# AVI Global Trust NAV falls 4.1% in June 2026 - Link: https://www.directorstalkinterviews.com/avi-global-trust-nav-falls-4-1-in-june-2026/4121256034 - Published: 2026-07-13T07:10:02+01:00 - Author: Ethan Davis **AVI Global Trust plc** (LON:AGT) has presented its Update, reporting performance figures for the month ended 30 June 2026. This investment management report relates to performance figures to 30 June 2026. | **Total Returns (%)** | **Month** | **1Y** | **5Y** | **10Y** | | NAV p/s^(1) | -4.1 | 5.5 | 40.1 | 183.3 | | MSCI ACWI^(2) | 0.7 | 27.7 | 75.2 | 235.4 | | MSCI ACWI ex US^(2) | 1.0 | 31.8 | 58.6 | 159.5 | All performance shown net of fees in GBP Total Return as at 30/06/2026. ^(1)Net Asset Value cum-fair. ^(2)From 1st October 2023, the comparator benchmark was changed to the MSCI ACWI Index. Prior to this, from 1st October 2013, the comparator benchmark was the MSCI ACWI ex US Index. Source: Morningstar, S&P Capital IQ AVI Global Trust has reported that its NAV declined by -4.1% in June 2026. Samsung C&T added +31bps to NAV. Beyond this, however, there were few meaningful other positive contributors. On the other hand, the list of meaningful detractors was longer. Partners Group Private Equity was the largest detractor (-90bps), followed by HD Hyundai (-68bps) and Vivendi (-52bps). We have added a number of new positions to the portfolio in recent weeks and months, as well as exited some. We discuss two such names below. **Rohm:** We have recently initiated a position in Rohm (6963). It was Churchill who described Russia as “a riddle, wrapped in a mystery, inside an enigma” and whilst we wouldn’t be so preposterous to describe Rohm as such, there are certainly many levels of value to unpick. At the simplest level, it is a vertically integrated semiconductor component manufacturer, whose core business appears to be moving from the wrong part of the cycle to the right one. More interestingly, it owns a large stake in unlisted Toshiba, which in turn holds a meaningful position in Kioxia, now Japan’s largest listed company. Taken together, these layers create the opportunity to own a collection of attractively growing assets, at a discounted valuation, with numerous levers to unlock value. Rohm’s history is rooted in resistors, an electrical component that restricts the flow of electric current in a circuit. During the transistor radio boom of the 1950s and 60s it controlled roughly 60% of the Japanese domestic resistor market. Over the following decades it pushed deeper into semiconductors and related components. Today its main business lies in power electronics: integrated circuits that manage and condition power inside end-products, and discrete devices such as power switches and diodes. This is not a glamorous area of the semiconductor industry, but it is an important one. Power devices sit behind the more visible logic chips and GPUs, enabling electric vehicles, industrial automation and – increasingly – data centre servers to operate safely and efficiently. Within that, silicon carbide (“SiC”) has become the preferred switching technology for EV inverters, which can lead to 10% more driving per charge versus traditional silicon, and is now finding its way into next-generation AI servers, where efficiency and thermal performance matter. Both of Rohm’s main segments have always had some cyclicality and the business has felt the effects of the FY2024-25 downturn. However, the set-up now looks quite different. Rohm has been selected by Nvidia as a partner for a new 800V server power architecture, as part of a design-in relationship, co-developing power-delivery solutions specifically optimised for Nvidia’s next GPU platforms. If this design wins broader adoption, it should anchor a more durable earnings stream on top of the usual cyclical upswing. Forward estimates reflect this. From FY2027 to FY2029 Rohm is expected to grow EBIT at around 51% per annum, ahead of a peer average in the mid-30s. On headline numbers the shares do not look especially cheap – the reported NTM EV/EBIT multiple sits in the 50s – but that figure muddles the picture by treating Rohm as a simple semiconductor company and ignoring the value of its Toshiba stake. Once that stake is stripped out, the implied multiple on the operating business falls to around 12x, versus peers closer to the mid-30s. The Toshiba position is the second leg of the case. Toshiba was taken private in late 2023, in a ¥2tn buy-out led by Japan Industrial Partners (JIP), with a consortium of Japanese corporates – including Rohm, Orix and Chubu Electric Power – each investing ¥100bn for a “14% plus” stake. On our numbers, Rohm’s share of Toshiba is worth roughly two-thirds of Rohm’s current market capitalisation. Even under a more conservative scenario that applies a 50% haircut to Kioxia’s market value, the implied value of Rohm’s Toshiba stake still equates