A recent Globe and Mail article carried the tantalizing headline – “What management isn’t saying: A potential blind spot for Canadian stock investors”

Here’s an extract from the article, written by David Milstead:

• Late last year, Artis Real Estate Investment Trust told its investors it had made an accounting mistake and announced it would correct its financials – a “restatement” – saying it had understated its profits by a factor of three.
• Earlier in 2016, Northview Apartment Real Estate Investment Trust said it would restate its financials for the third time in less than five years. Two of the errors related to the accounting for its financing costs.
• However, when it came time to tell investors whether there were any weaknesses in their “internal controls” – the systems that are supposed to help produce robust, accurate financial statements and prevent such accounting errors – the companies said no, there were no weaknesses to report.
• These two companies are more the norm than the exception, according to a new study by Robert Pozen, a senior lecturer at Massachusetts Institute of Technology’s Sloan School of Management, and Olga Usvyatsky, vice-president of research at Audit Analytics. They found 78 restatements between 2009 and 2016 among Canadian-listed companies with a market capitalization above $75-million. Of those 78 companies, only 18 disclosed a material weakness in its internal controls…

The issue here relates to National Instrument 52-109 Certification of Disclosure in Issuers’ Annual and Interim Filings. Among much else, this requires that if a non-venture issuer identifies a material weakness in its internal control over financial reporting (ICFR), it must describe that weakness and its impact in its annual or interim filings, as well as plans or actions already taken to remediate the weakness. A material weakness is an ICFR deficiency that causes a reasonable possibility that a material misstatement in financial statements “will not be prevented or detected on a timely basis.” In a situation where a material misstatement has already been identified, it might seem all but obvious that such a weakness must have existed. But the article sets out reasons why it might not work that way:

• Pozen says a restatement, “next to fraud, is the biggest indicator of a material weakness in internal controls.” Yet it is not definitive proof. Indeed, Dushyant Vyas, an assistant professor in accounting at the University of Toronto Mississauga, says that a company can respond to changes in the interpretations of an accounting rule, or simply disagree with its outside auditor. “While restatements in general and material/multiple restatements in particular raise a red flag that an internal controls weakness may exist, they do not necessarily imply [the existence of] internal controls weakness.”
• In a number of cases in the study by Dr. Pozen and Audit Analytics, the companies may not be disclosing an internal controls weakness because they believe that the restatements or errors themselves, even though seemingly large, are not material.
• After discovering an error in translating foreign currency, Artis reduced a loss of $25.97-million in the second quarter of 2016 to $4.01-million, a difference of nearly $22-million, or 85 per cent of the previous net loss. But Artis chief financial officer Jim Green says “we … discussed [this] with our external auditors and they agreed with our assessment that it would not represent a control weakness and would not be material to a reader of the financial statements.”
• Green says the restatement had no impact on the balance sheet, cash flow statement or on any of the non-GAAP measures commonly looked at for REITs such as funds from operations or various debt metrics. He also claims the absolute dollar amount of the change was immaterial when compared with the company’s revenue, which was just under $135-million in the quarter.

All this really tells us, of course, is that the determination of a material weakness in ICFR is subject to a large amount of somewhat abstract and tedious judgment. But in a way, for companies that have already reported restatements, it doesn’t really matter – investors can see for themselves what happened, and can draw their own conclusions on whether the event is likely to be (say) an isolated technical matter, or whether it might have broader ongoing implications. The real risk-sensitive information content perhaps lies in material weaknesses reported by companies before they have to restate anything. However, on the rare occasions when these occur, they’re usually generic and unspecific, for example:

• The material weakness relates to a limited number of personnel assigned to positions that involve processing financial information, resulting in a lack of segregation of duties so that all journal entries and account reconciliations are reviewed by someone other than the preparer, heightening the risk of error or fraud. If we are unable to remediate the material weakness, or other control deficiencies are identified, we may not be able to report our financial results accurately, prevent fraud or file our periodic reports as a public company in a timely manner. Due to our small size and early stage of our business, segregation of duties may not always be possible and may not be economically feasible…

Such disclosure is better than nothing, but doesn’t go far in allowing an understanding of the real extent of the reporting risk attaching to this particular company, or of exactly what management does about it.

Anyway, although the Globe and Mail article provides ample reason to think this aspect of NI52-109 doesn’t work particularly well, it doesn’t really convey why that’s a major issue, beyond the following:

• “There are serious questions about whether there are enough disclosures for investors, and in the end, that’s the critical issue: Investors should be concerned if there’s an internal controls problem that hasn’t been fixed, or if they say it’s been fixed, and you get a second or third restatement,” Dr. Pozen says.

But the absence of a reported material weakness surely makes it more rather than less likely that the company and its advisors don’t expect the issue to reoccur, and there are few such instances of companies going through multiple rounds of restatements. The article pushes the idea that management’s assertions on the effectiveness of ICFR should be audited (as required in the US; Canadian regulators consciously stopped short of taking this step), but it doesn’t come close to making the case that the issues identified would justify imposing such a significant ongoing expense (and as the comment by Artis indicates, auditor opinions also rely on judgments…)

Those involved in ICFR sometimes comment that the process of assessing its effectiveness tends to focus too much on mundane, repetitive, low-risk procedures, rather than on the key areas of estimation and judgment and high-level debate (which by their nature, may not be as thoroughly structured by systemic controls and procedures). Despite this, I tend to believe the regime has been a positive contributor to the stability and credibility of financial reporting, but I think that’s probably more true if it’s regarded as a broad preventative measure, rather than as a communicator of risk on any particular entity. When a company’s compliance with IFRS has tripped up in the past, it’s certainly wise to be aware that it might do so again. But one should never accept the financial statements of any company, however reliable its controls may seemingly have been in the past, without being aware of the risk inherent in so many underlying processes and judgments and estimates….

The opinions expressed are solely those of the author