The hardest question in standard-setting?

When accounting standard-setters have to make decisions about recognition and measurement, they often face two competing claims:

  • some people argue that investors will under-react if companies are forced to recognise something; but
  • other people argue that investors will over-react if companies are forced to recognise that thing.

Standard-setters would love to get evidence that would help them decide which of those claims is more valid. Unfortunately, such evidence is rarely or never available, as I discuss below.

When do these questions arise?

These questions about how investors will react to information often arise if a standard-setter is deciding whether it should start requiring companies to do one or more of the following:

  • recognise an asset (or liability);
  • measure a recognised asset (or liability) at a current value, rather than on a cost basis;
  • recognise changes in the carrying amount of a recognised asset (or liability) in profit or loss, rather than in other comprehensive income.

I will discuss initially questions about whether companies should recognise a thing in the primary financial statements, rather than just disclose that thing in the notes. By ‘thing’, I mean an asset or liability, or a change in the carrying amount of a recognised asset or liability. Towards the end of this post, I will then discuss briefly questions about whether companies should disclose things at all.

Is evidence ever available?

If companies recognise a thing, rather than just disclose it, do investors react more? It is sometimes possible to find evidence that will help answer that question. But usually—indeed, I suspect, always—such evidence cannot show how much investors should react so that they can make optimal investment decisions. Consequently, such evidence cannot show whether investors:

  • under-react if companies disclose that thing but don’t recognise it; or
  • over-react if companies do recognise it.

Recognition versus disclosure

These questions arise if some people claim that investors will start reacting more when companies start recognising a thing than they did when companies were just disclosing that thing. If these claims are correct, at first sight it appears that there are only 2 possibilities:

  • investors were under-reacting when companies were just disclosing that thing—perhaps because the information about that thing wasn’t prominent enough.
  • investors will start over-reacting when companies start recognising that thing—perhaps because the information about that thing will be too prominent and investors will not see it in full context.

There is in fact a 3rd possibility. Investors may conclude that information about the thing is subject to less measurement uncertainty if it is recognised than if it is just disclosed. If that conclusion is (at least sometimes) well founded, then it may be rational for investors to place more weight (at least sometimes) on information about the thing when it is recognised than they do it when the information is merely disclosed without recognition.

I do not try to assess in this post whether investors do draw such conclusions. Nor do I assess whether such conclusions would be well founded.


Sometimes, questions like those discussed above also arise in discussions about whether a standard-setter should require companies to disclose information. One such question is whether companies should be required to disclose a current value of an asset (or liability) measured in the primary financial statements on a cost basis.

Sometimes, some people oppose requirements to disclose some such information, on the grounds that investors might mis-interpret the information and over-react it. A common version of this argument is that disclosing a 2nd measurement of an asset or liability might confuse those investors about the meaning of the measurements reported in the primary financial statements.    

For the same reasons as those given above, there is unlikely to be evidence indicating whether investors would over-react to information disclosed in the notes.  


Accounting standard-setters often have to decide whether assets and liabilities should be recognised, how they should be measured and where changes in their carrying amounts should be recognised. In making those decisions, they often receive arguments that investors might over-react or under-react to information produced by applying the outcomes of those decisions.

Unfortunately, evidence to help standard-setters assess those claims does not exist, and probably cannot exist. That is because evidence is not available to indicate how much investors should be reacting if they are to make optimal investment decisions. So, standard-setters need to ask themselves whether those claims are valid. Without evidence, that question is perhaps the most difficult question in setting accounting standards.

Leave a comment

Your email address will not be published. Required fields are marked *