If AI hasn’t already struck fear in the minds of the general population, humanoid robots are threatening to break the confines of the lab and enter the real world. It’s been hard to ignore those funny humanoid robot videos but it’s not all about party tricks and dance moves. Deployments are expected to surge from today’s tens of thousands to millions in the next decade, driven by broadening commercial and industrial applications.

What’s stopped full-scale adoption in the past?

Robots themselves are not new. They have been performing singular mechanical tasks like painting, welding or assembling for decades. There is no shortage of demand for real-world applications, but the biggest constraint for full-scale, multipurpose adoption remains a lack of real-world training data.

Unlike LLMs where most of the training data resides on the internet, robots need data from the physical world, including watching humans performing various mundane and complex tasks in unpredictable environments (called physical AI). However, the race is on to use real-world and synthetic data to create the data flywheel that will enable rollout at scale. AI-based simulation algorithms can now replicate physical-world training data and compress the learning cycle from years to days.

Additional constraints to humanoid robot adoption are the lack of dexterity and limited battery life that prevent robots from taking up complex and mission critical tasks – for now.

What’s changed now to drive full-scale adoption?

As per CLSA, there are four key trends beneath the surface that could be changing the paradigm when it comes to full-scale commercialization of humanoids.

1. Actuators used to convert electrical energy into kinetic energy (movement) are being standardized across the industry with energy efficient and precise electric actuators (vs. leaky hydraulic actuators).
2. A move toward modularization where the motor, reducer, inverter and controller are housed in one integrated unit which is easier to mass produce.
3. Reducers are what give robotic joints the power to lift weights. Advances in AI software are helping decrease inaccuracies in low-cost planetary reducers, therefore enhancing performance while bringing overall costs down.
4. Finally, the adoption of linear actuators which use convert rotational energy into straight-line thrust, mimicking the contraction and extension of human muscles. This helps with heavy load bearing capacity and better shock absorption on ground impact.

Where will demand come from?

• Manufacturing – particularly in automotive industries which have both the scale and assembly line processes to enable large-scale adoption. We also see auto component companies as fertile ground for the emergence of humanoid robot opportunities, given their scale manufacturing and quality control experience.
• Retail and facilities management – from tasks like customer service, cleaning and delivery-related work.
• Logistics – tasks around assembly, packaging, inspection and transportation.
• Defence and hazardous tasks – from mine clearing to dangerous material handling.

We have several holdings in the emerging market small cap portfolio that could potentially benefit from the coming robotics revolution. Sinbon Electronics Co. Ltd. (3023 TT) supplies sensors and connectors for both battery charging and signal transmission, and video capture harnesses directly to humanoid robot OEMs. Reliability and quality requirements are stringent and Sinbon is a sole supplier for at least two US humanoid clients.

Similarly, pneumatic actuators and linear motion components are among the most widely used parts in humanoid robot joint and limb assemblies and Airtac International Group (1590 TT) is an important supplier in this space. We also believe holdings such as WeRide Inc. (WRD US) and Dongguan Yiheda Automation Co. Ltd. (301029 CH) could potentially benefit from increased investment in humanoid robotics.

We believe emerging markets will be at the heart of the humanoid revolution. According to Bloomberg, China accounted for almost all global humanoid shipments so far this year. Just as how Asian supply chains deliver the picks and shovels enabling the current AI buildout, we think a similar story will play out in robotics as the current supply chain leverages its scale and manufacturing excellence to mass produce everything from actuators, reducers, sensors and cables that are needed to power this revolution.

The companies discussed are provided for illustrative purposes to demonstrate the investment team’s research into the robotics theme and are not intended as investment recommendations. There is no assurance that these companies will benefit from increased adoption of humanoid robotics or that the investment thesis will develop as anticipated.

Les guerres commerciales produisent rarement des résultats gagnants. Elles augmentent les coûts, perturbent l’investissement et injectent une incertitude inutile dans les décisions d’affaires. Le regain de confrontation entre les États-Unis et le Canada ne fait pas exception et, à court terme, il devrait peser sur la croissance canadienne. Toutefois, les répercussions à plus long terme pourraient en fait être plus constructives. Depuis des années, le Canada peine à convertir tous ses avantages, notamment ses ressources abondantes, sa stabilité politique et ses énormes réserves de capital institutionnel, en un investissement et une productivité plus solides. Les obstacles réglementaires, la lenteur des approbations de projets et la forte dépendance à l’égard du marché voisin des États-Unis ont constamment freiné l’économie. Le différend commercial actuel pourrait fournir la pression nécessaire pour susciter le changement.

