top of page
Roman Cassini

Roman Cassini

Head of ESG

  • We are often asked about our investment approach to China, and in particular the risk to the portfolio in the event of a conflict over Taiwan

  • The Hosking Partners’ portfolio’s underweight to China reflects a cautious approach to geopolitical uncertainty in the region combined with a distorted supply-side market environment

  • This article examines these factors in more detail and lifts the lid on how we think about this complex and fascinating topic

“A crisis is an opportunity riding a dangerous wind.”

Chinese proverb


The China-Taiwan-US relationship could produce a range of scenarios up to and including nuclear war. Much like the phenomenon in classical mechanics from which this note takes its title – whereby in a three-body system, any miniscule change in initial conditions can result in wildly different outcomes – the deep complexities of this three-way strategic impasse resist simplistic forecasting (1). Each movement by one actor prompts a kaleidoscopic range of responses by the other two. Expertise in one of the three promises none in the others. And viewing the issue through a singular lens of economics, military strategy, geopolitics – or any other isolated discipline – does little to confer an actionable advantage in predicting the future course of events in the region. This is demonstrated by the remarkable lack of consensus – whether within public or private spheres – about the exact nature or magnitude of the threat posed by this critical geostrategic relationship.

At Hosking Partners we are generalists, and we make no claims to specialised expertise or in-depth insight. Fittingly, our conclusion is one of uncertainty, and our portfolio’s approach to the region reflects that uncertainty. In this article, we will make some simple observations about the geopolitical situation as we see it, as well as comment on several other investment concerns which contribute to our underweight to China (see Figure 1)(2). In doing so, we hope readers gain a window not only into how we think about China from an investment point of view, but also how the consideration of geopolitical risk ties into our capital cycle approach.

A realist perspective makes war seem inevitable

Commentators tend to ascribe everything in geopolitics to idealism or everything to realism, and never the twain shall meet. Realists argue that international relations are driven by the cold logic of competition. Idealists, on the other hand, view international relations as an extension of domestic ideology and politics. The Taiwan situation has its roots in both. A 2015 Harvard Kennedy School study identified 16 historical examples of rising powers challenging established powers, and 12 of these led to war (this is what is sometimes called the ‘Thucydides Trap’)(3). A theme in most of the scenarios ending in war is that competition – for resources, land, military power – was combined with ideological difference. Conversely, where war was avoided, the two powers tended to share similar ideologies, for example Spain-Portugal in the 15th century and the US-UK in the early 20th. In the case of rising China versus the dominant US, realist competition appears to be combined with distinct ideological disagreement. And while Taiwan is only one of several possible flashpoints, history tells us that war – in some form – is therefore probable. But is history right?

Taiwan is just one thread in the grand tug-of-war between the US and China. Its importance is nonetheless significant. Geographically, Taiwan forms the central link in the chain of islands that stretches from the southern tips of Japan and South Korea, along the outer edge of the East China Sea, across the Bashi Channel and south to the Philippines (see Figure 2). This chain of US allies guards China’s maritime access to the Pacific. Alongside the Malacca Strait between the Malay Peninsular and Sumatra, these key naval channels could cut off China’s maritime trade routes relatively quickly if blocked by a hostile navy. As we wrote in ‘The Gambler’, a key consideration in both Xi Jinping’s approach to Russia and his build out of EVs is reducing China’s reliance on imported seaborne oil, thus minimising this geographical vulnerability (4). Reunifying Taiwan, however, would be the real geostrategic gem. Taking control of the major Kaohsiung naval base which sits at the southern tip of Taiwan would break the US-controlled island chain in half. This would not only guarantee China reliable passage to the Pacific, but also provide it with the ability to turn the tables and block US access to the South China Sea. Without that reliable military and trade access, the US’ ability to compete with China as a regional power in South-East Asia would suffer a critical blow.

