Monthly Fund Focus: A defensive income and growth fund, Agentic AI and a Future Supply Chains ETF

David Stevenson returns with his latest fund roundup, highlighting a defensive income fund focused on financials, a new ETF targeting the reshaping of global supply chains, and a simple momentum strategy for ETF investors. He also examines Morgan Stanley’s latest thinking on Agentic AI.

In this month’s funds roundup, we look at a high-yielding defensive fund focused on the financial sector, a new ETF aiming to capitalise on supply chain challenges reshaping the global economy, a simple momentum strategy for investing in country exchange-traded index tracking funds, and Morgan Stanley analysts’ top Agentic AI picks.

A defensive income and growth fund: The Algebris Financial Income Fund

If you’re looking for a fund that generates both income and capital growth while keeping well away from anything AI-related, the Algebris Financial Income Fund is worth a look. It’s relatively unknown amongst UK private investors but has a genuinely interesting thesis built around European banks, and it’s available on most UK investment platforms.

Algebris is a London-based manager that has been doing exactly one thing since 2006: investing in financial companies and the debt they issue. That singular focus has given them serious expertise in bank stocks and credit. They manage around €35 billion across a range of strategies, and the Financial Income Fund is one of their flagship products at around €2.7 billion in assets, launched in August 2013.

The core idea is straightforward. The fund blends two types of financial sector exposure, equities and credit, extracting income from both while also targeting some capital appreciation. On the equity side it holds high-dividend-paying bank and insurance stocks. On the credit side it focuses on AT1 bonds, also known as CoCo bonds, which are the more junior, higher-yielding parts of bank balance sheets. The split between the two can swing between roughly 30% and 65% each, giving the manager flexibility to tilt toward whichever looks more attractive. Right now the allocation sits around 32% equities and 46% bonds, with the rest in cash.

Performance since launch has been solid: around 208% in total, roughly 9.3% annualised, with volatility of 12.1% and a Sharpe ratio of 0.8. Recent years have been particularly strong, with gains of 20%, 15%, 19%, and 21% in 2021, 2023, 2024, and 2025 respectively. The 2026 year-to-date return is under 1%, reflecting the choppy broader market backdrop.

What is an AT1 bond?

AT1 stands for Additional Tier 1, a type of bond that banks issue to meet regulatory capital requirements. They sit just above equity in a bank’s capital structure and were created after the 2008 crisis to ensure banks had a layer of capital that could absorb losses before taxpayers stepped in. If a bank’s core capital ratio falls below a certain threshold, the AT1 can either convert into equity or be written down entirely, hence the CoCo nickname, short for contingent convertible.

European bank capital ratios now average between 12% and 15%, comfortably above any trigger point. That said, Credit Suisse is a reminder that blow-ups can still happen and AT1 investors can be left with nothing. On that score, Algebris avoided Credit Suisse entirely. That tail risk is precisely why AT1s offer the yield they do: the fund’s bond portfolio yields around 6% to worst, with AT1s offering around 257 basis points of spread over European corporate bonds, more than most conventional high-yield alternatives, from banks that carry investment-grade ratings overall. Supply is also shrinking, as the initial regulatory build-out of the asset class is now complete and the market has moved from net new issuance to largely just refinancing.

Why European bank equities?

European banks spent the decade from 2012 to 2022 in a near-zero rate environment, the worst backdrop imaginable for an industry that makes money by lending at a spread above funding costs. When rates normalised from 2022 onwards, net interest income surged and earnings were revised sharply higher. Algebris argues that profitability is durable even as the ECB considers cutting, because the yield curve is steepening in a way that actually favours banks. Loan growth is recovering, asset quality is holding up, and, at around 10 times forward earnings, European banks are the second-cheapest sector on the continent.

The fund’s top equity names include Intesa Sanpaolo, Deutsche Bank, Barclays, and UniCredit. The bond portfolio is anchored by UBS, Barclays, Santander, and BNP Paribas. Overall gross yield is around 5.7%, management fees are 0.90%, and the fund offers daily liquidity. The one note of caution is meaningful exposure to US banks and insurers, where the risk profile looks somewhat higher than in Europe right now.

Lazy momentum strategies using ETFs

One of my favourite strategists, Joachim Klement at Panmure Liberum, directed his readers recently to an excellent research paper on Lazy investing and momentum strategies. The idea of lazy portfolios – there’s a whole subculture of Bogleheads in the US dedicated to it – is to use diversified ETFs to build a long-term portfolio around a single, simple-to-understand strategy. That strategy is sometimes based on simple asset-allocation diversification but frequently borrows heavily from technical trading and trend-following. The core ideas tend to either be a:

• Focus on momentum trends upwards to capture the best, risk-adjusted returns
• Or use a strategy to minimise downside risk by looking at technical indicators such as the moving average over 20 and 200 days.

One of the more recent iterations, via Joachim Klement here, is Javier Estrada’s The Lazy Man’s Momentum Strategy, an academic at the IESE Business School. You can read the paper here: IESE Business School Working Paper. His idea is incredibly simple: build a diversified portfolio of ETFs from around the world and reshuffle it every six months using momentum metrics.

