Manchester&lon. Share Price (MNL)

For discussion....

Compared with Blue Whale Capital / Polar Capital Technology Trust / Allianz Technology Trust, MNL increasingly looks like the highest-conviction vehicle of the group.

A few things stand out to me:

MNL’s portfolio concentration is extreme even by tech trust standards. Recent disclosed holdings show huge weightings in NVIDIA, but also meaningful allocations to less crowded names like Lumentum Holdings, Ciena, Coherent Corp and Bloom Energy.

That’s different from ATT/PCT, where the Mag 7 exposure is still generally more dominant and diversified across software/platform winners. MNL seems more willing to hunt the “picks and shovels after the obvious picks and shovels”.

The gearing matters. If they’re effectively ~118% invested, that creates real torque if this AI capex cycle broadens beyond the hyperscalers. But it also means drawdowns can become vicious very quickly. This is not a “sleep at night” trust.

The interesting thing — and I think this is what you’re reacting to — is that the AI narrative may now be shifting from:

model creators/software,

to

physical infrastructure bottlenecks.

And MNL seems unusually aligned to that phase:

optics,

networking,

power,

cooling,

memory,

semiconductor equipment,

data-centre infrastructure.

That’s potentially where alpha could come from if the market starts rewarding the enablers rather than simply recycling into the same seven mega-caps.

The other thing in MNL’s favour is that it genuinely behaves differently. A lot of tech trusts have quietly converged into “slightly active Nasdaq”. MNL still looks idiosyncratic and manager-led. Sometimes that’s dangerous; sometimes that’s exactly where outperformance comes from.

That said, I’d still push back on one part slightly:

MNL is still heavily dependent on the AI capex cycle remaining hot. Even if it’s “outside Mag 7”, many of the holdings are economically downstream from the same spending wave. So diversification may be less than it appears in a stress scenario.

And because of the leverage + concentration, the volatility profile could be dramatically worse than ATT/PCT during any AI wobble or recession scare. You probably know this already, but MNL is closer to a hedge-fund style expression inside an investment trust wrapper.

What does feel different versus 2022–2024 is:

there’s now actual earnings acceleration across the supply chain,

not just narrative expansion.

The recent factsheet commentary on companies like Taiwan Semiconductor Manufacturing Company, Vertiv and Bloom Energy points to real infrastructure demand rather than speculative future demand.

So overall: yes, I can understand why this is the first time in years you feel MNL may actually have a structural edge versus ATT/PCT/Blue Whale rather than simply being “higher beta tech”. The combination of:

concentrated conviction,

gearing,

does create a plausible route to outsi

No advice

There are documented issues with Nvidia's Vera Rubin (sometimes called Rubin or VR) platform rollout, primarily supply and technical challenges rather than fundamental design flaws.

Key points from recent reports (as of early-mid 2025):HBM4 memory bottlenecks: Delays in qualifying and scaling high-bandwidth memory (HBM4) from suppliers like SK Hynix and Micron. This has led analysts (e.g., TrendForce, KeyBanc) to cut expected 2026 Rubin GPU shipment share from ~29% to ~22% of Nvidia's high-end GPUs, with production potentially reduced from ~2M to ~1.5M units.

Other engineering hurdles: Higher power consumption, advanced liquid cooling requirements, networking transitions (e.g., CX8 to CX9), and overall rack-scale integration complexities in the unified Vera Rubin platform.

These are short-to-medium term headwinds (next few months/into 2026 ramp), not existential. Nvidia has a strong track record of overcoming them, and Blackwell continues to drive revenue. Rubin is still expected to ship in volume in H2 2026 with big efficiency gains (e.g., much higher performance per watt).

iMark Sheppard (@MLCapMan) explicitly cited these rollout struggles, over-specification in the rack-scale design, potential near-term market share losses to AMD/others, and the need to reduce single-stock concentration risk (from an extreme ~42% to 24.5%). He remains bullish long-term, targeting $500 NVDA in 3 years, but wants more patience and diversification.

This aligns with broader market chatter: It's a real but manageable setback in a supply-constrained AI boom.

On the new holdings layout/preferencesReducing NVDA from ~42% to ~24.5% (still by far the largest position) and broadening exposure makes sense for risk management in an investment trust. Extreme concentration worked well in the AI rally but amplifies downside on any hiccup (supply delays, competition, valuation resets).

From recent factsheets/portfolio updates, the shift includes:Keeping big core AI names like Broadcom (~11%) and TSMC (~10%).

Adding/increasing diversified plays: VanEck Semiconductor ETF, Morgan Stanley Broad AI Basket, stronger Bloom Energy (data center power), and AMD entering top holdings.

I prefer the new, more diversified layout. 40%+ in one stock is aggressive even for a tech-focused trust—reducing it while staying heavily AI-exposed lowers volatility without abandoning the thesis. It gives breathing room for other winners (e.g., if AMD gains share short-term or power/infra names like Bloom run). Sheppard has a history of decisive moves (e.g., slashing Microsoft earlier), which has generally served the trust well over time.Bottom line: The concerns are valid and public (not obscure), justifying a trim. Long-term, Nvidia's AI dominance looks durable, but diversification here is prudent. Always check the latest factsheet/newsletter from M&L Capital for full deta

No Advice

NVDA is still our biggest position and we still believe that the stock will see $500 in 3 years time, but the next 3 months may see Vera Rubin delays and further losses in Market Share that we have already anticipated. NVDA had the opportunity with Vera Rubin to grasp the CPU market but by over specification to the Vera Rubin racks, they have dropped the ball through delays. It is a set back that we believe will be overcome but it gives AMD and ARM some space to run. More on this matter can be read in the section about AMD results below.

Interestingly, when we produced our full model in the last Newsletter one eagle eyed shareholder pointed out that Morgan Stanley are now forecasting Hyperscaler Capex spend in 2027 of $1.1trn which is now HIGHER than our model forecast of $1trn. Note this is the first time any broker has punched in with a forecast higher than ours and we would note that when we first published these forecasts many people said we were deranged. This is why we retain a high level of confidence that NVDA can hit our $500 target price but some painful patience may well be required.

Our Portfolio is now more diversified. We have positions in Memory, CPUs, GPUs, TPUs, Optics, Networking, Ai Model Labs, Semi Fabrication, Semi Equipment, Energy, Power & Cooling, Financial Applications, Neo-Clouds, Robotics, and Space.

No Advice

BUT we also but we remain big bulls of NVDA on a 3 year view, but NVDA must escalate and accelerate their Share Buy Backs in H2 2026. Results are forthcoming in the middle of May. We are data dependent, and will flex accordingly post the input of further data.

Important portfolio changes announced last night.

NVDA holding reduced to 25%

No Advice