XML 105 R38.htm IDEA: XBRL DOCUMENT v3.10.0.1
EARNINGS PER SHARE (EPS) (Tables)
9 Months Ended
Sep. 30, 2018
EARNINGS PER SHARE (EPS)  
Computation of basic and diluted EPS
Three Months EndedNine Months Ended
September 30,September 30,
(dollars in millions, except per share data)2018201720182017
Numerator for EPS:
Income (loss) from continuing operations$(1,220)$(1,712)$661$609
Less: Net income from continuing operations attributable to noncontrolling interests-26540
Income (loss) attributable to AIG common shareholders
from continuing operations(1,220)(1,738)656569
Income (loss) from discontinued operations, net of income tax expense(39)(1)(40)7
Net income (loss) attributable to AIG common shareholders$(1,259)$(1,739)$616$576
Denominator for EPS:
Weighted average shares outstanding — basic895,237,359908,667,044902,081,555938,130,832
Dilutive shares--14,736,71423,165,114
Weighted average shares outstanding — diluted(a)(b)895,237,359908,667,044916,818,269961,295,946
Income (loss) per common share attributable to AIG:
Basic:
Income (loss) from continuing operations$(1.37)$(1.91)$0.72$0.60
Income (loss) from discontinued operations$(0.04)$-$(0.04)$0.01
Income (loss) attributable to AIG$(1.41)$(1.91)$0.68$0.61
Diluted:
Income (loss) from continuing operations$(1.37)$(1.91)$0.71$0.59
Income (loss) from discontinued operations$(0.04)$-$(0.04)$0.01
Income (loss) attributable to AIG$(1.41)$(1.91)$0.67$0.60

(a) Shares in the diluted EPS calculation represent basic shares for the three-month periods ended September 30, 2018 and 2017 due to the net losses in those periods. The shares excluded from the calculation were 13,538,168 and 22,459,868 shares, respectively.

(b) Dilutive shares included our share-based employee compensation plans and a weighted average portion of the warrants issued to AIG shareholders as part of AIG’s recapitalization in January 2011. The number of shares excluded from diluted shares outstanding was 5.8 million and 4.7 million for the three- and nine-month periods ended September 30, 2018, respectively, and 2.4 million and 2.0 million for the three- and nine-month periods ended September 30, 2017, respectively, because the effect of including those shares in the calculation would have been anti-dilutive.