Momentum investing has a long and well-documented track record. Buy the recent winners, ditch the recent losers, rinse, repeat. Institutional money has been running versions of this playbook for decades, and the academic literature has consistently found that stocks that perform best over the past three to twelve months tend to continue outperforming over the next three to twelve months. The problem for most private investors is that classic momentum strategies demand constant attention: monthly portfolio shuffles, large stock universes, and the kind of operational infrastructure that individual investors just don’t have.

Estrada looks at a dataset covering the full MSCI universe of developed markets — 23 countries in total, with the earliest data dating back to December 1969. Every six months, each country is ranked by its total return over the preceding half-year. The top 8 to 11 countries go into the Winners portfolio, each weighted equally. The rest go into the Losers portfolio, also equal-weighted. Then you sit on your hands for six months, check the scores again at the next rebalancing date, and shuffle accordingly. That’s it.

How did it perform? Looking at the period from June 1970 through December 2024 (54 years), the Winners portfolio returned an annualised 13.1%, compared with 10.3% for the MSCI World benchmark. That 275 basis-point annual advantage is valuable: over time, £100 invested in the Winners portfolio in 1970 grew to £81,600 by the end of 2024, compared with just £21,320 in the benchmark. The volatility premium you pay for that outperformance is relatively modest. The Winners portfolio has an annualised volatility of 16.9%, compared with 14.8% for the benchmark — roughly 2 percentage points higher.

Source: Estrada (2025). Returns nominal, USD, including dividends. Terminal values assume constant compounding at the annualised geometric mean.

Transaction costs and taxes aren’t factored into the analysis, which is a standard academic caveat but worth keeping in mind. The paper’s argument is that the 275-basis-point annual return advantage is large enough to absorb realistic trading costs and still leave a meaningful excess return.

Agentic AI: Morgan Stanley’s top bets

The Ai story is changing fast. Until recently, the main narrative was centred on super-fast, smart LLMs that answered all your prompts and queries. The new, new thing is what’s called Agentic AI : your own artificial intelligence that goes away and solves all your problems without you interfering. This new AI model puts different pressures on both the AI models and the equipment suppliers selling to them, i.e., different chipsets might be required. Analysts at US investment bank Morgan Stanley recently crunched the data and identified how agentic Ai changes token usage. As a result of this key switch in AI usage, Morgan Stanley has revised its total addressable market projections upward, materially so:

Morgan Stanley estimates that there will be around 1 billion knowledge workers globally by 2032, with AI adoption among that group rising from 78% in FY26 to 99% by 2030. Right now, they estimate 6 agents running per session on average -but that figure is expected to grow to nearly 100 agents per session by 2032, with the analysts noting even that “may still be conservative as exponential growth would imply trillions of agents per session”. Meanwhile, simultaneous agentic sessions are modelled to grow at a staggering 176% CAGR from FY26 to FY30.

CPU pricing also has a counterintuitive tailwind here. While the price per core is expected to decline 10% annually, core counts per CPU are set to leap from 64 today to 130 by FY27 and 200-500+ by 2030. The net result is that the average selling price per CPU rises from over $2,000 in FY27 to at least $3,000 by 2030.

As for the companies most likely to benefit from this trend, the Morgan Stanley analysts list some well-known candidates, including:

AMD tops the list of potential beneficiaries. On its Q1 2026 earnings call, management raised its server CPU TAM outlook, projecting growth of more than 35% annually to over $120bn by 2030, up from a prior CAGR of around 18%, explicitly citing agentic AI as the driver. Data Centre revenue came in at $5.8bn, up 57% year-on-year, with server CPU revenue growing more than 50% YoY and Q2 guidance pointing to growth above 70% YoY.

Arm Holdings is another big winner. The company reported record Q4 FY2026 revenue of $1.49bn and full-year revenue of $4.92bn, with data centre royalties more than doubling year-on-year. More telling still: customer demand for Arm’s new AGI CPU, co-developed with Meta and explicitly positioned for agentic infrastructure, already exceeds $2bn for FY27-28, more than double what was indicated at its “Arm Everywhere” event. Arm also says its CPU compute share sits at around 50% among top hyperscalers.

Intel has also caught the agentic tailwind, explicitly stating on its Q1 2026 release that the shift from foundational models to inference to agentic AI is increasing demand for its CPUs, wafer capacity and advanced packaging business. DCAI revenue hit $5.1bn, up 22% YoY, with Xeon 6 singled out as the host CPU for NVIDIA’s own DGX Rubin NVL8 platform and Google continuing its Xeon deployments. Intel’s foundry and packaging angle gives it a somewhat unique position in the agentic supply chain.

The Meta and AWS deal might be the clearest single proof point in the entire report. Meta signed an agreement with AWS to deploy tens of millions of Graviton cores specifically for agentic AI workloads. AWS described the demand as CPU-intensive, driven by real-time reasoning, code generation, search and multi-step task orchestration. Meta’s own characterisation was refreshingly honest: “agentic AI is evolving to require more CPU and no single chip architecture can serve every workload”.

