So we're tempted to think, well, iterations is the answer to everything. We have a circular reference, just turn iterations on and it will be fine. I want to demonstrate to you why that's not true. I'm gonna find some free space at the side here, and I'm going to type in a few numbers. 1, 2, 3, and you'll see that the numbers at the side we're recalculating every time I input those numbers, even though none of these are linked into this calculation. So having iterations on does slow your model down because it recalculates every time you input something, even if it doesn't affect the calculation that is circular. Now, let's say I would like to add all of these up. So I do a sum starting with one, but I make a mistake. Instead of stopping there and just adding up K14 to K17, my finger slips or I get distracted and I go too far.

Normally, if I press enter, now Excel would warn me that's a circular reference, but because I've got iterations on it accepts that circular reference and it does what it's told to do. It goes round the loop 100 times or until the numbers stop changing by more than 0.001. So in this case, 1 plus 2 plus 3, it should give me an answer of 6, but because the sum includes itself, it goes round that loop just like any other circular reference, and it gives me this erroneous answer of 1 plus 2 plus 3 gives me a total of 600. That's the problem with using iterations. It has no way of distinguishing between a deliberate circular reference like this one. I need to borrow enough money to pay the interest, which causes a circular reference or this one, which is just a simple mistake.