NPAT fell 74.8% on large events while underlying profit rose 6.4%

The gap is driven by catastrophic-weather claims, concentrated in Pacific Islands where the segment result swung from a NZ$6.0m profit to a NZ$2.3m loss. Group revenue (gross written premium) still grew 10.1% to NZ$214.1m, and New Zealand segment result improved to NZ$7.8m from NZ$6.0m.

Cash quality moved with the reported result. Net operating cash inflow fell 57.4% to NZ$25.1m and pre-lease free cash flow was NZ$16.4m, which Annolyse's historical baseline classifies as below the recent range (4-period mean NZ$39.4m, range NZ$17.6m-NZ$58.4m). Capex stepped up sharply from NZ$0.5m to NZ$8.6m (4.0% of revenue), almost entirely intangibles and customer relationships. Total equity declined 14.2% to NZ$306.9m despite a positive result, indicating a material capital distribution between periods.

• Underlying NPAT up 6.4% on revenue up 10.1% says the franchise is growing; the reported NPAT collapse reflects weather claims that Tower itself strips out. For an insurer, large events are a recurring feature year-to-year, so the underlying line is the cleaner indicator of pricing, growth and expense discipline, while the reported line measures how much catastrophe exposure landed in this specific half.

• Pacific Islands carried the event load. The segment swung NZ$8.4m year-on-year to a NZ$2.3m loss, while New Zealand's result improved by NZ$1.9m. The question this raises is whether Pacific Islands reinsurance, pricing or risk appetite need recalibration, or whether HY22 simply captured a concentrated cyclone/storm sequence that will not repeat at the same scale.

• Capital returned even as cash conversion weakened. Pre-lease FCF of NZ$16.4m is the lowest in the recent baseline, and equity dropped NZ$50.9m, consistent with the capital return flagged at FY21. The 2.5c interim is unchanged, but payout against this half's reported NPAT is 357.1% (versus 92.6% prior) and 63.7% of pre-lease FCF (versus a 3-period mean of 38.3%). Distribution capacity now depends on the underlying line, not the reported one.

With underlying H1 already at NZ$18.2m, the implied H2 underlying is NZ$2.8-6.8m, which is meaningfully lighter than H1. The supplied seasonality (HY21 was 61.6% of FY21 reported NPAT) is consistent with H1-heavy earnings, so the guided H2 step-down is shape-consistent rather than a downgrade signal. Management commentary points to continued business performance with large-event impacts mitigated.

What the release does not support is a clean read on reported FY22 NPAT, because that figure depends on whether large-event budgets hold for the remainder of the year. Guidance is stated only on the underlying basis.

Against that, the cash and balance-sheet read is softer. Operating cash inflow fell more than NPAT did, pre-lease FCF is below Annolyse's recent range, and capex stepped up 18-fold on intangibles, signalling a technology spend cycle that will keep FCF below historical norms even if claims normalise.

The effective tax rate of 46.7% is flagged as an unprecedented high against a 4-period mean of 26.2% (range 10.3%-33.0%), versus 33.0% in the prior comparable. This widened the NPAT decline relative to PBT by 5.9 percentage points and is not explained in the supplied excerpts; it should be treated as a one-period distortion until reconciled. ROE of 1.0% (versus 3.2%) reflects both the suppressed numerator and the smaller equity base after the capital return, so it overstates the underlying return decay.

This briefing cannot assess Tower's reinsurance terms, claims development triangles, or the catastrophe budget assumptions embedded in maintained guidance.