My point on the guidance was really about the mandatory ones. That's what I wanted to confirm, because a limitation-on-benefits rule is easier to apply objectively—you know when it applies—but it's much more complicated and, therefore, it's generally done on a bilateral basis. That's the approach the U.S. has taken, and that's in large part why the U.S. has not signed on to the multi-level instrument at all.
Canada and a number of other countries have gone with the much more subjective, ambiguous and difficult to understand, but much easier to draft, principal purpose test. That's why I'm saying we need more examples of when the principal purpose test would apply. I think the OECD's were limited, because they wanted consensus among members and all members to be able to read each example in whichever way they wanted, to determine whether they wanted it to apply or not.
In my view, Canada should use much more difficult, real-life examples, to be able to then ask if there is a certain amount of substance required for this rule not to apply. What can businesses look to in determining whether it will apply or not?
I think a quick example is with collective investors. If there were any sort of threshold, would it be enough if 99% of investors had identical treaty benefits? Is 50% enough? Is there anything below 100% that would be sufficient not to apply the rule? If it requires 100%, I think it would be helpful if Canada at least said that it's 100%. That way, taxpayers would know to invest in other countries and to avoid Canada. They'd at least get a sense of what their after-tax returns would be. After-tax returns are quite simple: they are just your revenues, costs and taxes. If you take one of those three factors, the tax part, and make it unclear what taxes are going to apply, either on distributions or on a sale, then investors won't know how to properly value that investment.