Aktien & Multi Asset Ausblick - Q3 2023

Die bisherigen Gespräche mit japanischen Unternehmen im Jahr 2023 haben gezeigt: Das Bewusstsein und auch die Investitionen in KI – besonders in generative KI – sind hoch. Dies entspricht nicht unbedingt den Erwartungen. Im Allgemeinen sind japanische Unternehmen bei der Umsetzung der digitalen Transformation im Rückstand. In diesem digitalen Bereich hat die KI in den letzten Jahren jedoch besondere Aufmerksamkeit erhalten. Das Telekommunikationsunternehmen NTT etwa trainiert bereits sein eigenes großes Sprachmodell (LLM), das speziell auf den japanischen Markt zugeschnitten ist. Dieses wird es noch 2023 auf den Markt bringen.

Unternehmen wie Panasonic nutzen LLMs bereits in großem Umfang. In anderen Industrieunternehmen sehen wir, dass sich Investitionen in maschinelles Lernen und LLM-Technologie auf die Geschäftsaussichten auswirken.

Auf Sicht von mehreren Jahren sind wir überzeugt: Die Implementierung von KI-gesteuerter Technologie wird ein entscheidendes Unterscheidungsmerkmal zwischen Gewinnern und Verlierern sein – in allen Branchen. Für uns als „Stockpicker“ geht es dabei nicht nur um Halbleiter. Der Grad der KI-Bereitschaft und KI- Implementierung in den von uns analysierten Unternehmen ist für uns ein wichtiger Aspekt, wenn wir die mittelfristigen Gewinnaussichten einschätzen.

Die bereitgestellten Informationen sind nicht als Empfehlung zum Kauf oder Verkauf eines bestimmten Wertpapiers zu verstehen. Die in diesem Dokument zum Ausdruck gebrachten Ansichten sollten nicht als Empfehlung, Beratung oder Prognose aufgefasst werden. Die frühere Wertentwicklung stellt keinen Hinweis auf die künftige Wertentwicklung dar. Der Wert eines Investments kann sowohl fallen als auch steigen. Dies führt dazu, dass Preise steigen und fallen können, und Sie bekommen möglicherweise weniger zurück, als Sie ursprünglich investiert haben.

Von Naver in Korea bis zu Alibaba und Baidu in China: Viele der Softwareunternehmen der Region betonten, dass sie KI-Tools anbieten können, die das Kundenerlebnis verbessern und zu einer intensivierten Kundenbindung führen. Verständlicherweise haben sich diese Ankündigungen eher wenig auf die Aktienkurse ausgewirkt. Der Markt versucht festzustellen, wer sein Geschäft und seine Rentabilität mit KI verbessern kann, anstatt nur seine aktuelle Marktposition zu halten.

Auch wenn es nicht besonders attraktiv klingt: Für große Finanzinstitute und Logistikunternehmen könnten sich echte Chancen ergeben. Sie könnten ihre Kunden mit KI-Einsatz viel individueller und maßgeschneiderter bedienen – und das zu deutlich geringeren Kosten, da der Bedarf an menschlicher Mitwirkung sinkt. Entscheidend ist, dass Unternehmen eigene Datenpools mit der Kundenerfahrung verknüpfen können. Aus diesen Daten müssen sie „maschinell lernen“ können, um die Kundenbedürfnisse zu verstehen und zu antizipieren. Diese Trends waren bei diesen Unternehmen bereits fest verankert. Doch die Verbesserungen im Bereich der KI dürften sie noch beschleunigen.

Verstärkt beschäftigen wir uns beispielsweise mit indischen Software-Dienstleistern. Diese sehr erfahrenen IT-Unternehmen arbeiten und experimentieren schon seit einiger Zeit mit KI. Daraus ergibt sich eine echte Chance für sie, die Prozesseffizienz ihrer Kunden mit wirksamen KI-Lösungen zu erhöhen. Es besteht jedoch auch ein reales Risiko: Disruptoren könnten mit Billiglösungen in den Teil des Marktes eintreten, in dem die Wert-schöpfung geringer ist. Die negativen Auswirkungen auf das Preisgefüge könnten dann einen Großteil der KI-bezogenen Gewinnen wieder zunichte machen.

²Quelle: Bloomberg, 30. Juni 2023.

