Ann: Preliminary Final Report, page-3

The warrants were on issue in 2020. They should not be in the EPS number because they are out of the money. If they were included then the subscription price of $0.50 would also need to be accounted for - and would be a agin because $0.50 is more than the NTA.

There are two ways I can get close to the reported EPS for 2021 is by taking the 1.8b shares on issue prior to the CR, multiplying it by 2.21 to get 5.8b and then applying this number to the $45m which at 0.78c is still too high. Of course the CR was 1 new share for 2.21 existing shares, not 2.21 new shares for each existing share. Would have BA made this mistake? Unlikely.

The other way, and the more likely way, is that the uplift in NPAT from the "gain from debt modification" has been deducted from the EPS calculation as an extraordinary/abnormal item to give a pro-rata normalised EPS.

$46.2m - $27m = $18.2m over 2.6b shares get me an EPS of 0.7c - if there was a tax expense associated with the $27m (likely as the rescheduling would have affect interest cost) then we could be at 0.072c if the tax shield is $600k. This difference might be attributable to the non-controlling interest, but because this adjustment probably relates to Medco's 5% stake in EEES, and because the debt rescheduling is applied to Slipform/EWI debt attributed to the new projects - I do not think this is likely.