Criteria: 

  • Market cap > $2bn: avoid illiquid small caps and stick to investable names.
  • Regions : North America, Europe, Asia, Oceania: focus on deeper, more liquid and better-covered markets.
  • Growth > 7/10: favour dynamic revenue/earnings-per-share and FCF trajectories.
  • Valuation > 7/10: favour cheaply priced companies without sacrificing the quality of fundamentals.
  • Quality > 7/10: select solid companies (returns, profitability, financial health).
  • Profitability > 7/10: keep a foundation of margins and value creation.
  • EPS revisions over four months > 7/10 and over one year > 7/10: look for an improving consensus, not growth seen only in the rear-view mirror.

SK Hynix

SK Hynix makes DRAM and NAND memory, essential for servers, PCs and smartphones, and is very well positioned in advanced memory (HBM, DDR5) for data centres. 


The South Korean group has exposure heavily geared to major tech markets (63% United States, 23% China). Between 2025 and 2027, the company steps up a gear: revenue rises from KRW 97,147bn to KRW 234,823bn, with a spectacular acceleration in 2026. The driver behind this surge? Demand for HBM (High-Bandwidth Memory) for AI. SK Hynix holds 57% of the HBM market and supplies Nvidia, Google and AWS. Memory is a segment where demand tied to digital infrastructure and data-intensive uses can trigger rapid expansion phases.


The most interesting signal comes from investment. Capex rises in absolute terms, from KRW 25,000bn in 2025 to KRW 39,000bn in 2027, but its weight in sales eases sharply: capex/revenue falls from 25.86% in 2025 to around 17% in 2026-2027. In plain terms, SK Hynix is growing revenue faster than investment. By comparison, Samsung Electronics shows lower intensity, with capex/revenue sliding from 15.07% in 2025 to 12.61% in 2027: investment therefore weighs less on Samsung's profitability equation than on SK Hynix's.


Finally, the balance sheet and cash generation confirm a very favourable scenario. Net debt/EBITDA improves from -0.15x to -0.96x, a sign that net cash (more cash than debt) is strengthening and provides a real cushion in the event of a downturn. Above all, FCF accelerates sharply: it rises from around KRW 21,361bn in 2025 to KRW 97,000bn in 2027. Relative to sales, that implies an FCF margin (FCF/revenue) improving from around 22% in 2025 to 42% in 2027 - in other words, generating nearly KRW 22 and then KRW 42 of free cash for every KRW 100 of revenue. This firepower is such that the investment effort becomes almost negligible: the capex/FCF ratio collapses to 0.04% in 2027, a sign the company no longer needs to reinvest more than a tiny fraction of its profits to maintain its dominance. SK Hynix is therefore portrayed as a cash machine. 

H. Lundbeck

A change of scenery with the Danish drugmaker. Lundbeck specialises in neuroscience, developing and marketing prescription medicines for brain diseases such as Alzheimer's, bipolar disorder and depression.

It generates most of its revenue in the United States (51%), ahead of Europe (23%) and international operations (24%). Over 2025-2027, the business grows at a measured pace: revenue rises from around DKK 24.5bn to DKK 25.4bn, driven by two blockbusters: Rexulti for agitation linked to Alzheimer's and Vyepti for migraines. This is a name where the challenge is not hypergrowth, but the ability to turn modest growth into cash generation and a stronger balance sheet.

On that front, the numbers are reassuring. Lundbeck's net debt/EBITDA drops from 1.10x to 0.06x between 2025 and 2027: the company moves from moderate leverage to a near debt-free position. Financing risk recedes sharply, giving Lundbeck breathing room to fund its pipeline - notably its two Phase III molecules, Amlenetug and Bexicaserin, which are very promising - pay dividends or pursue acquisitions. For comparison, US peer McKesson stays at very low leverage, close to zero over the period (a slight peak around 0.3x in 2026). At the same time, Lundbeck converts sales well into cash: an FCF/revenue ratio around 24% from 2026 means a significant share of revenue turns into available cash.

This conversion is helped by limited capex (capex/revenue around 2.3%), meaning little industrial investment to fund. The watchpoint is relative valuation: very attractive in 2025-2026 (0.4x and 0.8x), it climbs to 6.97x in 2027, implying that by then the stock becomes markedly less "GARP” if profit growth does not keep pace. 

Perseus Mining

Finally, Perseus Mining, an Australian gold producer operating in Ghana, Côte d'Ivoire and Tanzania, shows a clear ramp-up over 2025-2027: revenue rises from AUD 1.9bn to AUD 2.7bn. The increase is most visible in 2027, driven by two major projects: the Nyanzaga project in Tanzania, where first production is expected by 2027, with a target of 200,000 oz/year. And the CMA Underground project at Yaouré, where production is also expected around 2027. This more comfortable base helps absorb sector costs and secure internal funding capacity.

The most interesting element here is the investment cycle. Capex rises sharply in 2026, with the capex/revenue ratio jumping to 31.7%, from 16.6% in 2025, then falling back to 20.6% in 2027. This capex increase is mechanically reflected in cash generation: FCF/revenue drops to 7.3% in 2026 after 26.4% in 2025, before rebounding to 17.4% in 2027. In 2026, Perseus clearly prioritises investment over available cash: the higher capex/revenue mechanically pushes FCF/revenue lower. The rebound seen in 2027 suggests these outlays are starting to deliver (or, at a minimum, that investment pressure is easing), allowing cash generation to recover.

Finally, the balance sheet remains a relatively strong point: net debt/EBITDA is negative throughout the period (from -1.01x to -0.77x), meaning a net cash position despite the investment peak in 2026 (USD 755m of cash at end-2025). By comparison, Northern Star Resources invests even more heavily (capex/revenue around 36-39%) for similar growth, but Perseus retains a far more protective liquidity buffer for shareholders.