Die bereitgestellten Informationen sind nicht als Empfehlung zum Kauf oder Verkauf eines bestimmten Wertpapiers zu verstehen. Die in diesem Dokument zum Ausdruck gebrachten Ansichten sollten nicht als Empfehlung, Beratung oder Prognose aufgefasst werden. Die frühere Wertentwicklung stellt keinen Hinweis auf die künftige Wertentwicklung dar. Der Wert eines Investments kann sowohl fallen als auch steigen. Dies führt dazu, dass Preise steigen und fallen können, und Sie bekommen möglicherweise weniger zurück, als Sie ursprünglich investiert haben.

Healthcare services

In healthcare services, US firm Nuance provides a conversational AI virtual assistant platform to collect patient data. With better data analytics, NLP and deep learning models can be applied to improve patient outcomes by predicting conditions and decreasing the cost of care. New York-based medical centre NYU Langone Health is using LLMs to predict the risk of 30-day re-admission after patients are discharged from the hospital.

Drug discovery

LLMs are also being developed for drug discovery. While still in its nascent stage, we believe as generative AI becomes more widespread and mature, we will see greater adoption by healthcare companies as it pertains to drug discovery. Pharmaceutical firm AstraZeneca is an early adopter of Nvidia’s GPUs, building LLMs for molecular biology to explore datasets of small molecules, proteins and, soon, DNA. 

Financial services

Traditional enterprises make up 30% of Nvidia’s customer base for its GPUs. More recently, the company has noted particular interest from asset managers who are exploring further use of ChatGPT in their day-to-day business. The provision of services such as client reporting, customer service, and back-office operations can become significantly more efficient by utilising generative AI.

Utility companies

Increasingly, utility companies are embracing AI to help improve the electrical grid – both for ‘smart grid’ optimisation and also to help reduce outages from unforeseen incidents, such as lightning strikes. Our global equities colleagues, who run a listed infrastructure strategy, offer more colour below on the applications of, and opportunities from, AI in the utilities sector. 

Goliath vs. goliath: evolving competitive landscape as generative AI goes ‘mainstream’ 

Nvidia’s rise opens the door for rivals to take a share of the pie. As a technology gets more broadly adopted, levers of competition often shift from ease of use to cost and economics. This is where Nvidia's 80% gross margin on GPUs leaves a very wide pricing umbrella for competition to come in. Hyperscalers, such as Microsoft, Meta Platforms (formerly Facebook) and Amazon, are already working together with fabless chip design houses to make custom chips (ASICs – Application-Specific Integrated Circuits) for accelerating workloads. Google already uses its own custom processors, TPUs (Tensor Processing Units), in many workloads.

As the size of these AI models keeps growing at an exponential pace, there are more problems to be solved. 

• While computer chips have kept pace, the memory bandwidth hasn’t. This means incremental performance gains can also be made by solving this bandwidth problem through software (instruction sets) and hardware (interconnects, CXL – Compute Express Links). 
• Furthermore, larger models require multiple pods of GPUs connected through high bandwidth networking gear. 
• Finally, generative AI reduces the barriers and costs of content creation. For instance, Adobe Stock is a library of circa 200 million images. Over the last couple of months, while Adobe Firefly (Adobe’s generative AI) has been in beta, users have created 200 million images using AI. This has implications for storage demand.

Democratising the use of generative AI 

Data has already been hailed as the new oil. Generative AI might as well be the internal combustion engine of our age that helps unleash its value. Businesses that own proprietary datasets will become valuable, especially since those datasets are required to create the generative models. 

Earlier this year, Getty Images sued Stability AI for copyright infringement⁶, alleging the latter copied 12 million images to train its AI model. The results of this case will have serious implications in the world of content generation and copyright protection – whether it is images, music, videos or advertisements. 

Most software businesses will be turbocharged with the use of generative AI. At its core, generative AI is a piece of software that needs specialised chips to run. The value of this technology will be realised in real world software applications that help improve productivity across the board for its users. 

Microsoft and Adobe have already teased us with what’s coming next. We will have meeting notes transcribed and summaries made, with any follow-ups or action items highlighted. Institutional knowledge that sits in unknown corners of an organisation could be aggregated and codified. The chatbots on the phone or on a website will become exponentially more helpful. Each one of us could become a graphic designer, making art using language with little to no knowledge of how to operate a Photoshop tool. 

When language becomes the “API⁷” to access and benefit from most software tools, their usage gets democratised. Applying generative AI to existing software will be a significant and durable pricing growth opportunity for incumbents. 

