So we've arrived at thin cap and we've been avoiding this so far because we haven't had our interest line. And now we've got an interest line, we can start creating our thin capitalization working. In terms of context, thin capitalization is a set of guidelines and those guidelines are then adopted in varying ways by different countries, and the guidelines are designed to catch out highly indebted structures and make sure that they're not being manipulated for tax purposes. This has come about from some of the scandals of the 2000s where especially group accounting led to non arms length transactions with subsidiaries, which led to high indebtedness in high tax jurisdictions and kind of tax manipulation and funneling of profits. So to catch this out, what tends to happen is countries that adopt rules like this will have some sort of limit. And here we've simplified a bit and we've said 30% of EBITDA is your limit for allowable tax. You can see our EBITDA in year 52, 16. Now, 30% of that is definitely below that number. That means that some of this is going to be disallowed from an accounting point of view, that number will not change. But from a tax allowable point of view, that number is about to change, and that's what we're goning to do now. I think I warned you earlier in the NOLs section, and I'll say it again, if this were to be highly significant in your project, I would urge you to take tax advice because it differs from country to country and it differs from year to year. Now, if we start building, you can see we've got our interest as a proportion of EBITDA that immediately comes up with an error. So we can wrap that in an IF error.

If we pull that to the right, it's good for discussion.

You can see I need to flip the sign, which is one of my favorite errors to make myself.

Now, as we sort of felt out earlier, the interest here is way too high. So 5% of that is going to be disallowed, but you don't lose that permanently. That would be unfair. And so most jurisdictions allow you to put that 5% away and top up to the 30% in subsequent years. So here for example, we'll get 3.9%, and here we'll probably get the balance. That means we need two lines, one for disallowed, which wants to say 5% there and nothing here. And one for, okay, you disallowed before now you could go up to the limit, which will be this line here. The first thing is we'll need another IF error and we will happen is, we'll, we take this and minus this, and then we'll say, IF that's an error, just give me N/A please. And you'll see that initially it's got encouraging results, but I need a limit so that there's no negative figure here. So I need to go back into this thing and say, okay, I need that to be the max of what you just came up with or zero. And that creates a limit.

The next line will, for all intents and purposes be identical to this line, but will be a min.

So what we could do is we could F2 into this thing, copy the whole lot, come out F2 into the next slide, and that could be described as a hard paste. So really we're just regenerating the line and we can't just press control C, control V because we'll end up with all sorts of absolute and relative cell referencing problems.

Now, if I put a min in, you can see that we've got the extra up to the allowance and it's appearing nicely. If we want that to be a positive figure, however, we should go in and just make sure, that we flip the sign on that.

Okay, so we've just analyzed the situation as to in percentage terms, we're now going to move to absolute terms and we're going to create a base account for the disallowed interest. Given that we're starting this up from scratch or we start from zero, the amount that we disallow in the year will be the percentage of EBITDA.

Okay? Now it's tempting to say, ugh, the disallowed is picking up from here and then it's picking up an error. So why don't we, instead of putting an N/A here, put a zero. The problem is if we actually go ahead and do that, if we put a zero here, really strange things happen because we're deemed to have zero interest. And so although this seems to work, we'll end up with very strange distortions down the line. So we can't do that. What we need to do is yet another IF error, and this time perhaps we could put a zero in.

Now let's get the base count working. And so we'll do an auto sum like this copy to the right, and what we should find is that we've got the disallowed being built up, and then basically staying put, were we to increase our interest expense in subsequent years artificially very heavily. So for example, to 10,000, we'd find out that there is more being disallowed and you would find the account further being built up. Now that stuff is working, what we need to do now is get the usage in the year. This is complex because we can't just have it reach in and grab 3.9 or 11.6 because we'll find that we'll run out of extra interest very quickly and we'll end up making up interest or driving this account into negativity.

So a little bit like the NOL account, and if we take a quick look at the NOL account, you can see that we've initially got some logic to say if there's still funds in the account, the NOL account that is then go ahead and use them. Otherwise don't. We'll have to do the same thing here for disallowed interest, there is money in disallowed interest. Then go ahead and grab the minimum of what's left or the amount that is being charged, and we'll probably find there's another error that's going to pop out of this.

So let's let it work out. Okay, it seems to work initially and it seems to be working fine. Now that's probably only working fine because it's not interacting with the N/As in any meaningful sense. If we were being super neat and tidy, we would need to wrap this in an IF error as well. So let's go ahead and do that. Although it's going to lead to a slightly clunky formula, it is a neat and tidy thing to do, and it may save us some headaches down the line when we fully populate the model. All sorts of unpredictable things can happen, and so seemingly innocent things can come out to play. You can see we've revealed another error, which is, as you are starting to see, one of my favorite errors, which is that we need to get the sign right, and so we've just attached a minus 1 to that min outside the parentheses, and you can see this is now working as intended. We're building up a disallowed interest and we're subsequently using it. So if we were to stress test this, let's say we had that massive, we would create more and then subsequently use it. To round out the tax, we'll need to let the thin capitalization interact with the NOLs. You might recall that losses, for example, here are being created very often by the interest itself. And so given that we've just taken that interest away a lot of the time, okay, so we've just taken some of that interest and disallowed it, then it would be unfair for the interest to fully work against the profit before tax to create NOLs. It'd be double counting it. So to round things out, what we need to do is a thin cap adjustment, and what we're going to do is drop in the product of the disallowed and then used.

And what it does is it really tames that NOL down because a lot of the NOL was being created from interest, which we have disallowed, and this is the kind of complexity that might be missing if you look at these on an individual basis. But they do interact with each other in quite a complex way, but they need to do that again down in taxable profits. Just as a summary, say thin cap adjustments, and you can see now the two aspects working together because you can see that thin cap has taken some of the loss away in the form of disallowed interest, and NOLs have taken the remaining loss away, which has led to zero taxable profit. Then in subsequent years, those losses be they, because of disallowed interest or net operating losses are allowed to come back out again and create taxable expenses for the project going forwards.