spend a small amount of cash on it, giving it perhaps a lick of paint and putting a few new lampshades here and there. This will have little impact on trading but the company then excludes it from like-for-like sales figures because it has seen capital expenditure, and therefore it would be distort the figures. In reality a poor site is removed from comparisons.

But that was then and this is now. Now we have a term called "invested like-for-likes", which includes sites that have seen substantial capital expenditure. You lump in both invested and uninvested pubs into one band, which, let's face it, isn't really comparing like with like.

Although analysts aren't as concerned with invested like-for-likes as one might expect as long as it forms part of a wider picture.

Then there is the situation with Yates Group, where the company reports like-for-like sales just for revamped sites, or 21st Century Yates as it calls it.

Yates is, according to analysts, one of the biggest offenders for reporting inconsistent numbers. "A year ago it was shouting about like-for-likes in its Ha! Ha! Bar & Canteen brand but it failed to disclose the same numbers in its latest update, presumably because they weren't as favourable," said one. "It only tells you what it wants you to hear not what it should be disclosing."

Companies may have different rules but when you give a figure to the market surely it should be calculated on the same basis each year? In the latest trading statement Yates also fails to disclose current like-for-like trading on an uninvested basis - the most basic measure.

Although Yates needs to adopt a more consistent and transparent approach it is by no means in the SFI league of "massaging" like-for-like figures. Before the troubled group hit the buffers it was known for its selective approach in choosing pubs for like-for-like comparison. At one results presentation just 35 per cent of its estate was included - a meaningless measure.

SFI used some classic techniques such as the "lampshade" phenomenon. Another favourite was to identify poorly performing sites that either needed capital expenditure or were to be disposed of and ringfence them, excluding them from the yearly comparisons. This is known as the "we still run them, but don't really want them so won't bother including them" technique. It will be interesting to see how Luminar reports its "ringfenced" disposal estate in forthcoming results.

JD Wetherspoon "back-ends" its pub openings. It opens a load before its year end, so there is an immediate uplift in the comparable figures in the beginning of year three.

However, analysts believe this is acceptable, especially as the company is generally transparent. "Everyone can grow turnover but it's a question of doing it profitably," says Geof Collyer from Deutsche Bank. "If you are quoted, people want to see the earnings but also, importantly, how you've grown the earnings. JD Wetherspoon gives people a string of measures to assess the whole picture including like-for-like profit, which is crucial."

Average profit per pub is also a key measure of how the business is progressing. Mitchells & Butlers' average profit per pub fell seven per cent last year, yet this was not reported.

The point about like-for-like profit is emphasised further by festive trading information from Regent Inns. Over the autumn 2002 period like-for-like sales at its comedy brand Jongleurs were down nearly 10 per cent. This was because it removed some largely uneconomic performances from its schedules - actual like-for-like profits were up 18 per cent.

When Inventive Leisure recently issued a trading update, it reported a like-for-like sales decline of 14.6 per cent. It wasn't the brightest news release from the operator of Revolution vodka bars and predictably its share price took an absolute hammering.

It was similar to when Regent Inns issued results showing a less dramatic but hefty 7.4 per cent decline in like-for-like sales for the half-year to January 2004.

But both companies issue perhaps the most "honest" like-for-like figures, reporting a pure, uninvested figure. Both are doing better than perceived. Average sales across the estates are up as