A review of generative AI would be incomplete if we didn’t note the ongoing Cambrian explosion in open-source domain around smaller models and applications of generative AI. 

In February 2023, after Meta publicly released their 65B parameter LLM to help researchers in the field of AI, multiple fine-tuned derivates showed up. This fine tuning sometimes cost as little as US$300⁸. The availability of an open-source foundation model lowered the capital barrier, and broad interest has fuelled open-source innovation as well as disruptive market entrants. Stability AI, a company we referred to earlier, is an open-source company with 200,000 members in its community. Midjourney, another generative imaging tool (used to create the cover image and the image above), has only 11 full-time staff and has chosen to remain private and self-funded⁹. 

Thriving and surviving

It’s a very interesting time for innovation and competition. The adoption of generative AI in various industries will lead to significant productivity benefits and consumer surplus. Of course, as with any new technology, many businesses will jump on the generative AI bandwagon. So, while it’s a very exciting time to be investing, it will be important to separate the wheat from the chaff. 

² https://www.reuters.com/technology/chatgpt-sets-record-fastest-growing-user-base-analyst-note-2023-02-01 
³ https://www.coca-colacompany.com/media-center/coca-cola-launches-masterpiece-campaign
⁴ https://www.nytimes.com/2022/09/02/technology/ai-artificial-intelligence-artists.html 
⁵ https://www.bbc.co.uk/news/world-us-canada-65735769
⁶ https://www.theverge.com/2023/2/6/23587393/ai-art-copyright-lawsuit-getty-images-stable-diffusion
⁷ An application programming interface (API) is a way for two or more computer programs to communicate with each other. It is a type of software interface, offering a service to other pieces of software.
⁸ https://lmsys.org/blog/2023-03-30-vicuna/
⁹ It is cash flow positive, unlike many start-ups of the last decade

The information provided should not be considered a recommendation to purchase or sell any particular security. The views expressed in this document should not be taken as a recommendation, advice or forecast. Past performance is not a guide to future performance. The value of investments will fluctuate, which will cause prices to fall as well as rise and you may not get back the original amount you invested.

Water

AI can play a crucial role in optimising water resource management. By utilising machine learning algorithms and real-time data analysis, smart water systems can detect leaks, predict water demand, and optimise water distribution. American Water Works – the largest US water utility – has been using machine learning to monitor its ageing network of water pipes, detect deteriorating tank coating, and predict water leakage. This allows for more efficient use of water resources and ensures the availability of clean water for customers. 

Water utility companies that are adept at leveraging these systems will earn higher regulated returns through existing incentive mechanisms that encourage water resource efficiency and reliability. 

Waste Management

AI can help waste management and recycling companies improve waste collection, sorting, and recycling processes. AI can also automate waste collection routes, which will generate cost savings for waste collection companies. AI-powered sensors and robotics can significantly improve sorting accuracy for recycling, so companies adopting these systems can generate new revenue streams from higher recycling efficiency. These advances also help accelerate the move towards a circular economy. 

Waste Management and Republic Services – two of the largest waste management companies in the US – have invested in modern systems that use artificial intelligence and robotics to automate waste sorting. These advanced systems can separate recyclable materials from the waste stream with extremely high accuracy. 

Transportation

Transportation infrastructure companies can leverage AI to improve mobility efficiency and safety. Traffic management systems using AI for predictive analytics can optimise traffic flow, reduce congestion, and enhance road safety. Toll road operators can utilise data analytics to predict demand patterns and dynamically adjust tolls to maximise revenue. Ferrovial – the largest Spanish toll road operator – is developing the AIVIA orchestrated connected corridors, a project that uses AI to integrate connected autonomous vehicles safely into our future roads. 

Demand for data centres 

The explosion of AI applications can directly translate to data centre demand. Companies are queuing to get access to more computational resources. Increasing demand for AI computational resources can lead to both higher revenue growth and better pricing power for data centre operators such as Equinix and Digital Realty. In addition, AI workloads can be computationally intensive, resulting in increased energy consumption. Data centres will need to invest more to ensure sufficient power and cooling infrastructure that can support AI operations. 

Risk management

Finally, for infrastructure construction projects, AI can analyse historical project data such as usage patterns and maximum load to assess the risks associated with each project. It can also help identify potential risks during the project lifecycle, such as predicting delays, cost overruns, and environmental impact. By providing accurate risk assessments, AI helps infrastructure companies and infrastructure investors develop effective risk mitigation strategies and make more informed decisions.