Taiwan is also central to the battle for control of semiconductor technology. The small island has become a specialist in chip manufacturing. The flagship Taiwan Semiconductor Manufacturing Company (TSMC) is one of just three companies able to produce the world’s highest-end chips, thriving as the industry rationalised and weaker competitors fell away (see Figure 3, next page). TSMC now produces around 60% of the world’s semiconductors, and almost 90% of the most advanced types. Both China and the US regard production and supply of semiconductors as a critical strategic aim. China is some distance behind, and – as Chris Miller writes in his excellent book Chip War – is “staggeringly reliant on foreign products”. Most of these products are made by the US or a close ally (5). In response, China has put the construction of a domestic chip industry at the heart of its industrial policy (6). Success would not only provide massive support to cutting-edge Chinese military technology – including AI – but would also put a significant dent in US export revenues. China spends in the region of $300 billion annually importing semiconductors, much of which flows to the US and its close allies South Korea and Taiwan (7). However, thus far, China has proven poor at replicating the most advanced, precision manufacturing techniques of its adversaries. Again, as with the geographic problem described in the previous paragraph, reunifying Taiwan and taking control of its chip industry would offer a convenient shortcut if it could be achieved in the right way.

These powerful realist incentives to seize Taiwan have been countered in a similarly realist manner – with hard deterrence. The US has almost 100,000 troops stationed in the region, alongside the 50-60 vessels that make up its 7th Fleet. These numbers will increase significantly in coming years. Meanwhile, close ally Japan has recently committed to spend 2% of GDP on defence, a level unheard of since WW2 and which will make Japan the third largest defence spender in the world. Elsewhere, the AUKUS pact between the US, UK, and Australia has been designed to increase the number of nuclear-capable submarines operating in South-East Asia. These measures have been implemented – to a substantial degree – with the aim of deterring the Chinese threat of invading Taiwan. More broadly, they are to respond to China’s recent build-out of military power. The Chinese armed forces now incorporate the world’s largest standing army and have been adding to their naval fleet – in pure tonnage – the equivalent of an entire Royal Navy every four years. Amassing military assets does not necessarily mean they will be used, especially in the age of nuclear deterrence. However, Russia’s invasion of Ukraine is a reminder that the very existence of military mass does assert a form of gravitational pull towards conflict, particularly when wielded by an autocratic state with few formal checks or balances. If you build for war, war might well come knocking.

The idealist picture is more complicated, but offers hope

This is not just a story about autocracy versus democracy. Although it is tempting to paint it in those terms, their simplicity obscures a tapestry of competing and overlapping ideologies. Clearly, there is a fundamental disagreement about how societies – nationally and internationally – should be organised, and the values which those societies should promote. These differences have their routes in history, reaching back far beyond the 20th century tussle between capitalism and communism to Aristotle and Plato in the West and Confucius in the East. If each of the US and China is pursuing a deeply different vision of global society, surely the ideological basis for conflict could not be stronger? Francis Fukuyama has pointed out that neither China nor the US are very good examples of either autocracy or democracy, respectively, and in turn neither system is successfully upholding the underlying values it claims to protect (8). The degradation of US democracy may be more visible – particularly here in the West – but China is also exhibiting all the flaws inherent in an autocratic approach. Opposing ideological systems that happen to be broken probably make top-down miscalculation – and so war – more likely. But broken systems also produce bottom-up effects that may pull them away from conflict.

In China, the strategic importance of the one-child generation should not be overlooked. This generation, now anywhere between 10 and 40 years old, may hold the keys to China’s future. Keyu Jin describes how many were raised under conditions of intense competitive pressure in an education system based on rote-learning and frequent examination. Despite a graduate population highly qualified on paper, high-tech roles nevertheless go unfilled; Jin reports that as of late 2022, there are as many as 300,000 vacancies in China’s semiconductor industry, even as many Master’s and PhD holders are working manual jobs (9). A remarkable 20% of 16 to 24-year-olds are unemployed. This uncomfortable situation is a popular topic on social media in China. Recently, an obscure short story about a scholar-turned-beggar became a viral representation of the generation’s plight. Written in 1918, the story’s main character – Kong Yiji – rejects much of Chinese society and longs for a kinder world. He is mocked for his views. What do the parts of a generation that identify with such a character make of China-Taiwan? It is difficult to know – censuses do not exist in China – but Jin writes that the anecdotal evidence suggests a general aversion to outright war, even among those who believe Taiwan is rightfully a part of China (the dominant view). This seems to have increased following Russia’s disastrous experience in Ukraine. If true, this is significant. China may not be a democracy, but the one-child generation is large and accordingly wields political influence. This was clearly demonstrated during the 2022 protests which led to an abrupt U-turn on the zero-Covid lockdown policy – which were led in a large part by overqualified manual workers. Ignoring this critical segment of Chinese society’s views and ideas – which are born of the failures of Chinese autocratic governance – would be at President Xi’s peril.