Choc à court terme, catalyseur à plus long terme

Les dernières négociations ont échoué après que les modalités d’un accord potentiel eurent changé tard dans le processus. Ottawa a finalement quitté la table, soutenant que l’accord proposé affaiblirait des secteurs clés et limiterait la capacité du Canada à diversifier ses relations commerciales. L’administration Trump a ensuite annoncé des droits de douane de 50 % sur environ 20 G$ US d’exportations canadiennes, ce qui a amené le Canada à répliquer par des contre-tarifs équivalents sur des biens américains à compter du 8 septembre. L’incidence économique immédiate ne doit pas être minimisée, mais elle demeure également relativement circonscrite. Pour être clair, l’incidence totale est gérable. Environ 85 % des exportations canadiennes demeurent exemptées en vertu de l’ACEUM, tandis que les nouveaux droits de douane touchent environ 5 % des exportations totales. L’incertitude commerciale retardera probablement certains investissements et pèsera sur la croissance, mais il ne s’agit pas encore d’un désastre économique. Le Canada dispose aussi d’une capacité budgétaire plus importante que celle de bon nombre d’économies développées pour amortir le choc à court terme tout en soutenant l’investissement, et il aborde cette guerre commerciale avec une inflation relativement plus modérée.

Ce qui pourrait compter davantage, c’est la façon dont le Canada réagira. Le premier ministre Carney est entré en fonction avec un programme ambitieux visant à accélérer les projets d’infrastructure et de ressources, à réduire les obstacles à l’investissement et à attirer beaucoup plus de capitaux privés. La question persistante était de savoir si le Canada pourrait surmonter les obstacles réglementaires et politiques qui ont ralenti les grands projets par le passé.

Ce différend commercial pourrait catalyser le changement. Les nouveaux pipelines, les infrastructures de GNL, les projets de minéraux critiques et le transport d’électricité peuvent de plus en plus être présentés non seulement comme du développement économique, mais aussi comme des éléments de résilience nationale. L’élimination des barrières commerciales internes devient urgente lorsque le commerce extérieur est moins fiable. En ce sens, le différend augmente le coût de l’inaction. Un contexte politique unifié pourrait donner à Ottawa la marge de manœuvre nécessaire pour faire avancer des projets dont on discute depuis des années, mais qui se concrétisent rarement.

Il en résulte des perspectives inhabituelles à deux horizons : une croissance plus faible à court terme, mais un investissement intérieur et une productivité potentiellement plus solides au fil du temps.

Le Canada paraît différent vu de l’extérieur

Le contexte mondial des placements évolue également, alors que les flux commerciaux et de capitaux sont de plus en plus influencés par la politique. Les investisseurs accordent davantage d’attention à la stabilité institutionnelle, à l’accès aux ressources et à la fiabilité des contreparties. Dans ce contexte, les forces relatives du Canada prennent de la valeur. Des signes préliminaires indiquent que les capitaux internationaux deviennent plus réceptifs à l’égard du Canada. L’investissement direct étranger a atteint tout juste moins de 26 G$ CA au T2, en hausse par rapport à la moyenne de 21 G$ CA des quatre trimestres précédents. La demande étrangère pour les actifs financiers canadiens a également été forte. La majeure partie de ces achats s’est concentrée dans les obligations, mais les flux étrangers vers les actions canadiennes sont récemment redevenus positifs après plusieurs années de ventes persistantes (voir le graphique 1).

Graphique 1 – Les flux étrangers vers les actions canadiennes redeviennent positifs

Source : Statistique Canada

Le Canada dispose d’une occasion particulièrement opportune de susciter davantage d’intérêt. À la mi-septembre, Toronto accueillera le premier Sommet de l’investissement du Canada, conçu pour attirer des capitaux vers les entreprises et les infrastructures canadiennes. Il survient maintenant au moment où les investisseurs mondiaux réévaluent activement leur concentration géographique et leur exposition aux chaînes d’approvisionnement. Le Canada n’a pas besoin de remplacer les États-Unis comme destination mondiale des placements. Il peut simplement devenir plus attrayant à la marge.