Undoubtedly there will be risks and challenges associated with the rapid uptick in AI usage but, as highlighted above, by incorporating AI technologies into transportation, energy, water, and waste management systems, companies have the opportunity to unlock higher levels of infrastructure efficiency and resilience, supporting future growth and revenue gains across multiple business units. 

The information provided should not be considered a recommendation to purchase or sell any particular security. The views expressed in this document should not be taken as a recommendation, advice or forecast. Past performance is not a guide to future performance. The value of investments will fluctuate, which will cause prices to fall as well as rise and you may not get back the original amount you invested.

In recent months, we have spoken to management teams who are already witnessing significant increases in the speed and efficiency of drug development, thanks to large language models’ (LLMs) ability to analyse enormous datasets. It seems AI could potentially enable these companies to create new medical discoveries and healthcare advancements that were previously unimaginable.

The UK’s large finance sector may also begin to undergo an AI transformation. In banking, the technology could potentially revolutionise risk assessment and fraud detection. Financial institutions are also likely to leverage AI to analyse vast datasets, drive efficiencies and offer clients better solutions. HSBC, for instance, recently launched its AI Markets analytics service for institutional clients, which it claims uses a proprietary Natural Language Processing (NLP) engine¹⁰.

Elsewhere, we would note developments in retail, with companies like Ocado and ASOS using AI to enhance their customer engagement strategies. Meanwhile, manufacturers such as Renishaw and Rolls-Royce are leading efforts to boost productivity and operational efficiency using AI-powered automation and predictive maintenance. And across companies of all sectors and sizes, AI could help to address the UK’s particularly poor productivity levels and labour market shortages.

Of course, technological change creates uncertainty, and UK investors are also likely to navigate some significant challenges in the period ahead. AI’s threats could take years to materialise, but the market tends to act first and ask questions later. In May, US-listed education provider Chegg’s share price fell almost 50% when the company cited AI as a potentially systemic threat. On the same day, UK-listed Pearson’s share price fell 15%, despite no compelling evidence that its business model faced the same risks. Pearson’s share price has subsequently recovered around half of its decline, while Chegg still trades at close to its lows¹¹.

At this stage, it is difficult to fully comprehend the ramifications of emerging AI technologies. However, it has quickly become a key area of focus for investors across all markets, and the UK is no exception.

In time, we may be required to revisit the fundamental investment case for any given company in our portfolios. Our priority will, therefore, be to anticipate and move ahead of perceived risks, while at the same time identifying potential opportunities, which seem likely to present themselves in a UK market whose potential is currently overlooked, in our opinion.

¹⁰ Source: HSBC launches AI Markets | Insights | HSBC, May 2023
¹¹ Source: Datastream; PSON, CHGG, local currency share price returns; as at 22 June 2023
¹² Source: Datastream; PSON, CHGG, local currency share price returns; as at 5 May 2023

The information provided should not be considered a recommendation to purchase or sell any particular security. The views expressed in this document should not be taken as a recommendation, advice or forecast. Past performance is not a guide to future performance. The value of investments will fluctuate, which will cause prices to fall as well as rise and you may not get back the original amount you invested.

Emerging Markets (EM) investors have experienced a sharp divergence year to date between the performance of banks compared to IT stocks (Figure 3). Banks and insurers have been hit by the sentiment fallout from the US banks debacle in March and general risk-off sentiment, as investors question the strength of the recovery in the Chinese economy, in particular. Meanwhile, the IT sector, which was hit hard last year (down 33% in US dollar terms)¹⁴, has bounced despite the still sluggish nature of IT product demand, particularly within consumer products such as laptops and smartphones. 

So, why the optimism? 

In a word(s), Artificial Intelligence (AI). As generative AI fever has swept the investment world, IT stocks have been lifted in the wave of excitement regarding sectoral opportunities. The IT sector in EM is Asia-centric, dominated by semiconductor heavyweights, such as TSMC and Samsung Electronics, in Taiwan and Korea, respectively. Investors have rushed to price in the more obvious AI benefits for the giant semiconductor players in the region.

Demographics as a catalyst

If we look down the list of countries contributing the most to global traffic on the ChatGPT platform in the first quarter of this year, the usage in Asia is dominated by India and Indonesia. What unites those two countries? Their age demographic, with c.50% of their populations below the age of 30 – driving an acceleration of tech adoption and innovation. 

Africa

That same demographic characteristic is shared by Africa, with 70% of the continent's population below the age of 30. AI offers the continent huge opportunities to accelerate development and drive an improvement in productivity. We are already seeing the vast agricultural sector embracing technology (including AI) in order to drive greater crop reliability and yields, from Ghana to Kenya¹⁵. There will surely be considerable investment opportunity to come, spurred by such development, with tech adoption driving better data collection and transparency over time. We would look to the operators in the traditional telecoms sector such as Safaricom (Kenya) or MTN (South Africa) as natural conduits for AI investment.

Middle East

The Middle East has the capital to match its ambition and is moving quickly to take advantage of AI opportunities. With the United Arab Emirates (UAE) the transport, capital, cultural and linguistic bridge between the Middle East and the rest of the world, its sovereign wealth funds (SWFs) have already taken a prominent role looking to shape and invest in AI locally and globally. 

Mubadala-backed Emirati AI firm G42 set up a US$10 billion tech fund last year to operate as a private equity (PE) firm, looking to invest in emerging technologies globally, including AI. G42-backed Astra Tech¹⁶ has already unveiled the first ‘Arabic ChatGPT’ in the Middle East and North Africa (MENA) region to integrate into its BOTIM app, as the language is not well-served by the global tech players.

In Saudi Arabia, the government has made clear their ambition with its Vision 2030 to diversify away from economic reliance on hydrocarbons by investing heavily in information and communication technology (ICT) infrastructure and services. Global players like Microsoft, Google, Oracle, Meta and Apple are all investing in the Kingdom, attracted by its deep pockets, high internet usage and plans for the future. 

Again, the local drive is led by the local telecommunications incumbents, Saudi Telecom Company (STC) and Etisalat (e&). But non-telcos are also adopting fast, with companies as diversified as Dubai Electricity & Water Authority (DEWA), online travel company Almosafer, edtech start-up Abwaab and concierge service provider Yanzo integrating ChatGPT into their respective systems.

Latin America

Across Latin America, financial firms have moved quickly to adopt ChatGPT and AI into their tech arsenal, often deploying on an experimental basis. Such firms are in a strong position to do so, given the vibrant fintech scene, along with the sophistication and years of experience incumbents already possess in responding to technology threats and opportunities for their businesses. 

Banorte, a leading bank in Mexico is a good example, trialling ChatGPT in their corporate communications. The potential for contextual AI use in financial services is enormous, but the search goes on for appropriate real-world application, especially on the customer-facing side. The industry is naturally moving cautiously, given the very real regulatory sensitivities around operational risks. 

The more obvious enterprise use cases for ChatGPT and AI in general are within automated customer care, conversational e-commerce, AI-powered virtual assistants and work efficiencies. By example, Latin America's largest e-commerce ecosystem MercadoLibre has adopted Vue.ai (a leading AI provider for retailers) in their vendor onboarding processes to process and assess text and images faster and more accurately¹⁷. This has cut human intervention by 50% and reduces the prospect of human error, with a commensurate 56% collapse in false positives detected in text assessed. 

Advantages of AI unevenly accrued

Alongside feverish corporate activity to better understand how companies can integrate AI and ChatGPT into their businesses comes natural anxiety regarding the risks of hasty adoption in terms of exacerbating further economic inequality in vulnerable communities across EM. 

The emergence of ChatGPT and other AI applications has catalysed a debate regarding the implications of these technologies for the future of work. There is real concern that the benefits of AI may accrue narrowly to those at the top of the economic pyramid, while the downsides in terms of jobs displacement will be felt mainly by the poor and working class. 

In a recent report, McKinsey & Company concludes that adoption of AI is likely to be faster in developed countries with the associated implication that productivity benefits will accrue in greater share to those already technologically savvy. High wages in rich countries create a greater rush for automation to cut costs. By comparison, lower wages in countries like Mexico put less pressure on companies to adopt AI. 

It is incumbent on policymakers and politicians across the emerging world to prepare for the AI revolution and equip their populations with the ability to reap its rewards.