In the Chinese Communist Party (CCP) itself, the structural challenge of guaranteeing lifetime rule may limit President Xi’s institutional agency. Autocracy is often heralded as having the advantage of facilitating unilateral decision-making, which enables the implementation of drastic policy. Putin’s actions in Ukraine have recently served as an example of this dynamic. However, the extreme concentration of power can also have the counter-intuitive effect of limiting meaningful policy action, as China expert Dr Victor Shih writes (10). In China, as more power is concentrated around the President, the top elites – who are responsible for implementing Party policy – are disincentivised from taking on operational risk for fear of reprisals. As such, they tend towards carrying out tasks on paper only, thus limiting real institutional development. This was the case in the late-Mao era, and increasingly it is emerging again today. Maintaining this “coalition of the weak” serves President Xi’s purposes by reducing direct internal threats to his nominal authority, but at the cost of reducing the real-world utility he derives from that authority. Shih points out that the way power is exercised is very different in Beijing to Moscow, with far more bureaucracy and less decision-making devolved to subordinate actors as in the ‘capo di tutti capi’ (boss of bosses) model that Putin appears to have borrowed from the mafia. As such, the view that China’s drift towards one-man autocracy heightens the likelihood of the invasion of Taiwan may be oversimplistic. Not only does the will of the people still matter, but the structural and bureaucratic reality of maintaining a one-man state may obstruct rather than lubricate the path to war by reducing structural risk appetite. Of course, although this dynamic may complicate top-down policy implementation, it can also raise the risk of bottom-up miscalculation as relatively junior powerbrokers – such as Chinese military officers – jockey for favour. Recent near-collision encounters between American and Chinese warships may be an example of this sort of behaviour.

Socioeconomic factors – arising in part from China’s broken autocracy – may also act as a counterbalance to war. Demographic aging and a slowing economy mean deflation is real risk in China. The GDP deflator – which measures growth in real terms – has recently turned negative (see Figure 4). Local government debt servicing has surpassed RMB 1 trillion per month, and nationally it represents over 100% of local budgetary income and growing. Employment figures – as described in the previous paragraph – are poor and both housing sales and exports are weak. Unsurprisingly stimulus is on the policy agenda, but here the CCP’s options are limited. President Xi has long been wary of outright QE, which risks capital flight. But his long-favoured fiscal alternatives of infrastructure spending and property investment are constrained by high debt and low revenue growth at the local level, where many regional entities and state-operated enterprises are already in a balance sheet recession. Tellingly, China’s credit impulse (the amount of new credit issued as a percentage of GDP) is starting to decline, suggesting monetary conditions are tightening despite policy noise to the contrary. And all of this gets worse if US-China relations collapse, which will only further impact exports and employment figures. As such, in the near-term, the Chinese government has relatively little control over its own economic fate, and so a stable global macro-outlook and relatively constructive trade relationships with the West remain in its interests despite the hostile rhetoric. This restricts China’s optionality around Taiwan in the near to medium term and incentivises maintenance of the status quo.

So although cold realism makes war seem like a logical outcome, party politics, idealism and economics are pushing back. It is unlikely that many in China want a ‘hot’ – i.e. conventional military – war with Taiwan any time soon, and even fewer with the US. And interestingly, only around 5% of Taiwanese citizens support independence “as soon as possible” (11). A large – and still rising – majority favour maintaining the current status quo. President Xi must also recognise the significant downsides to an outright invasion. Tactically, amphibious landings are a military nightmare, while strategically Russia has demonstrated the folly of poorly thought through adventurism. However, as author Fred Kagan has written, “compellance” – war – is only one route to reunification, and it will be the last that is attempted because the costs are the highest (12). He identifies two others, “forceful persuasion” and “coercion” which use a combination of hard and soft means to achieve their effect. Xi has publicly staked his legacy on “resolving the Taiwan question” (13). So while in the near-term war seems unlikely, we should nevertheless expect tensions with the US to tighten further, especially if domestic economic conditions ease. And as they do, the collective will to avoid full-scale war may increase the range and scope of the means that are employed short of that end. Similarly, the scope for miscalculation rises. We do not know what this will look like. The range of outcomes is broad. But we should expect financial measures – already proven ‘in battle’ against Russia – to be implemented far quicker, and more aggressively by both sides. This means there is an uncomfortable trade-off between the potential for a near-term economic recovery – which could be positive for Chinese equities – and the longer-term implications that recovery infers for the CCP’s optionality in Taiwan. Closely monitoring this situation as it develops will be critical to determining the level of exposure we retain.

It is not only about geopolitical risk

Uncertainty around the Taiwan question forms an important part of Hosking Partners’ thinking on China – but another reason for caution is rooted in our capital cycle approach. To recap, the basic logic of this approach is that observing the flow of capital into and out of industries (and geographies) provides an insight into their behavioural and competitive dynamics. These dynamics – the addition or removal of capacity, the rationalisation or expansion of competition, the spending or withholding of investment – influence future returns on capital in a directionally predictable way. And over the long run, it is returns on capital that primarily determine the direction of share prices. Figure 5 (next page) provides a graphical depiction of this process.

Through the lens of the capital cycle, large parts of the Chinese market remain unattractive. The flow of capital into certain Chinese industries – and its relationship with forthcoming supply – is difficult to observe. Company filings are opaque, national-level data is often redacted or edited, and third-party audit is of consistently poor quality (14). A relatively well-known example of this is data for Chinese coal production – which is of enormous international significance due to its pivotal role in global supply chains – but is extremely unreliable (15). Nevertheless, the advantage of considering supply rather than demand is that being directionally correct is valuable in itself even if granular exactitude is lacking. And directionally, the supply picture – particularly in the extractive industries that have driven Chinese growth – is one of state-driven overcapacity. Ultra-subsidised coal power and questionable employment practices buy China an advantageous position on global cost-curves, and thus domination of upstream and midstream market shares across a range of commodities. The practice of running these industries for share rather than returns shows no signs of abating, and in fact the CCP has been consolidating its influence via a series of reforms that blur the line between private companies and SOEs. Firms are required to establish formal CCP committees, the activities of which are opaque and are carried out with no shareholder oversight (16). Meanwhile, industrial subsidies in China are worth an astonishing $250-400 billion per year (17). This equates to almost 2% of Chinese GDP and a level which is three times higher than in South Korea, which is itself the world’s second most subsidised industrial economy. The web these pay-outs have spun has incentivised an ever more complex network of joint ventures (‘JVs’) which – combined with direct CCP involvement – creates a hugely significant additional driver (or otherwise) of flows of capital. The capital cycle approach tells us that – all else equal – if the CCP continues to flood these industries with capital while concurrently distorting free competitive behaviour, then future returns will remain compressed.

Where Hosking Partners does retain direct exposure to Chinese companies is mostly in the online retail sector. Our ownership of several Chinese internet companies is predicated on the view that they have deeply entrenched market positions with high consumer mindshare, allowing them to sustain high returns on capital. The marketplaces of Taobao and Tmall, the super-app WeChat, and’s extensive logistics infrastructure act as a toll road on the rising living standards of the Chinese population, which remains the key priority of the CCP. In recent years, we underestimated the impact of government regulations and the rising competitive intensity within the industry, but see early signs that these headwinds are abating. The $1 billion fine imposed on Ant Group has been signalled as the concluding step in the regulatory crackdown. At the same time, management teams are placing a greater emphasis on higher quality growth and cost efficiencies, which in most cases has led to improved profitability. Lastly, we have been positively surprised by the magnitude of shareholder returns. In 2022, Alibaba reduced its share count by 4% and Tencent returned >10% of its market capitalization, including the distributions of and Meituan shares. Many investors believed these companies would never be allowed to return meaningful amounts of capital to Western shareholders, but instead there are signs that the higher level of returns will continue. In contrast to the extractive industries – where government meddling is leading to over-supply – the capital cycle rationale for investment in this part of the Chinese market seems persuasive.