Le capital est déjà ici

Le Canada dispose également d’une énorme source de capital intérieur. Les grandes caisses de retraite du pays gèrent collectivement environ 2,5 T$ CA, mais seulement environ le quart de leurs actifs sont actuellement investis au Canada. Toutefois, l’argument en faveur de l’investissement intérieur ne peut pas simplement reposer sur un appel à « acheter canadien ». Il exige de meilleures occasions. C’est pourquoi les réformes des politiques et le développement de projets comptent. Les infrastructures, l’énergie, la production d’électricité, les minéraux critiques et le transport sont des actifs de longue durée qui peuvent bien convenir aux investisseurs en régimes de retraite. Si le Canada peut accélérer les approbations et créer des projets plus attrayants sur le plan commercial, l’investissement intérieur pourrait augmenter parce que les occasions elles-mêmes seraient convaincantes.

Tirer le meilleur parti du moment

La confrontation commerciale demeure un frein économique à court terme, mais elle pourrait aussi mettre en évidence certaines des faiblesses structurelles dont le Canada discute depuis des années sans les corriger. C’est là que réside l’occasion. Rien ne garantit que le Canada transformera ce moment en changement durable. Les annonces de projets doivent encore se traduire par de véritables projets, et les réformes réglementaires doivent encore produire des résultats durables. Mais l’argument en faveur des placements est assurément devenu plus intéressant.

Stratégie de portefeuille

Les marchés canadiens ont fait preuve d’une résilience notable malgré l’escalade des tensions commerciales. Les actions canadiennes ont continué de bien se comporter, l’indice composé S&P/TSX surpassant l’indice S&P 500 tant depuis le début du trimestre (voir le graphique 2) que depuis le début de l’année (en monnaie locale). Parallèlement, le dollar canadien a mieux résisté que prévu depuis la reprise de la guerre commerciale, compte tenu de ce qui constituerait normalement un choc négatif important pour les perspectives intérieures. Cette résilience laisse croire que les investisseurs pourraient regarder au-delà de l’incidence immédiate sur la croissance. En même temps, les rendements obligataires ont augmenté à l’échelle mondiale, reflétant les pressions persistantes attribuables à une croissance solide du PIB nominal, aux dépenses budgétaires et, peut-être surtout, à la combinaison croissante des besoins de financement des secteurs public et privé. Aux États-Unis, la hausse des rendements est devenue suffisamment importante (voir le graphique 3) pour que le secrétaire au Trésor Bessent augmente les achats de titres du Trésor à long terme dans le cadre du programme de rachat. Les flux directs étaient faibles par rapport à la taille du marché, mais le signal était notable : les responsables du Trésor sont de plus en plus mal à l’aise devant des hausses désordonnées des rendements à long terme. Les obligations à long terme ont d’abord progressé, mais le mouvement s’est rapidement estompé, ce qui donne à penser que l’intervention des autorités pourrait ne pas éliminer la pression sous-jacente sur les taux à plus long terme. Parallèlement, les communications de plus en plus restrictives de la Fed, plus récemment de la part du président Warsh au symposium de politique économique de Jackson Hole, laissent entendre qu’une inflation persistante aux États-Unis pourrait encore forcer un resserrement de la politique monétaire.

Graphique 2 – Les actions canadiennes surpassent les autres marchés malgré la guerre commerciale

Sources : Bourse de Toronto, S&P Global, MSCI, Nasdaq, Macrobond

Graphique 3 – Les rendements des obligations du Trésor américain à 30 ans ont atteint leur plus haut niveau en près de 20 ans

Sources : Département du Trésor des États-Unis, Macrobond

Dans ce contexte, les portefeuilles équilibrés conservent une orientation globalement défensive, avec une exposition neutre aux actions. Au sein des actions, nous privilégions les actions canadiennes par rapport aux actions mondiales.

Au sein des titres à revenu fixe, la croissance canadienne plus faible et l’incertitude commerciale apportent un certain soutien tactique à la durée, mais les pressions persistantes à l’extrémité longue de la courbe commandent la prudence. Nous observons actuellement des occasions relativement attrayantes liées à l’accentuation pentification de la courbe des taux canadienne.

La stratégie des portefeuilles d’actions fondamentales demeure constructive, soutenue par des révisions positives des bénéfices et une activité économique résiliente. Nous continuons de favoriser des thèmes comme les infrastructures d’IA, les terres rares et la défense, tout en surveillant les principaux risques, notamment les pressions inflationnistes potentielles et tout ralentissement des dépenses d’investissement liées à l’IA.

