Marlin Global faces another test after portfolio falls 6.3%
Marlin Global’s portfolio fell 6.3% before fees and tax in FY26 while its benchmark gained 30.5%, producing a NZ$17.3 million net loss. The Fisher Funds-managed vehicle is responding with broader sector coverage, faster portfolio turnover and an expanded research team.
- NZ$17.3m net loss versus NZ$0.3m profit
- Gross portfolio return of -6.3% against a 30.5% benchmark gain
- Adjusted NAV return of -8.5% and total shareholder return of -7.6%
- 7.35 cents per share paid in quarterly dividends
- 57.3m warrants issued with an April 2027 exercise date
Portfolio misses benchmark by 36.8 percentage points
Marlin Global Limited (NZX:MLN) has delivered a bruising full-year result after its quality-growth portfolio landed on the wrong side of an AI-fuelled market rotation. Gross performance fell 6.3% for the year ended 30 June 2026, compared with a 30.5% gain for the company’s benchmark, leaving a 36.8 percentage-point gap before fees, tax and other costs.
The result translated into a NZ$17.3 million net loss, versus a NZ$0.3 million profit a year earlier. Adjusted NAV fell 8.5% and total shareholder return declined 7.6%, while net asset value per share dropped to NZ$0.80 from NZ$0.95. Marlin’s share price closed the year at NZ$0.77.
AI concentration exposed sector positioning
Chair Fiona Oliver said the board was “disappointed” with the result. The report attributes the underperformance to an overweight position in healthcare, an underweight position in US financials, limited or no exposure to some semiconductor stocks, and several company-specific setbacks.
That diagnosis matters because the global market’s strongest gains came from semiconductors and other direct beneficiaries of AI infrastructure spending. Marlin’s manager said high-quality businesses underperformed lower-quality companies by 40% during the year, with healthcare, software and consumer holdings among the areas left behind. The portfolio still held AI-linked names including Microsoft, Amazon, Nvidia, Meta Platforms, ASML and Taiwan Semiconductor, but its exposure was concentrated in what Fisher Funds described as businesses with more durable economics.
The winners and losers show the scale of the dispersion. ASML returned 150% in local currency, Alphabet gained 104%, while Boston Scientific fell 59%, Zoetis declined 54%, Salesforce dropped 42% and Gartner fell 61%. Gartner was ultimately exited after the manager concluded that AI-related risks to its research and consulting business had increased.
Manager changes portfolio process
Fisher Funds says it has added two analysts, taking the research team supporting Marlin to seven, and moved towards a sector-specialist model covering technology, healthcare, consumer, industrial and financial stocks. It also added nine holdings and exited four during the year, with Old Dominion, Keyence and TSMC among the newer positions that contributed positively.
The manager says it is moving faster to sell holdings where the investment thesis has changed or valuations have become stretched, while modestly increasing the number of portfolio companies across sectors and geographies. Those are stated process changes rather than evidence of a recovery: the manager’s own outlook is that improved returns are possible, but it acknowledges that a lasting shift away from AI leadership remains too early to call.
Dividends continue as warrants add a capital test
Marlin maintained its policy of distributing 2% of average NAV each quarter, paying 7.35 cents per share during FY26. The next dividend is 1.62 cents per share, payable on 25 September 2026. Management fees were reduced by NZ$1.019 million through a credit note after the portfolio underperformed the relevant hurdle, and no performance fee was charged.
The company also issued 57.3 million warrants for free to eligible shareholders, with one warrant for every four shares held. The warrants are exercisable on 23 April 2027 at an exercise price initially set at NZ$0.87, reduced by relevant dividends. At year-end, the adjusted exercise price was about NZ$0.85 against a NZ$0.77 share price, meaning the warrants were not included in diluted earnings per share because they were antidilutive.
Board reset and next performance checkpoint
There were no share buybacks during the year, while two long-serving directors departed: Carol Campbell retired in December 2025 after almost 14 years, and former chair Andy Coupe retired on 30 June after 13 years. Fiona Oliver became chair, Dan Coman joined the board during the year and Simon Flood was appointed effective 1 June. The current board comprises four directors from 1 July 2026.
The next test is not the dividend calendar but whether the manager’s process changes can narrow the performance gap. Marlin’s portfolio remains heavily exposed to North America, which represented 74% of the portfolio at year-end, while information technology, consumer discretionary and healthcare together accounted for 65%. The 6 November annual meeting and the April 2027 warrant exercise date offer shareholders two formal points to assess whether the portfolio reset is producing more than a change in explanation.
Bottom Line?
Marlin has acknowledged a severe performance failure, but the expanded research team and broader portfolio now have to demonstrate that the problem was positioning rather than a deeper weakness in the investment process.
Questions in the middle?
- Can broader sector coverage improve returns without diluting Marlin’s quality-growth discipline?
- Will the manager’s faster exit process prevent legacy holdings from becoming repeat detractors?
- How many of the 57.3 million warrants will be exercised in April 2027, and what effect could that have on the share count?