Back in October 2018 Patisserie Valerie (PV) shocked the markets by announcing they had been “notified of potentially fraudulent, accounting irregularities” in their accounts. From nowhere, PV had a huge, almost £40m, cash black hole. What followed was a lot of head-scratching. How could this have happened? I did my best to try to make sense of the tumult of events (you can have a read here).
On January 22nd 2019 Patisserie Valerie fell into administration (much like how I fall out of the bar after having far too much to drink – it was never going to end well). The administrators, KPMG, announced 71 store closures with the loss of over 900 jobs. This had followed an earlier frantic, and unsuccessful, scramble by the company to delay repayments to its banks.
All in all, it’s very sad for the thousands of staff who look likely to lose their jobs through no fault of their own.
So what happens now? To answer that question, let’s look at what we’ve found out since the news broke in October 2018.
What happened? – an update
In my October blog post, I reached the conclusion that there was likely to be wholesale long-term manipulation of Patisserie Valerie’s general ledger (the accounting records). That appears to have been the case. PV stating:
The work carried out by the Company’s forensic accountants since then has revealed that the misstatement of its accounts was extensive, involving very significant manipulation of the balance sheet and profit and loss accounts. Among other manipulations, this involved thousands of false entries into the Company’s ledgers.
That said, the scale of the manipulations is, even to me, surprisingly large.
It’s still not clear how the general ledger was manipulated. For now, I’d still place my money on inappropriate recognition of voucher/coupon revenues and inappropriate capitalising of expenses (more detail in my first post). It’s possible that the ‘secret’ bank accounts were also being used to pay suppliers and those costs not being recognised. Being frank, it’s such a blatant wheeze that it didn’t cross my mind back in October.
It seems safe to conclude that the accounts weren’t worth the paper they’re written on.
That leaves us with the age-old question…
What were the auditors doing?
Much like the gawping of passers-by when there’s a road accident, the usual: “auditors are scam artists” line has been trotted out. I’m no shill for the audit industry (as should be clear from my earlier Patisserie Valerie post, and my Carillion Series) but we need to get one thing straight. It’s not the auditors’ job to stop fraud. That is down to management. The directors have the primary responsibility for the prevention and detection of fraud. The best way to stop fraud is to not hire bad dudes (something that Enron failed at a little bit).
It’s also not the job of the auditors to decide if fraud has occurred. That’s because fraud is a legal concept that the courts alone can judge. That said, it is the job of the auditors to gain assurance that the accounts are free of material misstatement – be that from error or fraud. To do so, the auditors must follow a set of guidelines. These set out the processes and principles for performing a good audit.
Chief among those is Professional Scepticism. In English, this means the auditors should always be alert to potential fraud and consider whether fraud is a possibility regardless of their views on the integrity and honesty of management.
Many people argue that Scepticism is a lost art in the auditing world. I’d say it ranges on a spectrum from obliviousness (fresh lambs in the audit profession) through to paranoia (forensic accountants – you do come to second guess everything as a forensic accountant). I think the public perceives that auditors should be (or are) more towards the forensic accounting end than they actually are. Whether that is right or wrong is a big open question.
What should the auditors have been doing?
So how do we apply Professional Scepticism? Generally speaking, there are two tools: tests of control and substantive testing.
Tests of control
Tests of control are where the auditor tests to see if the company’s processes are effective and appropriate. For example, for PV, this would involve checking to see if two (or more) people are required to bank store takings at the close of business or whether two or more people are required to approve big expenditures.
These are the less time-consuming and ‘bread and butter’ of the audit. In theory, if the right controls are in place and are working effectively then it less likely for fraud or errors to occur.
Substantive testing is where the auditor checks the numbers in the accounts. It takes more time and is more rigorous.
There are two types of substantive tests: analytical procedures and tests of detail.
Analytical procedures are where the auditors do some number crunching and create an expectation of what the numbers should look like. They then check to see if they are close and if they aren’t, investigate why. For example, the auditor will look at the rental expense for a random month then extrapolate that to a yearly expected figure. They’ll then check to see if the actual figure marries up.
Tests of detail are given away in the name. The auditor goes straight into the transactions to test the figures are right. The classic example is the bank reconciliation. Does the company have the cash it says in the bank. In PV’s case – oops.
Over time there has been a general tendency to move from substantive testing to tests of control and from tests of detail to analytical procedures. Some argue, and I’m sympathetic to that argument, that it leads to less rigorous auditing.
Cash in the attic?
Let’s take the cash reconciliation as an example. One that appears to have failed at PV. The auditor asks management for the bank account details. It then goes to the bank and asks for the account transactions. The auditors should then go back to the general ledger to check they match up. With PV there appears to have been two big failings.