The global manufacturing PMI new orders index rose in August, though remains below an April high – see chart 1.

Chart 1

The solid August result is consistent with a rise in global six-month real narrow money momentum into early 2026. Growth, however, has eased since February, suggesting a moderation in new orders over the remainder of the year.

A July fall in real money momentum reflected deepening contractions in Europe and Japan, which offset a further pick-up in the US – chart 2.

Chart 2

The suggestion is that the US economy is running hot with the Fed behind the curve, while ECB and BoJ policy tightening is misguided, risking a sharp economic slowdown, or worse.

Meanwhile, six-month growth of global real narrow money is estimated to have fallen below that of industrial output in July, implying a less favourable liquidity backdrop for markets – chart 3.

Chart 3

A previous undershoot in April preceded a sharp correction in momentum stocks but real money growth recovered to close the gap in May / June, following which equity indices reached new highs.

The global services PMI survey for August was stronger than for manufacturing, with new business reaching a 20-month high. Services buoyancy, however, has little implication for manufacturing prospects. Granger-causality tests show that manufacturing new orders predict services new business but not vice versa. Manufacturing deceleration is likely to be reflected in services cooling.

Dans une récente entrevue accordée à Soundbites d’Investment Executive, Lisa Conroy, CFA, spécialiste des produits au sein de notre équipe des actions fondamentales, présente ses perspectives pour les actions canadiennes et explique pourquoi plusieurs tendances structurelles créent des occasions intéressantes pour les entreprises canadiennes. Du rapatriement des activités de production à l’électrification, en passant par les investissements dans les infrastructures liées à l’intelligence artificielle, Lisa explique pourquoi le Canada est bien placé pour tirer parti des forces qui transforment l’économie mondiale et où notre équipe trouve des occasions sur le marché canadien.

Japan: Oasis of (political) stability

In February 2026, Japan’s first female prime minister Sanae Takaichi announced a snap election in the National Diet’s Lower House. She proceeded to win a commanding majority of roughly two-thirds of seats for her ruling conservative party, the Liberal Democratic Party (LDP). With conservative Osaka-based coalition partner, the Japan Innovation Party (JIP), the ruling coalition control about three-quarters of seats, providing a strong mandate to advance policy plans. After cycling through four prime ministers (including Sanae Takaichi) since the pandemic, Japan finally has political stability. In contrast, G7 peers like France, Germany, the UK and Italy are beset by fragile ruling coalitions or political infighting.
 

Figure 1: National Diet Lower House seats split by party before and after the snap February 2026 election

Source: Nikkei Asia

Fiscal investment, not consumption

The Takaichi administration is taking advantage of this electoral supermajority to pursue a more growth-oriented fiscal strategy with a ¥370+ trillion fiscal investment package into 17 strategic sectors through FY 2040E. Previous fiscal stimulus programs focused on consumer support and public works projects to prevent economic stagnation or to alleviate downturns and associated unemployment. However, Prime Minister Takaichi’s plan is aimed at boosting Japan’s long-term productive capacity.
 

Figure 2: Estimates of Japan’s real GDP potential growth rate from Bank of Japan and the Cabinet Office

Source: Bank of Japan, Cabinet Office via Bloomberg Economics

Figure 3: Annualized potential real GDP growth estimates across major economies

Sources: International Monetary Fund, Bank of Japan, Bank of Canada, Bank of Korea, USA Congressional Budget Office

Automation to offset a shrinking labour force

Against other advanced economies, Japan’s economic growth potential is low. This is partly due to its aging and falling population, but also because of low productivity relative to peers. The government and private firms both see automation, rather than mass immigration, as the solution to low productivity and structural labour shortages caused by an aging and declining population as well as insufficient technological adoption. By directing capital toward sectors such as automation, semiconductors, data centres, batteries and advanced healthcare, Prime Minister Takaichi’s fiscal plans seek to lift productivity, which will drive Japan’s long-term economic growth. Specifically, Prime Minister Takaichi’s fiscal roadmap includes allocations toward themes such as physical AI, advanced health care and soft power.
 

