Blaik Wilson was surprised when he read the review draft of IFRS 9: Financial Instruments in September. A paragraph never seen before had been inserted in the application guidance.
“It has caused a quite a lot of consternation and frustration among those who are looking to interpret and deploy this standard’s treatment of hypothetical derivatives,” says Wilson, who is Solutions Consulting Manager for Asia Pacific at Reval, a software company that provides Software-as-a-Service treasury management solutions, and a Chartered Accountant specializing in financial instruments, hedge accounting and treasury policy.
He spoke to CFO Innovation’s Cesar Bacani about what he calls a “fatal flaw” in the proposed IFRS 9, which is expected to be issued as a final standard by the International Accounting Standards Board (IASB) any day now. Excerpts:
Talk to us briefly about IFRS 9.
It’s going to replace the current standard, IAS 39, and they’re doing it in three phases. One is on classification and measurement, which has already been released and is a standard. The second phase is impairment, which is actually not going to be released probably until late next year.
The third phase is hedge accounting, which after a long time came out as what they call a review draft in September this year. From there, you’ve got 90 days [to finalise and issue it as part of the standard].
In compiling IFRS 9, there’s been a lot of feedback and a lot of things have been taken on. I think most people agree there’s in general a significant improvement on the current rules. However, the IASB added a paragraph deep in the application guidance that has never been seen prior to the review draft in September.
And that is Paragraph B6.5.5. [Click here to read this section.]
It has caused quite a lot of consternation and frustration among those who are looking to interpret and deploy this standard’s treatment of hypothetical derivatives.
Has the IASB explained why they inserted Paragraph B6.5.5 in the September review draft?
I don’t know why it wasn’t exposed earlier for comment. The application guidance has a lot more detail than what they earlier put out and they felt that probably they needed to define what a hypothetical derivative was because it was never really properly defined in the current standard [IAS 39]. They probably did not fully appreciate the application of what they decided and how it would work.
In your view, this is a ‘fatal flaw.’ Why?
It is a fatal flaw on a number of fronts. The primary one is the fact that we’re now going to have a situation where the accounting outcomes won’t reflect the risk management outcomes. One of the objectives set up in IFRS 9 is that the objective of hedge accounting be oriented to represent the entity’s risk management activity, which is a very sensible objective to my mind and what people quite like.
But now we’re going to have a situation where the typical FX hedging strategy – where all the foreign cash flows are completely offset economically by being converted to the local currency – [will lead to] p&l volatility through movement in items such as currency basis. Economically you are hedged, all exposures are offset, but you’re still going to experience p&l volatility, which I think is not in line with accounting reflecting economic risk management.
At worst, you can have a situation where once again the accounting will start driving the economic hedging decisions, rather than the other way around. People maybe will not use certain hedging instruments and they might not hedge as far forward as they used to because of this accounting impact, even though economically that’s what they would prefer to do.
Paragraph B6.5.5 says that “this is one possible way of calculating the change in the value of the hedged item.” Presumably there are other ways to calculate the change in value other than using hypothetical derivatives?
You don’t have to use the hypothetical derivative method, but it is by far the most popular because it is easier in terms of calculating fair value . . . . And they don’t really talk about any other method. But either way, the implication is the same. The problem [of the accounting not reflecting the economic risk management] is there regardless.
You should get the same accounting outcomes with the use of hypothetical derivatives or some other method. They just must fairly represent the exposure. You’re going to get basis risk in the hypothetical and the hedge, and where there is basis risk there is p&l volatility. And it’s unpredictable volatility.
For instance, if I had US dollar debt, and I was converting it to Hong Kong dollar debt using a cross-currency swap, if I wanted to measure the fair value of that US dollar debt, the implication is I can’t use currency basis in the way I measure that fair value of the US dollar debt, whether I do it straight from the debt or use a cross-currency swap as a hypothetical derivative.
How does the proposed change differ from what is the current practice?
Most people nowadays interpret the hypothetical derivative as a perfect hedge of the exposure. They really say: How do I hedge that exposure in the market? They use observable market data, and so that would include elements like currency basis.
In your view, is the current method preferable to what IASB wants to do?
Yes. Although there’s a lack of guidance, what we recommend – we wrote a letter to the IASB on this fatal flaw – is that the IASB define a hypothetical derivative as a perfect hedge of the exposure, which is consistent with US GAAP.
What would be IASB’s objection to that particular approach?
I think they find it difficult to think of the exposure being measured with something that is not priced into the exposure. If you issue USD debt, you typically don’t price in currency basis when you issue that debt. They are uncomfortable with measuring the value of that debt using a hypothetical derivative that does include currency basis.
Has Reval looked at hedge instruments and valued them under the current method and the proposed changes? What have you found?
We’ve looked at USD-EUR, USD-HKD and other instruments. In the currency basis movement of the euro over the last four years, [the variance between the two methods] is as much as over 40 basis points. Relative to interest rates, that’s massive, when you think interest rates are so low.
Is it the case that under the current method the hedge comes out as “perfect” in the accounting?
Obviously, there are other factors that can cause ineffectiveness. But if you’ve got a perfectly matched up cash flow and typical FX cash flow hedge, yes. That’s typically zero ineffectiveness. There are a few provisos around that, but certainly this issue [of the accounting outcome not reflecting the economic outcome] does not arise.
So what did you recommend to the IASB?
The alternative is to consider the hypothetical as it must be a perfect hedge of the exposure, given the nature of the risk or hedging. This is what they do under US GAAP. It’s a concept that people can understand because they are in the hedging business, so they know what it takes to create a perfect hedge. It uses observable market data so you’re able to access that data and it’s verifiable.
One of the fatal flaws in the approach in [Paragraph B6.5.5] is that, to value the exposure as they describe it, you have to create some FX forward curve that does not exist in the market. So you’ll have to calculate and maintain your own synthetic forward curve for measuring an exposure. That adds a massive layer of complexity especially in Asia, where FX hedging is by far the most common strategy.
What is the experience in the US with regard to hedge accounting under US GAAP? Are companies happy with it?
I think they’re pretty happy with the way this part of the standard works for them, but they don’t hedge FX as much as people outside of the US do, so it’s a little bit hard to say.
If Paragraph B6.5.5 is enshrined in IFRS 9, what is the impact on Reval?
We will need to build the capability to create a synthetic forward curve for measuring the exposure. We have all the inputs, we have the interest rates, we have the spot rates, but it will be another layer of work to create a synthetic curve.
Reval will work with our Big Four partners to make sure we use an approach they expect. Our clients will probably rely a lot on us to do that. For companies that do not have Reval, they will have [to create] their own synthetic forward curve.
But if anyone can create their own synthetic forward curve, it will be different from one organisation to the next?
That’s a good point. The way you create your theoretical forward curve [that does not impute a charge for exchanging currencies] is you look at the spot rate but then you also bring in the interest rates in the two countries that you’re hedging and from those interest rates you derive your theoretical forward point.
That brings out the audit risk, the IT cost, it becomes more resource heavy. You’re going to get room for error.
But to be fair, if you do very short term hedging, it’s not really impacted. It’s when you hedge beyond one year forward, that’s when it becomes a significant number.