First, there were additional bank accounts that weren’t noticed and were overdrawn. It is possible management didn’t provide those account numbers for the auditors to check. That said, as a forensic accountant (not an auditor) one of the first things I would do is to check the general ledger to find all the cash accounts and make sure they are all captured (called completeness in accounting jargon). I suspect that the auditors may not have done this.
Second, there should be a thorough check of transactions and matching to the general ledger. Especially around year-end where there is usually a flurry of activity. According to The Times, PV was cashing millions of pounds in cheques at year-end, only for them to bounce a few days later. This is called Cheque Kiting. It’s one of the oldest tricks in the book. It is very surprising that this happened without the auditors knowing if they were performing the bank reconciliation properly. Together, with the secret accounts, it would seem that there was Cross-Firing going on. Another classic scam where webs of transactions are used to give the appearance of solvency (‘cash’ is deposited in the known accounts and the overdrafts are hidden away).
In my first post, I talked a bit about tests of control. And above I mentioned a few examples. Principally where two or more people are required to ‘do stuff’. This is called Segregation of Duties. One of the classic controls in a business. With good controls, one person can’t just do whatever they want. A second person needs to approve transactions. This can usually be tested easily by checking purchase orders have two signatories or checking that large general ledger transactions have two people posting them (one being the finance director or senior management).
A problematic proposition
This leaves us with a problematic proposition. If the auditors did their tests of control adequately, the “very significant manipulation” of the general ledger could not have manipulated by one person alone – i.e. there was collusion among several people – a serious issue. On the other hand, if we accept that this was one person acting rogue, then the controls were inadequate and should have been picked up by both the external auditors and the internal audit committee.
So what next for the auditors?
Well the auditors, Grant Thornton (GT), are being investigated by the Financial Reporting Council (FRC). It’ll take several months to get to the bottom of what’s gone on and whether GT erred in their audit.
Unfortunately, the FRC is stymied by a lack of funding, resourcing and powers. I’m very unsatisfied by the politicians that condemn the FRC (and other regulators) for being a soft touch but repeatedly refuse to fund or endow them with the powers they need to bring bad practices to task.
The cynic in me feels that we are likely to have a publicly unpopular result to the investigation and the many talking heads taking the opportunity to push out the standard tropes.
What about the banks?
We talked about Cheque Kiting above. The Times report on bouncing cheques is troubling. Therefore, it should be no surprise that the Financial Conduct Authority (FCA) is investigating. Put simply, millions of pounds of bouncing cheques should be raising alarm bells. This was also in contrast to publicly available figures that showed big piles of cash and no overdraft.
Serious questions about the procedures and controls of the two banks in question (HSBC and Barclays) will need to be asked. Given the possible accounting fraud that has taken place at PV, the banks may have fallen foul of Anti Money Laundering regulations, which is deeply concerning.
As with the FRC investigation, I wouldn’t hold my breath to hear back from the FCA. Any investigation is likely to take a long time.
The Administration process
As mentioned up top, Patisserie Valerie is now in administration. The administrators, KPMG, will be looking to restructure the business and potentially save it, through cutting costs (closing stores and reaching agreements with lenders) and/or finding a buyer for the remaining business.
Positive sounds have been mooted on finding a potential buyer. I’ll confess I’m a little more pessimistic. Looking at the accounts (I know…) PV had very little in the way of realisable assets. It had less than £2m in land in buildings. It had a book value of almost £9m on leasehold improvements, though it would seem difficult for the company to now realise this value through its leasing agreements.
The principal assets are plant, equipment, fixtures and fittings (£30m) and goodwill (£17m). It’s difficult to see the company able to make anywhere near those values for those assets in light of the collapse of the business. Especially given the huge question marks on the provenance of the financial figures.
Having spoken to some retail experts, there is a market for good food retail businesses. Even in these challenging times for the high street. That very much reflects my experience in valuing several food retailers over the years. Therefore the big question is: is there a viable business in the ruins? Unfortunately, it is very difficult to say at this moment.
If no buyer is found, and the company winds-up insolvent, then a court-appointed Official Receiver (OR) from the Insolvency Service (IS) will start the process of winding the business up. That process can take a long time. The IS will investigate to work out when the company was insolvent and whether any civil or criminal actions should be brought (you can read more about this process in my Carillion post). The OR and IS will rely heavily on external managers (read accounting firms) to do a lot of the work.
Will anybody go to jail?
Proving fraud is very hard. The Serious Fraud Office (SFO) have announced they are investigating. Though you shouldn’t get your hopes up. Yesterday (23/01/2019) the SFO gave up on its last court trial in the Tesco fraud case (having had the court dismiss the earlier trials). The jury is very much out on whether the SFO can successfully bring prosecutions for complex accounting fraud (pun intended).
Having worked big forensic accounting cases (typically commercial, not criminal), the expertise, costs and time involved are exceptional.