Figure 4: The Japanese economy’s capital intensity stagnated for two decades despite labour shortages

Sources: Bank of Japan, Cabinet Office, Ministry of Internal Affairs & Communications via Bloomberg

Figure 5: Prime Minister Takaichi’s fiscal stimulus plan through public-private partnerships into FY 2040E

Source: Cabinet Secretariat via Bank of America Global Research

Physical AI

Prime Minister Takaichi’s fiscal roadmap allocates ¥10.5 trillion to “physical AI.” This refers to the manifestation of Artificial Intelligence (AI) in the physical realm through robotics and automation. In Japan, labour shortages are concentrated in sectors where robots struggle to displace workers. For example, manufacturers have advanced machining tools, but too few operators to handle them. Logistics firms have sufficient trucks but too few drivers. Humanoid robotics and autonomous vehicles should eventually enable Japan to expand productivity via a capital-for-labour substitution.

In total, physical AI, semiconductors, data centre and battery investments comprise ¥101.6 trillion of the government’s 15-year ¥370 trillion fiscal investment plan. Our portfolios’ exposure to these core themes is concentrated in semiconductor production equipment makers and semiconductor material producers.

Tokyo Seimitsu Co. Ltd. (7729 JP)

Founded in 1949, Tokyo Seimitsu manufactures and sells metrology instruments and semiconductor production equipment for automotive, machine tools, semiconductors and aerospace, as well as charge and discharge testing systems for BEVs.

Micronics Japan Co. Ltd. (6871 JP)
Micronics Japan designs, produces and sells probe cards, which are used to inspect integrated circuits (ICs). The firm also manufactures wafer probers, probe card testers, IC handlers and inspection and testing devices used for liquid crystal display (LCD) manufacturing.

Horiba Ltd. (6856 JP)
Horiba manufactures and markets metrology instruments and analyzers. Key product lines include scientific/medical/emissions analyzers, environmental monitors for air/water and semiconductor testing equipment. Horiba has a local presence across China, Japan, Korea, India, Singapore, Thailand, Austria, France, Germany, the UK, the United States, Canada and Brazil.

Sumitomo Bakelite Co. Ltd. (4203 JP)
Sumitomo Bakelite Limited is an integrated processor of synthetic resins and a member of the Sumitomo Chemical Group, which retains a 10.5% equity stake. The firm’s materials are used during the production of electronic components such as chips and PCBs as well as in automotive where BEVs use more encapsulants. With an industry-leading 50% market share after acquiring Kyocera’s encapsulant business, the company is well-positioned to expand further, supported by the growth in automotive applications.

Kurita Water Industries Ltd. (6370 JP)
Kurita Water manufactures, sells and maintains water treatment equipment and facilities. It produces chemical consumables for precision cleaning and water purification. The company also manufactures equipment for wastewater treatment, purification, sanitation, soil remediation, sanitation and HVAC applications. Kurita Water remains Japan’s largest water treatment engineering firm.

Advanced health care

As Japan is a pioneer leading the world in aging, its government recognizes their firms’ “first-mover advantage” in tackling ailments. “Advanced health care” refers to pharmaceutical solutions and medical devices that improve human health. Prime Minister Takaichi’s fiscal plan allocates ¥64.1 trillion to pharmaceutical therapeutics such as antibody drug conjugates, bispecific antibodies, infectious disease vaccine R&D and AI-enabled medical device diagnostics. We have portfolio holdings that are positioned to benefit from the investments in this theme.

Asahi Intecc Co. Ltd. (7747 JP)
Asahi Intecc is a Japanese medical device manufacturer. Asahi operates through two segments. The medical segment develops, manufactures and sells private-label and OEM SKUs. The industrial devices segment develops, manufactures and sells components related to both medical and industrial products. It is the leading producer of interventional guidewires and microcatheters, perfected over 40 years with a longstanding presence in niche steel wire tech.

Sysmex Corporation (6869 JP)
Founded in 1968, Sysmex is the leader in hematology, hemostasis, invitro diagnostics, immunochemistry, urinalysis and the challenger in surgical robotics. The company designs, produces and supplies reagents, instruments, services and other products used in diagnostic tests.

Soft power

The government also understands the importance of “soft power.” This is the phenomenon of exerting geopolitical influence through cultural content such as manga, anime, music, and games. The fiscal plan outlines content investments worth ¥33.7 trillion to enable intellectual property (IP) monetization and new IP development, as well as the localization of Japanese cultural IP overseas, IP exports and tourism.