to around half of Rohm’s market cap. There are reasons to think Toshiba itself may be worth more than current implied figures. Operationally it has stabilised and begun to grow again, with operating margins recovering to just over 8% in FY2025 and management targeting 10% for FY2026. Its end-market exposures are attractive – grid and energy infrastructure, Japanese defence rearmament, and hard drive storage for AI data centres – and management and JIP appear to be working towards a relisting on the Tokyo Stock Exchange as early as FY2028, with segment carve-outs a plausible precursor. Each of those steps offers a way for value to be crystallised. Kioxia adds a third layer. Historically NAND flash memory has been one of the most difficult parts of the semiconductor industry, characterised by rapid technology- driven supply growth and noisy cycles. The rise of AI, and inference workloads in particular, has shifted that backdrop. Demand for NAND in high-performance SSDs and data-centre storage has become more structural at a time when key competitors are prioritising HBM. Kioxia combines that tailwind with a meaningful cost advantage, and margins are expected to be extremely high through the middle of this decade. As Toshiba’s largest shareholder, Rohm participates in that upside indirectly, with further stake sales already underway since Kioxia’s listing. Putting this together, Rohm trades on an estimated 36% discount to NAV with the market effectively assigning little value to the improvement in its core business and only partial value to its balance-sheet assets. On our numbers, applying a conservative 50% haircut to Toshiba and assuming Rohm’s stub re-rates in line with peers by 2027 gives a price target of 7,811 yen, around 46% above today’s level. It is always possible that cycles disappoint or that corporate events take longer than hoped. The attraction here is that Rohm does not rely on a single narrow outcome. As with a number of our Japanese holdings, the combination of good assets, over capitalised balance sheets and wide discounts continues to provide fertile ground for patient capital. **Cordiant Digital Infrastructure:** Over the month we exited the last of our position in Cordiant Digital Infrastructure(“ CORD”), after selling down steadily into the demand that arose from the company’s entry into the FTSE indices. The investment – which was initiated in February 2024 – was a successful one, generating an ROI of +85% and IRR of +40%. This compares favourably to the MSCI AC World Index which returned +36%/+18% respectively. Our average entry discount was 43% vs an exit discount of 18%, with NAV growth of 31% (12% annualised, ahead of the company’s target at IPO of 9% per year). The investment was predicated on the misperception of CORD as having similarities to troubled peer, Digital 9 Infrastructure. In reality, CORD’s balance sheet, acquisition discipline, portfolio characteristics, and management quality were markedly different. CORD’s portfolio, mainly comprised of two broadcast companies in Poland and the Czech Republic, was seen as far riskier than it actually was. Both companies offered highly secure contractual and inflation-linked revenues from essential services with a clear plan under Cordiant’s stewardship to redeploy income from this high yielding core business into longer duration, higher growth, and higher terminal value segments such as data centres and mobile towers. As the market came to appreciate this CORD’s discount to core plus infra peers Pantheon Infrastructure and 3i Infrastructure narrowed substantially. While we still see scope for more upside from CORD (noting a SOTP analysis yields a much higher NAV than the official one reported by the company), the substantial discount narrowing we have seen mathematically means that future returns are likely more limited than they have been historically. As such we took advantage of index- related buying to exit the position given the compelling opportunities we see elsewhere within our universe. **Contributors / Detractors (in GBP)^(4)** | **Largest Contributors** | **1- month contribution****bps** | **% Weight^(3)** | | Samsung C&T | 31 | 6.4 | | Mitsubishi Logistics | 13 | 6.2 | | Symphony | 13 | 2.2 | | Dai Nippon Printing | 12 | 2.1 | | Cordiant Digital Infrastructure | 8 | Exited | | **Largest Detractors** | **1- month contribution****bps** | **% Weight^(3)** | | Partners Group PE | -90 | 3.0 | | HD Hyundai | -68 | 1.8 | | Vivendi | -52 | 5.9 | | Hyosung Corp | -43 | 3.3 | | Jardine Matheson | -31 | 5.4 | ^(3)All Figures shown as % of Net Asset Value ^(4)Contributors and detractors from Factset [AGT-Newsletter-JUN-2026](https://www.directorstalkinterviews.com/wp-content/uploads/2026/07/AGT-Newsletter-JUN-2026.pdf) [Download](https://www.directorstalkinterviews.com/wp-content/uploads/2026/07/AGT-Newsletter-JUN-2026.pdf) Share on: