9 July 2024

The Hosking Partners strategy is starting to show strong relative performance of just under 5% relative to its benchmark over the last twelve months, despite the most recent quarter’s falling behind by 0.6%. This quarter was a curate’s egg – good in parts, although the overall relative result was negative. A strong start and finish notched up net relative performance of c. 3.2%, (1) but this was more than undone by a four-week period which saw a c. -3.9% (2) net return for the strategy.

Not for the first time, our commentary on the quarter starts with the paradox that the MSCI ACWI benchmark - which is intended to be a proxy for the entire market - is more concentrated (at least in its largest positions) than Hosking Partners’ bottom-up strategy of individual stocks with its active share of c. 84%. Dealing first with the strategy absentee Apple, which had a 3.8% benchmark weight and cost us 0.7% of relative performance during the quarter, weary readers will note that during the last four quarters Apple has made a positive contribution twice and a negative contribution twice, delivering a total of 0.5% of positive relative performance for the strategy over the twelve month period, suggesting the short-term volatility is an acceptable price to pay for not owning such a highly valued share.

Next up is Nvidia, which had a c. 0.5% weight in the Hosking Partners strategy but a 3.5% weight in the benchmark and cost the strategy 0.9% following strong results that demonstrated its dominance of the artificial intelligence (AI) space. However, taking a broader basket of AI-enablers consisting of Nvidia as well as Broadcom, Micron and SK Hynix, the weighting in the strategy was around 4% during the quarter versus 4.6% in the benchmark, narrowing the relative cost to 0.6%. Broadcom has been a long-standing constituent in the strategy, the essence of the investment thesis being the outstanding capital allocation ability of chief executive Hock Tan, and we have been holders of the memory stocks SK Hynix and Micron based on a capital-cycle analysis that increasing capital requirements would induce industry consolidation. We own Nvidia because its combination of both hardware and software (its Cuda platform) make it likely that it will take the lion’s share of the profit pool from parallel processing, of which generative AI is just one application. We retain a sceptical position on the question of the size and duration of AI demand until use cases become clearer and killer apps truly emerge, but our pack of AI-enablers have strong competitive positions and technological advantages which continue to extend their lead ahead of competitors. In the meantime, we have been taking gradual profits in all four holdings, and the basket maintains our exposure to the emerging AI theme.

Two other top ten contributors to performance that can be found towards the crest of the capital cycle curve were 3i and Costco. Via its ownership of European hard discount retailer Action, 3i is very much Costco’s disciple in creating a business based on scale economies shared, resulting in an expanding moat to protect returns which continue to grow.

A recent paper from Empirical Research Partners shows that historically, when the US market’s returns have been concentrated in a few leaders, the following year has generally seen outsized returns to the value factor, at the expense of growth and momentum. Perhaps in anticipation of this, two of the three largest contributors during the quarter were shipping company Hafnia and diversified miner Anglo American, exemplars of the value type, and both trading at a discount to their replacement value. Anglo American’s strong contribution was the result of a bid approach by rival miner BHP, and while we are not holders of BHP, it is encouraging that a leading industry participant clearly sees the merit of a “buy rather than build” strategy, suggesting that disciplined capital allocation will continue to support the generation of steady returns on capital by companies in this sector to which we have broad exposure.

Other than the big tech underweights, the largest detractors from performance during the quarter were event-related rather than thematic. The landslide election victory of AMLO’s successor Claudia Sheinbaum and the weakening Mexican peso caused cement maker Cemex to fall 29% in US dollar terms, costing the strategy 0.2% of relative performance. Japanese paper company Hokuetsu started the month with its price inflated by the effects of a short squeeze in March (when we managed to take profits) and derated during the quarter as activist attempts to replace management failed to gain traction for the time being. Toyota Industries reported further certification compliance irregularities (we had met with management in Tokyo in February, so this was a disappointment), and the subsequent announcement of a buyback was insufficient to prevent the stock’s 19% US dollar underperformance during the quarter at a cost of 0.1% to the strategy.

During the quarter the Japanese yen fell from ¥151 to ¥161 against the US dollar, contributing to an overall 0.5% hit to performance from our Japanese names, despite strong contributions from Hitachi (an exporter with exposure to AI and grid transformation) and activist situations Dai Nippon Printing and Suncorporation. The yen weakness was despite US$60bn spent in April by the Ministry of Finance in an attempt to prop up the currency. Partly the weakness is a function of higher for longer rates in the US, partly it is due to the reluctance of the Bank of Japan to bring an end to ultra-low interest rates. During the last twelve months Japan has contributed a positive 0.5% to the strategy’s performance, and in the meantime the diversified nature of our 54-stock Japanese basket means we have exposure to both export champions as well as domestic-focused companies, mitigating the impact of currency. Our Japan weight during the quarter was 12.3% compared with its 5.3% weight in the benchmark.

Turnover during the quarter was 9%, elevated due to the introduction of Omar Malik to the multi-counsellor line-up and associated changes to fund his portfolio. We are very excited about Omar’s elevation and the differentiated contribution his sleeve will make to the strategy. Among purchases at the strategy level, financial services featured most strongly, both banks and insurers but each transaction stock-specific rather than part of a broader thematic programme. We added to our PGM (platinum group metals) exposure while rotating out of Anglo Platinum in anticipation of demerger flowback. A position was initiated in Argentina’s Cresud, a farmland operator and developer that is also the controlling shareholder in one of the country’s leading property companies. A sale of Kogan, Australia’s Amazon challenger, was recycled into e-commerce peers Asos in the UK and Alibaba in China. Sales during the quarter consisted predominantly of taking profits in shipping as order books begin to grow and in semi-conductors.

As the Northern hemisphere’s sun has just passed its zenith and we look to the second half of the year, it feels like we are in the middle of election season, with polls in France and the UK following earlier contests in India, Mexico and South Africa, and the big one yet to come in the US in November. Markets shrugged off Modi’s less than emphatic victory in India, stumbled with Sheinbaum’s landslide and were encouraged by the entry of the ANC into coalition with the DA. They appear to be less sanguine about the result in France compared with its cross-channel neighbour Britain. Barring a last-minute substitution in the US, Trump’s victory appears increasingly certain, but his policies and the market’s reaction frustratingly unclear. Will a hardening of the Biden administration’s protectionist policies result in inflation, or will China’s economic weakness cause deflation to be exported globally?

Meanwhile the Hosking Partners strategy not only benefits from the diversification offered by its c. 350 (3) names, but more than holds its own despite a c. 21.5% underweight to the US relative to the benchmark’s 63.9% weight, and an IT underweight of c. 15.4%. Even as the strategy has been able to keep up in the face of such headwinds today, the potential appears extremely exciting for when eventually the market broadens, dollar strength weakens, value is re-rated and money flows elsewhere from the FOMO frenzy which is the AI phenomenon. Whether this will happen simply as mean reversion kicks in with the passage of time, or is precipitated by some unanticipated event (hello, geopolitics) remains to be seen, but our unconstrained and contrarian strategy offers a wide selection of idiosyncratic and uncorrelated ideas whatever the conditions.