Sega Sammy Holdings Inc. (6460 JP)
Sega Sammy is the second largest gaming software and hardware producer by revenue, after Nintendo. Their entertainment content segment develops and sells games on third-party platforms (mobile, PC, consoles), licenses IP to film producers and goods manufacturers and sells equipment to arcade operators. Sega’s pachislot and pachinko machine segment manufactures and sells its products to game parlours. The resort segment operates hotels and golf courses at integrated resorts. The firm owns IP of major gaming franchises like Sonic the Hedgehog, Virtua Fighter, Yakuza and Angry Birds since 2023.

Kotobuki Spirits Co. Ltd. (2222 JP)
Kotobuki Spirits is a Japanese firm engaged in the manufacture and sale of confectioneries. It operates six segments: Sucrey, KCC, Seika Tajima, Sales Subsidiary, Kujuku Island and others. The firm is entering into retail after successfully operating via wholesalers.

The manufacturing PMI reached a 51-month high in August, according to flash data released last week. Recent strength was signalled by an upswing in six-month real narrow money momentum into July 2025 followed by a consolidation into early 2026 – see chart 1.

Chart 1

Real money momentum, however, has fallen sharply since February, turning negative in April and weakening further in July. Allowing for the usual six to 12 months lead, this suggests that the PMI is entering a time window to begin another sustained decline.

The fall in real narrow money momentum reflects a combination of a slowdown in nominal growth, probably explicable by misguided ECB policy tightening, and an energy-driven pick-up in six-month consumer price inflation – chart 2.

Chart 2

A country breakdown is available for the deposit component of narrow money but not currency in circulation. Six-month real deposit momentum is negative across the big four, with the largest contractions in France and Italy.

Chart 3

French monetary weakness suggests rising economic / fiscal risks, appreciation of which may explain a fall in demand for French government debt. Bond purchases by Eurozone banks in the 12 months to July were smaller than in the other big four markets, a reversal of the position a year ago – chart 4.

Chart 4

Further details show that purchases of French bonds by both French banks and other Eurozone institutions have slowed. French banks bought only €7.0 bn in the year to July, down from €48.9 bn in the prior 12 months.

The fall in demand for French bonds by French banks follows a reduction in the ownership percentage of French insurers and other domestic investors in recent years. Accordingly, the share of debt owned by non-residents rose to 57.5% in Q1 2026, a nine-year high – chart 5.

Chart 5

Chart 1

The pick-up is consistent with marked monetary acceleration since the start of the year, which continued last month. Six-month growth of the broad “M2+” measure calculated here reached 8.1% annualised in July, with expansion of narrow money M1A hitting 10.2% – chart 2*.

Chart 2

The rise in broad money growth appears to have been driven a combination of firmer commercial bank credit expansion, the Fed’s reserve management securities purchases and external monetary inflows, reflecting large-scale foreign buying of US equities.

Monetary strength suggests that near-term economic news will remain robust, while medium-term inflation risks (i.e. for 2028 and beyond) are rising.

Could money momentum be peaking? Three-month growth of commercial bank loans and leases has fallen since April, although the impact on overall credit expansion has been softened by a pick-up in securities purchases – chart 3. The slowdown has been focused on C&I loans and the “all other” category, which includes lending to non-bank financial institutions.

Chart 3

Additionally, the boost from Fed bill buying is moderating, with purchases suspended for the current operating period (ending 14 September) and uncertain prospects for a subsequent resumption at the previous $10 billion per month pace.

*M1A = currency in circulation + demand deposits. M2+ = M2 + large time deposits at commercial banks + institutional money funds.

Chart 1

Renewed weakness is unsurprising because the BoJ is continuing to ramp up QT, with monthly JGB purchases falling further behind the run-rate of redemptions – chart 2.

Chart 2

Recent f/x intervention will be a further drag on August numbers. (Yen purchases occurred on 30-31 July, so settled on 3-4 August.)

A post-covid fall in annual money growth accelerated from Q1 2024. This has been reflected in a slowdown in annual nominal GDP expansion since Q2 2025, to 3.1% last quarter – chart 3.

Chart 3

Nominal GDP growth may soon be at or below a pace consistent with the 2% inflation target, based on the BoJ’s estimate of potential expansion of 0.7% pa.

Annual money growth bottomed in Q2 2025, with a tepid recovery into Q2 2026 probably now reversing.