Microsoft FY3Q26 results didn’t just mention AI, they arguably coined a phrase for the whole era: “’agentic computing era’. Management said the company is focused on helping customers “eval-max” outcomes as this new computing paradigm plays out. When Microsoft invents a new term for a computing cycle, it tends to stick.

Google’s Alphabet / Google Cloud unit posted revenue of $20.0bn in Q1 2026, up 63% YoY, supported by enterprise AI solutions and AI infrastructure. The company launched a Gemini Enterprise Agent Platform as the connective layer for building and managing agents at scale, introduced an Agentic Data Cloud for real-time enterprise data handling, committed $750mn to accelerate partner-led agentic AI development across its 120,000-member ecosystem, and unveiled new TPU hardware (TPU 8t / 8i) purpose-built for agentic workloads.

A Future Supply Chains ETF

I’ll finish with a relatively new exchange traded fund listed on the LSE which aims to capitalise across different industrial sub sectors to the profound shift in supply chains and the slow unwind of hyper globalisation, The abrdn Future Supply Chains UCITS ETF, ticker ASCH on the LSE, is an actively managed global equity ETF aiming for long‑term capital growth by investing in companies expected to benefit from the reshaping of global supply chains over the next 5+ years. The brand-new fund is Ireland‑domiciled, physically invested, accumulating, with a TER of 0.60% and c. 12–16m AUM.

The strategy is built around three “future supply chain” pillars: Technology Independence (onshoring/IP control), Resilient Supply Chains (shorter, more diversified routes and logistics), and Decarbonisation & Energy Security (infrastructure and tech enabling lower‑carbon, more secure energy and transport). The managers, Blair Couper and Jamie Mills O’Brien, construct the portfolio from a themed universe using a proprietary quantitative model to balance exposure across pillars, stocks, and liquidity, and do not use a benchmark for portfolio construction, though MSCI ACWI NR USD is used as a comparator.

Across the portfolio, many positions straddle more than one pillar but can be broadly mapped out as follows.

· Technology Independence: TSMC and SK Hynix dominate the semiconductor stack, while Keysight and Advantest (in the wider top‑5 list on another date) provide the testing and metrology that allow advanced manufacturing nodes to operate reliably.

· Resilient Supply Chains: SITC, Hyundai Heavy (and its Korea Shipbuilding affiliate in alternative top‑5 data), and various logistics players provide shipping capacity and fleet renewal to support re‑routed trade.

· Infrastructure names: businesses such as Ferrovial, Promotora y Operadora de Infrastructure, and Grupo Aeroportuario del Centro Norte build and operate airports, roads and concessions, supporting diversified transport corridors and nearshoring in Mexico.

· Decarbonisation & Energy Security: Engineering, grid and equipment players such as Sumitomo Electric and Hyundai Heavy are exposed to LNG, offshore, and power grid projects that underpin energy security transitions. Companies engaged in efficient shipping, port automation and electrification also contribute to lower‑carbon transport chains by improving fuel efficiency and reducing congestion.

As of 31 March 2026, the ETF is heavily tilted to industrials (57.6%) and information technology (22.8%), with smaller allocations to materials (4.8%), consumer discretionary (4.5%), energy (3.5%), health care (2.2%), communication services (1.6%) and consumer staples (1.4%), plus 0.7% cash. This is consistent with a supply‑chain theme, capturing logistics, infrastructure, capital goods, semis and test equipment as key enablers of global trade.

Geographically, the US is the largest country weight at 34.9%, followed by Japan (14.8%), Mexico (7.9%), South Korea (6.3%), Taiwan (6.2%), China (6.0%), the UK (5.3%), Canada (4.3%) and a residual “Other” bucket at 13.7%. Concentration risk is manageable; the top 10 holdings account for 24% of NAV, while the top 5 account for around 15% of the fund.

Top 10 holdings and role in the theme – source March fact sheet

What about fund performance? The fund is only a few months old, but the last update revealed useful performance metrics indicating outperformance (though only over a very limited period since May 2025).

Source: Aberdeen, 31 March 2026

My bottom line? As you can see from these numbers, the portfolio is highly global and EM‑tilted relative to MSCI ACWI, reflecting the need to be close to manufacturing, shipping, and energy‑infrastructure nodes. One snag is that the fund size remains small (c. £12–16m), and the manager explicitly flags concentration and EM risk, including exposure to China A‑shares. I would also note that dealing in the ETF might be a bit tricky, as it’s very new, and I couldn’t find the shares on some key platforms (HL and IBKR). As far as I understand it, if you request it on these platforms, dealers will add the shares, and the fund does have a main market listing in London.

David Stevenson

Twitter: @advinvestor

This article is for educational purposes only. It is not a recommendation to buy or sell shares or other investments. Do your own research before buying or selling any investment or seek professional financial advice.