 P.S. This comment was prepared without the use of ChatGPT so all mistakes are my own. This may change in the near future…

¹⁴ Source: Bloomberg, data as at 31 December 2022
¹⁵ Source: Africa will be transformed by the potential of AI and data – if we can get investment | Mahamudu Bawumia | The Guardian
¹⁶ BOTIM launches region's first Arabic ChatGPT with MBZUAI (gulfbusiness.com)
¹⁷ Source: MercadoLibre automates image moderation with AI | Vue.ai
¹⁸ Source: McKinsey & Company, Jun 2023, the-economic-potential-of-generative-ai-the-next-productivity-frontier-vf.pdf (mckinsey.com)

The information provided should not be considered a recommendation to purchase or sell any particular security. The views expressed in this document should not be taken as a recommendation, advice or forecast. Past performance is not a guide to future performance. The value of investments will fluctuate, which will cause prices to fall as well as rise and you may not get back the original amount you invested.

Drug discovery

• Only about five in 5,000 of the drugs that begin preclinical testing ever make it to human testing. Only one of these five is ever approved for human usage¹⁹.
• Thermo Fisher Scientific, the global leader in R&D tools, incorporates AI tools within its analytic equipment. For example, its near atomic resolution images use AI to enable researchers to better understand druggable targets and potentially accelerate drug discovery.

Earlier diagnosis

• The human genome is 3,000,000,000 base pairs long²⁰ and AI is a crucial tool to help make sense of all that data.
• On average, a rare disease diagnosis takes 5-7 years and disproportionately impacts children²¹. Leaders in genomic sequencing are using AI software to identify rare genetic variants in newborn babies quicker, enabling earlier treatment.

Supported decision making

• The World Health Organisation (WHO) estimate there will be a global shortage of 10 million healthcare workers by 2030²². AI will play a key role in supporting healthcare professionals with decision making.
• Masimo, a leading medical technology company, launched a remote patient monitoring system. Capturing over 60 parameters the system uses machine learning to establish a personalised baseline for a patient to more accurately escalate patient distress with fewer false alarms, enabling greater healthcare worker productivity.

Personalised medicines

• Precision medicine is an approach to tailoring disease treatment based on a patient’s own genetics. AI will be key in facilitating the shift from a one-size-fits-all approach to personalised treatment plans.
• AstraZeneca’s targeted cancer treatment Lynparza is an example of personalised treatment for cancer patients. Genetic sequencing leverages AI to identify patients with the BRCA-gene mutation, who are potentially suitable candidates for the treatment.

Naturally, ethical guidelines and regulation lag advances in technology. An open conversation around both the positives and negatives are important. With this in mind, we must not lose sight of the transformative potential of AI in bettering healthcare outcomes.

¹⁹ https://ca-biomed.org/wp-content/uploads/2020/08/FS-DrugDevelop.pdf 
²⁰ https://www.genome.gov/genetics-glossary/Base-Pair#:~:text=One%20copy%20of%20the%20human,to%20300%20million%20base%20pairs. 
²¹ https://www.thelancet.com/pdfs/journals/landia/PIIS2213-8587(19)30006-3.pdf 
²² https://www.who.int/health-topics/health-workforce#tab=tab_1 

The information provided should not be considered a recommendation to purchase or sell any particular security. The views expressed in this document should not be taken as a recommendation, advice or forecast. Past performance is not a guide to future performance. The value of investments will fluctuate, which will cause prices to fall as well as rise and you may not get back the original amount you invested.

Unfortunately, such mega-cap infrastructure firms are absent from the convertibles universe (Nvidia’s latest convertible matured five years ago, while there is only a small synthetic convertible for Microsoft). Some convertible issuers in the semiconductor and hardware space may enjoy some exposure to the theme by providing the equipment needed for AI workloads (for example, memory or some wafer fab equipment makers).

A potentially richer vein of opportunities could be found in other technology, media and software names, which account for roughly 27% of the convertibles universe. With ferocious demand for AI-driven automation, firms that already provide a variety of solutions and have already incorporated some AI capabilities - in a previous incarnation, more likely to be machine learning rather than the newest large language models (LLMs) - are likely to benefit; both in terms of greater sales volumes and higher price points, leading to higher margins.

Moreover, we could see cost reductions and greater innovation if fewer software engineers are needed as generative AI disrupts the industry. Indeed, to date, the scarcity of programmers has been a big negative for many tech companies, and convertible bond (CB) issuers, that could not control R&D-related expenses to generate appropriate earnings and cashflow. However, some legacy software providers such as on-premise solutions, education and marketing/CRM, and contact centre solutions, that can be replaced by AI-driven automation could be at risk.