Money growth rates are far below 2010-19 means, when nominal GDP expansion averaged 1.4% pa, a pace associated with average annual CPI inflation of just 0.5%.

Optimists cite strong bank credit growth. Commercial banks’ domestic loans and discounts grew by an annual 6.7% in June, with corporate lending up by 7.8% – chart 4.

Chart 4

The “monetarist” view is that stronger lending has limited economic effects unless accompanied by faster monetary expansion. Otherwise, any rise in demand associated with the lending is balanced by weaker spending elsewhere – a monetary “crowding out” effect.

Lending buoyancy, in any case, is partly a consequence of the monetary squeeze imposed by QT, rather than being an independent positive signal. Rising yields due to the JGB dump have encouraged corporations to switch from bond market funding to cheaper bank borrowing.

Corporations may also have been borrowing domestically to finance rising FDI, contributing to downward pressure on the yen.

What would happen if QT were suspended? With the distortion of BoJ supply removed, domestic and foreign demand for JGBs would likely revive, resulting in lower yields and a rally in the yen. Money growth would recover but probably only to a moderate level, reflecting an associated slowdown in bank lending. Excessive monetary acceleration could be countered by raising rates. A stronger yen would damp near-term inflation while a recovery in money growth would reduce the risk of a medium-term undershoot.

Worth pushing for, Secretary Bessent?

Investors remained captivated by the AI capex cycle through a July selloff in tech hardware names across North Asia. A parabolic rally in tech hardware stocks controlling bottlenecks in the AI supply chains was followed by a sharp correction during the month. Pundits ascribed market jitters to concerns that hyperscaler capex may be curtailed, reflecting speculation that returns from investment in the technology might be less than compelling.

Second-quarter results from the megacap US tech companies flew in the face of these fears, with firms broadly reporting robust earnings growth in cloud computing services and hinting at healthy returns from their bets on AI. However, doubts are creeping in over whether earnings upgrades can be sustained, adding to concerns over the rising use of debt to fund the capex boom and the competitive threat posed by cheap open-source Chinese LLMs to more expensive frontier model providers in the United States.

Hyperscaler capex estimate by year (USD bn)

Source: Mizuho Securities Equity Research, July 2026.

Artificial Analysis Index – higher is better

Source: Artificial Analysis, August 2026.

Fragile liquidity backdrop

In our view, these risks were well understood before the sell-off. The unwind may instead have been driven by a deterioration in the global liquidity backdrop and an associated deleveraging in crowded trades. NS Partners Chief Economist Simon Ward flagged the risk early in the month (A “monetarist” perspective on current equity markets):

“Global six-month real money growth has fallen back since early 2026, crossing below industrial output expansion in April (see below chart). This suggests that the global economy will lose some momentum during H2, while the monetary backdrop for markets has become less favourable, at least temporarily.”

G7 + E7 industrial output & real money (% 6m)

Source: NS Partners & LSEG Datastream.

South Korean retail speculators driving parabolic rallies in AI hardware winners added to the vulnerability, with the correction appearing to coincide with systematic quarter-end repositioning. This triggered weakness in “speculation of choice” names that fed deleveraging and forced selling. According to Citi, more than 360,000 South Korean margin accounts were forced into liquidation with 62% of those individuals wiped out under the age of 35.

Leverage in South Korea

Source: Jefferies Equity Research, July 2026.

We had been trimming our AI exposure into the event, but in hindsight we should have been more aggressive ahead of what was the largest pullback in Asia momentum since 1999.

Asia Momentum (L/S) – monthly performance

Source: Bernstein Equity Research, August 2026.

Little has changed by way of the growth and profitability drivers of our tech hardware companies in Taiwan and South Korea

For example, TSMC is forecast to generate 30% EPS in 2027 with a gross profit margin of over 55% and trades on a PE of 15.8x 2027 and 12.7x 2028. Elsewhere in Taiwan a number of companies in the tech hardware supply chain continue to deliver EPS upgrades, with the recent de-rating providing some attractive entry points.

In South Korea, DRAM giant Samsung Electronics reported record revenues and operating profits for the quarter with the latter beating estimates driven by exponential growth in AI server demand. The company expects supply constraints to tighten further in 2027 despite investments in supply. The company has no debt and trades on a free cash flow yield of c.16% for 2026, with a PE of 3.1x for 2027 and 2.9x for 2028. We anticipate that the company will announce a shareholder return plan for c.50% of FCF in August including a special dividend, which should provide some support for valuations and reduce capital misallocation concerns.