By way of a personal example, it could sometimes take me a whole day to work through a single general ledger transaction. Why did it exist? Is the transaction appropriate? Is it correctly processed? How does it affect the financial statements? Where there are 1,000s of potentially manipulated transactions it could take a long time to build a picture of what went down and what should have been recorded. And despite my wife calling me a cheap b*stard, my charge out rates were quite expensive!
Another difficulty is making the step from recording incorrect/fraudulent transactions to the intentions behind those postings. It is very difficult to prove that somebody both knew and intended to commit the fraud through fraudulent postings. In effect, you need to overlay human thoughts and emotions to a series of numbers. This is where most fraud cases, including the SFO Tesco case, fall down.
Added (25/01/2019): Do Auditors ever go to jail?
Auditors can go to jail if they can be proven to have committed fraud. That is a high hurdle. Deliberately so, I’m not sure it’s appropriate to send people to jail just because they were incompetent. The Companies Act 2006 also provides for a criminal offence in relation to inaccurate auditors’ reports. The offence consists of knowingly or recklessly causing a report to include anything that is misleading, false or deceptive, or omitting a required statement of a problem with the accounts or audit (section 507). This is punishable with an unlimited fine.
A word on investigations
These investigations cost a lot of money – millions, or perhaps tens of millions, of pounds. Given the size of PV (‘market cap’ of £500m), there may not be the money or financial interest in a deep investigation. That would be an immense shame – but I think something people should be aware of.
It is entirely possible that we never get closure on the curious case of Patisserie Valerie.
On that rather depressing note, I shall leave you be. Thank you for reading.
All the best,
Young FI Guy
Two news articles this morning reported some interesting developments. First, the FT reported that PV is looking to get a rebate for overpaid corporation tax on previous years’ “profits”. Apparently, PV has paid £16m in corporation tax over the past five years. This could be quite a hefty asset for the company. Second, the Times reported that HMRC had previously raised concerns with PV that some of its invoices and cheques had been forged back in 2016. It’s unclear what, if any action, the tax authorities or the company took.
Administrators KPMG have announced the sale of the remaining parts of the business. The rump of PV, 96 stores, has been bought by an Irish-based private equity house, Causeway Capital. Raising £10m. New management, including the new CEO, Steve Francis have taken a significant non-controlling stake (between 10 and 50%). A separate deal was reached for the Philpotts bakery chain with AF Blakemore, raising another £3m. Collectively raising £13m. Baker & Spice has been sold for £2.5m to The Department of Coffee and Social Affairs, a pretty good coffee shop that started off on Leather Lane (and one of my old haunts). Altogether that’s a little over £15m, for a business ‘worth’ around £450m (market cap) before the shenanigans. IAfter paying the banks and Luke Johnson’s loans, I doubt there’ll be much if anything left. Hopefully, this means lots of jobs saved. Now a long wait to find out what the investigations turn up.
Accountants’ corner appendix
I’ve seen a few ‘commentators’ suggest that the PV ‘fraud‘ could have been spotted in advance. I think that comes with a huge dollop of hindsight bias. My view is that it was not possible to spot in advance (as articulated in my first post). As far as I’m aware there wasn’t a big short market in the stock before the implosion. The two common arguments put forward that this could have been spotted are:
- The business was too profitable – PV had big-profit margins (in the high teens) according to its accounts. Higher than competitors such as Costa (mid-teens) and Greggs (high single figures). From my experience valuing several food retailers, some well-positioned players have been able to make very good profits. Some achieving similar (and higher) figures than PV. Likewise, one particular company was making nearly £1m income per site per year (estimates for PV are around £500k). Whilst its valuation multiples were also very high (EV/EBITDA in the low-teens) again this was towards the high-end of what I’ve seen, but not completely unreasonable. The profitability of the company, which we now know to be a charade, wasn’t especially unreasonable for a company in this industry.
- The business was working capital positive – this is a bit more difficult. Working Capital is current assets less current liabilities. Most retail companies are working capital negative – that is, they rack up liabilities faster than assets. This is because retailers get cash in hand, whilst paying suppliers later down the line. So you’d expect an expanding retail business to generate negative working capital changes year-on-year. PV had the opposite. As with all accounting you need a narrative to work with the figures. Especially so for working capital which can be a brain-melter. Some things we know about PV: it was expanding rapidly (implying -ive WC), but making lots of rental prepayments upfront (+ive WC) it was moving into new areas such as vouchers/Sainsburys/Barclays coffee deal (providing possible delay in payments, increase in inventory +ive WC), it was expanding its pre-order cake business (-ive WC payment upfront, but holding more inventory +ive WC). So it’s much less clear that PV should have been generating negative changes in working capital. When I’ve gone a step further and stripped out prepayments (for upfront rent) working capital is still positive but the change is negative for 2015 and 2016 (though positive for 2017). This is much less alarming. In short, I don’t see the Working Capital balances alone as a give away of a fraud.