Here’s the background to an issue recently considered by Canada’s IFRS Discussion Group:

• On January 1, 20X5, an employee receives a share-based payment award that vests if Entity A’s share price hits a market performance condition (“target”) either by December 31, 20X5 (“Date 1”), or by December 31, 20X6 (“Date 2”).
• The employee could receive from zero to 100 per cent of the award depending on how much of the target is met. This means that:
  • the award has the potential to be 100 per cent vested on Date 1 if the full target is met on that date;
  • if the award does not fully vest on Date 1, the employee can earn any remaining amount on Date 2; and
  • the employee forfeits any amount not vested by Date 2.
• Assume that Entity A initially estimated that zero per cent of the award would be earned on Date 1 and 100 per cent of the award would be earned on Date 2. However, on Date 1, Entity A meets the full target. Hence, 100 per cent of the award is vested on Date 1.

This provides a nice opportunity to throw a few stones at IFRS 2, which I haven’t done for a few years. You’ll recall that the standard is based on the assumption that since no one gets something for nothing, an entity issuing share-based payments must have received some form of economic resources in return for them. That’s why IFRS emphasizes measuring the share-based instruments at the fair value of the goods or services received, unless that fair value can’t be measured reliably. It follows, in broad terms, that certain kinds of conditions (so-called vesting conditions) will (at least in theory) affect the economic resources to be received, for example by incentivizing the recipients to work harder or better. Other kinds of conditions though (non-vesting conditions) won’t affect the resources received: the conditions are beyond the control of the recipients, so they’ll put in the same amount and quality of work either way. In the first case, then, the outcome of the condition may result in revising the expense in the light of experience (by applying forfeiture-style accounting), but in the second case it doesn’t. A broad-based market condition, such as a target return on a share index, falls into the second category, because it reflects the performance not only of an entity itself but that of other entities outside the group as well, and so is inherently beyond the control of any employee. However, such market conditions are taken into account in estimating the fair value of the equity instruments granted. Where the length of the vesting period could vary, depending on when a performance condition is satisfied, IFRS 2 requires estimating the length of the expected vesting period at grant date, based on the most likely outcome of the performance condition. It specifies: “If the performance condition is a market condition, the estimate of the length of the expected vesting period shall be consistent with the assumptions used in estimating the fair value of the options granted; and it should not be subsequently revised.”

Against that backdrop, the group discussed whether entity A should accelerate the expense recognition when the target is met on Date 1, or should rather continue to recognize the expense over the estimated two-year vesting period. You might think acceleration makes sense because that’s consistent with the requirement to recognize the expense over an option’s vesting period, reflecting the fact that Entity A has received all the services to which it is entitled in exchange for the awards at the point that the market condition is met. Alternatively, based on the passage above, you might take the view that such acceleration is specifically prohibited, whether or not it might seem to better reflect the economics of the arrangement.

Most (but not all) group members agreed that acceleration is appropriate, while noting that diversity exists in practice. Under this view, accelerating the expense recognition needn’t be seen as a revision of the expected vesting period due to a change in estimate, but rather as a reflection of a vesting condition being met. It seems to me that that’s the better answer within the model summarized above. But still, the model assumes an extremely programmatic view of human behaviour, more so than I think non-specialist users can readily grasp and accept. The difficulty is compounded as the value of the share-based payment expense may not bear any relationship to what the individuals ultimately ended up realizing on exercise. Regulatory regimes (such as the Canadian Form F1-102F6) may compensate to some extent by requiring at least some disclosure of the payout value of vested awards and suchlike, but such a perspective is absent from the financial statements. Certainly it’s a justifiable absence in the context of the objective and scope of IFRS 2, but you can’t blame non-specialist users if they may think it’s an odd omission, even to the extent of rendering the whole accounting mechanism basically pointless.

Anyway, the Group recommended the Accounting Standards Board discuss the issue to determine whether any further action is required. Thereafter it seems the discussion understandably petered out…

The opinions expressed are solely those of the author