Caveat: TSMC and the memory giants are very different beasts

TSMC is the monopoly in advanced semiconductors that power all of the latest technological innovations, boasting structural competitive advantages that underpin lower earnings volatility through customer lock-in. While the dominant memory companies SK Hynix, Samsung Electronics and Micron exist in an oligopolistic industry enjoying a demand supercycle powered by AI spending, they remain at the mercy of severe supply-demand swings. Although we expect this memory cycle to go on longer than most expect, nothing cures high prices like high prices.

Tech monopoly vs. memory super cycle

Source: NS Partners & Bloomberg.

We maintain a modest overweight to the AI supply chain on a view that sharply accelerating demand for compute will continue to support growth and profitability for companies dominating key tech hardware chokepoints. However, navigating the cycle successfully will require disciplined adjustments of conviction levels as fundamentals shift, and ensuring this is tightly aligned with portfolio risk.

EM performance is broadening out

In our last commentary, Taking stock of EM performance, we highlighted the extent to which EM outperformance has been dominated by the tech trade. There are signs the rally is starting to broaden out with a number of regional bull markets bubbling away, as illustrated below.

EM MTD and YTD Returns

Source: MSCI

Over the year to the end of July, 18 of the 24 EM markets above delivered positive returns, while 17 of the 24 rose in July. Latin America was the most consistent region across both periods with Colombia the standout, followed by Greece, Hungary and Poland in emerging Europe. In MENA, the GCC was weak across the board, positive momentum returned in Egypt, while returns in South Africa have been negative over both periods.

EPS growth in Taiwan and South Korea has been explosive, the trend in the rest of EM is improving

Source: HSBC Equity Research, July 2026.

EM valuations are low both in absolute terms and relative to DM

Price to Forward Earnings Ratios MSCI Indices, 12m Forward Earnings, Source: IBES

Source: NS Partners & LSEG Datastream.

Style rotation?

After years of outperformance, quality stocks within EM equities are showing signs of an upturn.

MSCI EM Style Indices relative to MSCI EM, 5y ago = 100

Source: NS Partners & LSEG Datastream.

Overvalued exchange rates are a headwind for Mexico, Brazil and Eastern Europe, with Asian currencies mostly cheap

Real broad effective exchange rates
% deviation from 5y ma, Source: BIS

Source: NS Partners & LSEG Datastream.

Emerging market equities have been resilient despite Gulf War III pressuring energy prices, yields and the dollar higher. As noted earlier in this piece, a cross-over of global real narrow money growth below industrial output growth (what we call negative excess liquidity, or less money than economies need) has historically been a negative performance indicator for the asset class. While real narrow money growth reconverged with output in May / June, this has not yet reversed the April cross-over.

Global real money growth on a par with output growth

G7 + E7 industrial output & real narrow money (% 6m)

Source: NS Partners & LSEG Datastream.

Strength amid these headwinds could reflect the combination of earnings upgrades, valuations and cheap currencies, and may be a signal of good things to come if the United States, Iran and Israel can broker a peace in the coming months.

GDP (gross value added) rose by 1.1%, or 2.2% annualised, in the six months to June, having shown no growth in the prior six months (i.e. between June and December 2025).

Energy prices spiked at the end of Q1 with the increase only now feeding through to household tariffs, so the claim of resilience is premature, even ignoring monetary considerations.

Both the economic stagnation of H2 2025 and the H1 2026 pick-up were signalled by money trends. Six-month rates of change of real narrow and broad money turned negative in spring 2025 but rebounded into early this year – see chart 1.

Chart 1

Momentum has since softened again, with real narrow money returning to contraction. Trends are not yet as weak as a year ago but still suggest a significant H2 economic slowdown.

As an aside, the latest Monetary Policy Report included an analysis of broad money developments, focusing on whether the current stock is out of line with its “equilibrium” level, implying future changes to spending / prices to restore balance. The conclusion was that any “money gap” is small, in contrast to 2022, when the analysis would have suggested significant inflationary potential.

While any discussion of money in the MPR is welcome, the view here is that a focus on uncertain estimates of stock disequilibrium risks neglecting the information content of changes in money growth for future activity and inflation. Monetary acceleration / deceleration carries a message for policy even in the absence of an underlying stock imbalance.