by John Hughes
Pan American Silver Corp highlights one of the more commonly-cited differences between IFRS and Canadian GAAP in the area of income taxes:
(On this occasion I’ll leave you in suspense about what that second item might be!)
CICA 3465.34 explains the old Canadian GAAP approach like this:
- FOREIGN CURRENCY TRANSLATION, Section 1651, requires the use of historical exchange rates to measure the cost of non-monetary assets of an integrated foreign operation such as inventory, property, plant and equipment. Assuming that the Canadian dollar is the currency of measurement, when exchange rates change, the amount of foreign currency revenues needed to recover the Canadian dollar cost of those assets also changes but the foreign currency tax basis of those assets does not change. After a change in exchange rates, there will be a difference between: (a) the amount of foreign currency needed to recover the Canadian dollar cost of those assets; and (b) the foreign currency tax basis of those assets.
- Although that difference technically meets the definition of a temporary difference, the substance of accounting for it as such would be to recognize future income taxes on exchange gains and losses that are not recognized in accordance with Section 1651. In order to resolve that conflict and to reduce complexity by eliminating cross-currency (Canadian dollar cost versus foreign tax basis) computations of future income taxes, recognition of future income tax assets and future income tax liabilities for those differences is prohibited.
The treatment is consistent with US GAAP, but the explanation is perhaps a little odd in a couple of ways. The reference to complexity seems almost tossed in as an aside, and it’s hard to see how the calculations required here would often be any more complex than those arising in income tax accounting generally. The real reason for the exemption, and the only one cited in the illustrative example attached to this paragraph in the Handbook, is no doubt the perceived inconsistency of recognizing (say) a liability relating to a foreign exchange gain that’s not itself recognized. But this rationale seems a bit circular: it’s exactly because of not recognizing the foreign exchange gain that the temporary difference arises in the first place. It makes perfect sense of course to continue measuring “integrated” non-monetary assets based on the original exchange rate, but it’s also true that changes in exchange rates have a real impact on the tax benefits attached to those assets from a reporting currency perspective.
IFRS simply doesn’t contain this exemption. In the 2009 Exposure Draft of proposed changes to IAS 12, which has subsequently stalled, the IASB reconsidered the issue and summed up its views just about as pithily as can be imagined:
- BC50 Paragraph 9(f) of SFAS 109 prohibits recognition of a deferred tax asset or liability for differences related to assets and liabilities that are remeasured from the local currency into the functional currency using historical exchange rates and result from (a) changes in exchange rates or (b) indexing for tax purposes. In contrast, IAS 12 requires recognition of a deferred tax liability or asset for such temporary differences.
- Consistently with the objective of eliminating exceptions to the temporary difference approach, the Board proposes no exception for such differences.
And so there you are. It seems to me difficult to make a conceptual argument for anything other than this, for the reasons I mentioned. Now of course, one could ask: what’s the basic point of it all? Between the length of time that might expire before those temporary differences ever reverse, and the inherent volatility of exchange rates, the deferred tax amount arising – which could of course be enormous – may seem like an abstract kind of construct at best. True enough, but I think it’s a case of in for a penny, in for a pound. As I wrote previously, the point of deferred tax accounting doesn’t seem to me to be so much in the specific information content of the resulting numbers, but rather to avoid the tax-rate fluctuations, and utter challenge to predictive value, that would exist if it wasn’t applied.
It’s hard, I think, to argue rationally against the basic principle driving IAS 12, that an asset or liability’s tax costs or benefits are an inherent component of what’s acquired or assumed initially, and that by their nature they’re used up or tapped over the item’s economic life, even if the related cash flow impact may vary due to the quirks of the local tax code. No doubt there are cases where following the logic of that principle opens up costs and complexities that can never be justified by the resulting benefits – especially I suppose where the deferred tax assets and liabilities will only become current so far into the future that they’re barely relevant to any investing or other decision anyone might make today. But given the incomprehensible cookbook of rules that results from too many exemptions along those lines, I think the IASB was right to avoid allowing one here.
The opinions expressed are solely those of the author.