Mark Brumby, of Langton Capital, gives his view on a dramatic week at Patisserie Holdings and the questions yet to be answered.


• The numbers were wrong (1), so the earnings were wrong. Much of the growth was an illusion (2) so the rating was wrong (3).

• If the rating was wrong, then the share price was wrong (4). That being the case, transactions took place at the wrong level.

• That being the case, there were winners (5) and losers.

• Moving on we know that trust and reputations take longer to build than they do to destroy (6).


• The company has conceded in its RNSs that its numbers, earnings and growth profile were misleading or wrong.

• Implication. The units are less good than earlier believed, there will be a tail. If the numbers are wrong, many of the units will shift to the left on the graph and there will be a tail. Some units could be inprofitable. If EBITDA is to fall for the group by 60% or so, it is hard to see how this will not be the case.

• Implication. Trading tail-end units in this environment is tough. Exiting them is even tougher. Get in line behind the CVAs out there etc. Operators tell us dealing with a small number of tail-end units is more time consuming than trading an equal number of better sites.


• We would suggest that the 23x earnings relied heavily on the growth profile which, as the company says, is not what it appeared.


• The rating multiplied by the earnings gives the shareprice. There is no rocket science here. If the inputs were wrong, then the share price was wrong.

• Observation. We are stating the obvious here, the 50p placing price tells us as much.


• Implication. But hold on, until Tuesday evening, buyers were paying 420p plus is good faith.

• Example. There was plenty of trade earlier last week and, as the company itself RNSd on 28 September, some member (or members) of staff bought 10,000 option shares at 316p each. The shares may or may not have been sold as there is no requirement for this to be announced for non-directors.

• There have also been sellers of the shares.

• Example. On 20 July, now-suspended CFO Chris Marsh made a profit on the sale of 666,666 options of £680,000. On 7 Feb, Mr Marsh made a profit of £1.27m on the sale of a further 666,666 options. Mr Marsh has retained 465,000 shares in the company.

• Example. On 20 July, CEO Paul May made a profit of £1.02m on the sale of 1m options. On 7 Feb, Mr May made a profit of £1.6m on the sale of a further 1m of options. Mr May retained 4.55m shares in the company.

• Example. On 29 June, non-executive director Lee Ginsberg sold two thirds of his shares (40,000 of 58,823) at 469p per share. On 2 Feb 2018, non-executive director James Horler sold around a third of his 147,116 shareholding at 367p per share.

• Example. In May 2014 the group IPOd at 170p per share. Existing shareholders sold to incoming shareholders. Whilst the shares performed well thereafter, some of that outperformance may not have been warranted and the 170p represents a multiple of 3.4x the 50p price paid to rescue the company last week.

• Caveat. Directors need to be able to sell shares and there are limited windows during which they can do so. There is no implication of wrong-doing here. Indeed, the directors have indicated that they did not appreciate that their company’s results, rating and share price were all wrong. What is certain, however, is that the directors got the better end of the above deals. No directors have bought shares in the last three years, though this is not unusual in growth companies.


• CAKE’s shares should relist Thursday and, whilst the company’s future may have been secured (in the absence of more shocks), it will be a long road back.

• We don’t yet know what happened & building trust will take time. There may be more announcements, resignations etc.

• Regarding the suitability of certain players going forward, no doubt conversations have been had and more will take place over the coming weeks.

• Directors may need to stick around to help clear up the mess (the stab-vest principle, Theresa May is familiar with it) but, after that, it could be better to clean the slate.


• We (kind of) get that EBITDA to Sept 2019 will be lower than previously forecast but why will sales be c10% lower. Were they misstated or are closures being factored in?

• If the former, then how did that happen & if the latter then what will be the exceptional costs and (reality check here) will you really be able to get rid?

• The forecast £120m of revenue for Sept 2019 is only £11,500 per week. The Philpotts and bottom-end Pat Val units could be doing less than £10k. It’s hard work at this level of turnover.

• EBITDA of £12m is forecast. Add in, say £5m for HO costs give unit EBITDA of £17m. This is less than £85k per unit. Again, not large numbers.