Hosking Partners also retains indirect exposure to the Chinese economy in several important ways. As discussed above, cracks are emerging in the Chinese economy – particularly in real estate – and macroeconomic concerns centred on population decline are growing in prominence. But we would nevertheless hesitate to question its long-term potential. Even if the unprecedented growth rates of the early 2000s and 2010s are behind us, China will likely remain a pivotal player in global economics and politics for the duration of the 21st century and beyond. As such, while not directly expressed via an overweight to Chinese firms due the risks described in this article, the Hosking Partners portfolio remains indirectly exposed to China in several important ways. The Chinese market remains a key driver of both supply and demand of energy and materials, which as a theme constitute around 25% of the Hosking Partners portfolio. We have written about this previously, in our article ‘A Diverse World’ (18). Meanwhile, the capital cycle rationales for investing in many of the companies we own elsewhere in South-East Asia – and we are overweight Hong Kong, Malaysia, the Philippines, Thailand, and South Korea as markets – are often connected to China. These interrelationships are not well-captured by simple attribution metrics. But they form a common element in our global generalist approach, which uses the lens of the capital cycle to ‘look through’ sectoral and geographic differences to capture the opportunities concealed behind the headlines. Clearly, in a situation of full-scale war between the US and China, all bets are off. But indirect exposure to the Chinese economy – of the type described above – should prove resilient to a much wider range of sub-threshold scenarios.


The strategic uncertainty implicit in the Taiwan problem – combined with the misallocation of capital identified by our capital cycle approach in large parts of the Chinese market – leads to our underweight position. It seems to us that – for now – it remains in the interests of all three parties to continue the current status quo of mutual strategic uncertainty. Nevertheless, a war is conceivable, and the scope for miscalculation is high. In recent years, the checks and balances that serve to dampen such disturbances have been weakened by antagonistic diplomacy on the part of populist, protectionist Westerners and their increasingly ideological Chinese counterparts. This dynamic means that there is a lot that could happen that isn’t war, but which still carries risk for Western investors. And the West’s inability to prevent the invasion of Ukraine means that many of these measures – such as sanctions and the seizure of assets – will be deployed much faster next time, ahead of rather than in response to military conflict. This is a sliding scale that we are already riding on. Tightening capital controls, tit-for-tat trade restrictions, and import bans are already being deployed. And the way the CCP allows foreigners to invest in Chinese shares seems designed to facilitate their easy confiscation (19). The risk of capital loss for foreign investors in China is undoubtedly material, and although in the near-term attractive valuations in certain parts of the market give us the confidence to hold a small selection of names, we feel that overall the risk-reward equation remains skewed to the downside.

At Hosking Partners we abide by the aphorism that it is better to be roughly right than precisely wrong. In China, geopolitical risk and lack of transparency mean the risk of being ‘precisely wrong’ remains uncomfortably high, especially as we continue to reflect on our experience in Russia. However, our unconstrained, generalist approach means that whether it is via our limited investments in Chinese online retail, portfolio focus on energy and materials, exposure to South-East Asia more broadly, or keen interest in the capital cycle in semiconductors, we are hopeful that our holistic approach to China and the surrounding region strikes a balance that is ‘roughly right’. As ever, we will continue to interrogate this thesis – including via a research trip to the region later in the year – for the long-term benefit of our clients.

(1) We should note that by coincidence rather than design The Three-Body Problem is also a rather excellent Chinese sci-fi series!

(2) As of 30th July, the HGF has a 1.7% exposure to mainland China versus 3.1% in the Index. Including Taiwan and Hong Kong, our underweight is -0.3%.

(3) See 


(5) Chris Miller, Chip War: The Fight for the World's Most Critical Technology, Scribner 2022, p 245. 

(6) As articulated in its ‘Made in China 2025’ policy.

(7) See, for example, 

(8) See, for example, the ‘Leading’ podcast interview with Francis Fukuyama, July 2023.

(9) Keyu Jin, The New China Playbook: Beyond Socialism and Capitalism, May 2023.

(10) Victor Shih, Coalitions of the Weak: Elite Politics in China from Mao’s Stratagem to the Rise of Xi, Cambridge University Press, 2022.

(11) See here: 

(12) Dan Blumenthal and Frederick W. Kagan, China’s Three Roads to Controlling Taiwan, American Enterprise Institute, March 2023

(13) See the speech made by Xi Jinping to celebrate the 100th anniversary of the CCP, July 2021.

(14) See, for example, 

(15) Delu Wang et al., ‘Are the official national data credible? Empirical evidence from statistics quality evaluation of China's coal and its downstream industries’, Energy Economics, October 2022

(16) As laid out in Article 19 of ‘Company Law of the People’s Republic of China’ (2019).

(17) See, for example, 


(19) See, for example,

1 August 2023

Three-Body Problem

China and the issue of uncertainty

bottom of page