Financial markets spent much of the 2026 financial year climbing a wall of worry that, at times, resembled a cliff face. In February 2026, Israel and the United States launched strikes on Iranian nuclear facilities, resulting in a spike in oil prices and a sharp sell-off in global share markets. However, as is the case with most geo-political shocks, markets quickly shrugged this one off. The more enduring narrative of the year was the AI supercycle, the impact of which largely determined the market’s winners and losers. Without an AI story, the ASX200 returned a modest 6.1% for the year, underperforming the global index by around 10%. Unfortunately, it’s becoming a familiar story: UniSuper’s International Shares option has outperformed the Australian Shares option in eight of the last 10 years.
Pleasingly, all of UniSuper's investment options ended the financial year in positive territory. Our default Balanced option delivered 10.4% for accumulation members and 11.2% for pension members, the latter result representing the fourth consecutive year of double-digit returns.
As is customary, we reflect on the year’s best and worst performing assets. With funds under management now approaching $175 billion, our significant holdings tend to be concentrated in very large companies. Accordingly, this review focuses on investments in which we hold more than $1 billion.#
Best performing companies (major holdings)
Alphabet
A year ago, Alphabet ranked among our worst-performing holdings as investors questioned whether AI would erode Google's dominance in search. Twelve months on, that narrative has reversed entirely: the stock returned 93% and became our best-performing position. What changed was Alphabet's transition from perceived AI loser to AI winner. The company has built a compelling end-to-end AI stack, spanning its own custom microchips, large language models (Gemini), and infrastructure through Google Cloud, which has emerged as a major beneficiary of surging enterprise AI spending. Despite this exceptional run in the share price, the stock continues to trade broadly in line with its historical average valuation, leaving us comfortable with our position.
BHP
BHP delivered a standout 68% total for the year, knocking CBA off its perch to reclaim the title of Australia’s largest company. It’s been a rollercoaster ride for the Big Australian; a top performer in FY23, our worst in FY24, and now back in favour as one of our best. While people tend to think of BHP as an iron ore story, more than half the company’s profits now derive from copper. Copper prices are surging, driven by tight supplies, accelerating demand from electrification, and power needs of the data centre build out.
Despite the impressive recent share price performance, at current prices BHP still looks attractively priced in our view.
Apple
Long regarded as the laggard of the AI race, Apple returned almost 35% to take third place among our best performers. Sentiment shifted as investors warmed to the view that Apple can layer AI across its 2.5 billion-odd devices without the eye-watering spending shouldered by rivals such as Microsoft and Google. Demand for the iPhone 17 outstripped supply, revenue from Apple services set fresh records, and a new US$100 billion buyback added support. At the company's June developer conference, an overhauled Siri—for years the punchline of Apple's AI efforts—finally looked the part. The shares are not cheap, but few companies generate cash like Apple, so we remain comfortable holding.
Spare a thought for John Ternus who has the enormous challenge of filling Tim Cook’s shoes as the next CEO of Apple. Cook of course, had the seemingly impossible task of filling what many consider to be the biggest shoes in corporate history: those belonging to Steve Jobs. Cook has done a magnificent job.
Worst performing companies (major holdings)
CSL
CSL has followed a bad FY25 (-18%) with a horrendous FY26 (-51%). To be blunt, in a relatively short period, CSL has gone from a market darling to a dog. While some headwinds are structural and (arguably) beyond the company’s control, there have also been own-goals—chief among them the acquisition of Vifor, which has proved a costly misstep. Profit downgrades have been a recurring theme. CSL peaked at around $340 and, at time of writing, has fallen to $125. We expect it to find support around these levels. However, we offered a similar view in last year’s report, and we were wrong.
Chairman Brian McNamee’s stellar record as CSL’s former CEO is increasingly overshadowed by his lacklustre record as Chairman. He is due for re-election in October 2027, although the expectation is that he will step down sooner. It’s time to go.
Microsoft
Software and cloud heavyweight Microsoft returned a disappointing -28%, and not because the underlying business faltered. Revenue grew in the high teens all year, profits by as much as 25%, and Azure kept expanding at close to 40% year on year. What unsettled investors was the bill for artificial intelligence: Microsoft is spending heavily on the data centres underpinning the AI boom, with capital expenditure up and free cash flow down, and the market lost patience waiting for a return on those investments. Microsoft also has a sizeable applications business, such as Microsoft Office, that AI itself could disrupt. We think the pessimism is overdone. We can see a path where the smart integration of AI can further entrench Microsoft’s control of the office workstation.
So there you have it: three of the world's greatest technology companies, all exposed to the same AI super cycle, yet delivering vastly different returns. Alphabet was rewarded for proving it could compete at the frontier. Apple was rewarded for the opposite reason: restraint, and an installed base large enough to monetise AI without matching its rival’s capital expenditure. Microsoft was punished despite arguably doing everything right on the business side simply because the market has, for now, lost patience with the size of the investment required to remain in the AI race. The AI story has a long way to run, but the market is already picking winners and losers. Don’t be surprised if opinions are reversed by the next time we report.
Best performing option
International Shares topped the table with a 19.6% return for pension members and 17.9% for accumulation members, the fourth consecutive year of healthy double-digit returns. Unsurprisingly, the strong run coincided with the advent of the AI supercycle. The explosion in demand for AI models has fuelled an explosion in demand for power, memory, chips, and datacentres—and that’s just the first order impact. The second order story relates to demand for a vast range of inputs including commodities, critical minerals, cooling technology, and a whole range of services. The tech tide is lifting all boats and the countries best placed (US, Japan, Korea, and Taiwan) are enjoying healthy market returns. With around 65% of the option invested in the US and around 35% invested in the tech sector, the portfolio remains exposed to the same drivers that has underpinned its recent success. Of course, that also means it is exposed to a correction at some point. We just don’t know when.
Worst performing option
The Australian Bond option struggled in FY26, returning just 1.2%. Over the past financial year, the 10-year Australian Government bond yield rose from 4.16% to 4.72%, a significant move as rising inflation concerns prompted the RBA to reverse course and hike rates three times after a period of cuts. Rising yields equate to falling bond prices, and those losses more than offset the modest income the bonds generated during the year.
The silver lining is that the Australian Bond option now carries a higher prospective running yield than it did a year ago, which should support better returns ahead—assuming yields don't rise further. The challenge, as always, lies in predicting where inflation and central bank policy head from here. That remains a toss-of-the-coin.
Watch John Pearce’s latest video.
# All returns quoted include dividends. The Defined Benefit (DB) is excluded from the discussion as movements in share prices impact the DB portfolio returns but do not affect members’ balances given the (formula-driven) way in which benefits are determined.
John Pearce is Chief Investment Officer at UniSuper, a sponsor of Firstlinks. This article provides general information and may include general advice. It doesn’t take into account your financial situation, needs or objectives. Consider your situation before making financial decisions, because we haven’t, as well as the PDS and TMD relevant to you at unisuper.com.au/pds, and whether to consult a qualified financial adviser. Past performance isn’t an indicator of future performance.
For more articles and papers from UniSuper, click here.