Another area in convertibles that may deliver interesting investment opportunities is biopharmaceutical and medical technology/diagnostics companies, along with technology-enabled healthcare services. These roughly account for 12% of convertibles outstanding. Here, generative AI could hold the key to boosting productivity significantly, speeding up clinical research, reducing drug approval time, limiting unnecessary healthcare expenditure and improving clinical outcomes.

We expect a wave of AI issuers

Regardless of which companies turn out to be the winners and losers, we expect we will soon see a wave of AI-related convertible issuers hitting the market – possibly as soon as the second half of this year. Based on historical precedent and the market appetite for all things AI, we anticipate a flurry of fledgling AI firms looking to raise capital via CBs. If enthusiasm for AI remains elevated, these could come at less favourable terms for investors (and correspondingly, sweet deals for issuers). 

We have seen many similar bursts of issuance over the years, for example SaaS²⁴, social media, eCommerce / online marketplaces, payments, diagnostics / genetics, digital health, renewables, crypto and EVs and batteries, to name a few. Such an AI-wave could help meet, and even surpass, our beginning-of-the-year expectations of US$70-80 billion in new issuance in 2023, as we are currently annualising at about US$79 billion as of the end of June 2023.

So where does all of this leave investors?

Currently, we are witnessing a high level of excitement and elevated valuations in the market, and the potential for increased volatility as the market digests the implications of generative AI. 

While we expect a large wave of AI-related new issuance, we potentially face long lead times to achieve the highly-anticipated productivity boost of generative AI, while maturity windows for convertibles remain comparatively short.

Given this backdrop, we remain vigilant and maintain our selective approach; assessing the business prospects of potential investments in a clear-headed manner; mindful of the price we pay, the credit quality of bond issuers and the risk-reward in our capital allocation decisions. 

Regardless of what happens with AI, outside of technology we see increasing sectoral diversification in new issues (with more utilities and new REITs and financials coming to market), and attractive valuations in Asia where credit spreads are wider and equities have given back some of their China post-COVID re-opening gains. 

We think this is the way to generate attractive risk-adjusted returns when faced with the enticing prospects of generative AI, but also the many potential pitfalls that it could bring.

²³ Original Isaac Newton quote, ‘‘I can calculate the motion of heavenly bodies but not the madness of people’’.
²⁴ Software as a Service

The information provided should not be considered a recommendation to purchase or sell any particular security. The views expressed in this document should not be taken as a recommendation, advice or forecast. Past performance is not a guide to future performance. The value of investments will fluctuate, which will cause prices to fall as well as rise and you may not get back the original amount you invested.

US banking crisis in the rear-view mirror

Investors seem to have put the US banking crisis firmly behind them. With mixed economic data, inflation coming down at a stubbornly slow pace and central banks reiterating their willingness to rein in inflation, market expectations evolved over the second quarter from pricing in significant rate cuts in the US to a more central bank-aligned view of “higher for longer” rates. 

This mainly affected yields at the short end of the US curve, while longer-term yields have continued to flirt with the 4% level. In this phase, yields in the gilt market have also continued to push higher on the back of high inflation, a resilient labour market and strong wage growth. The expectation is that the Bank of England (BoE) has more work to do than the US and even Europe. However, as higher mortgage costs start to put a significant dent in UK household disposable income between now and the end of the year, the market could easily shift its focus to the need to cut rates, rather than hike. Overall, we have increased duration across many of our portfolios, with a preference for US and UK exposure. 

Impact of tighter monetary policy

Time will tell when and how the lagged effects of tighter monetary policy will work their way through the economic system. Emerging market sovereigns may offer us a glimmer of hope, with many (having started their inflation fight early) now seeing inflation decelerating more convincingly than developed markets, driving attractive returns across a range of emerging market debt and currencies. South Africa has been an outlier: macroeconomic concerns on the back of extended power outages were exacerbated by the news of Russian ties, potentially impacting the country’s relationship with the US. 

While we have taken some profit across emerging markets over the quarter, the resultant increase in South Africa’s risk premium provided us with an opportunity to add exposure to the country’s government bond and/or currency across several of our portfolios, as the legitimate worries of the market were more than compensated by the observed price moves, in our view. 

Elsewhere, Japan has provided exciting returns for investors that have been willing to stay the course. Earnings delivery has been strong but valuations at an index level are now significantly higher than they were at the start of the year and potentially less attractive than major European equity indices. Although we continue to find Japanese equities attractive on a longer-term basis, we have scaled back on broader market positions in some of our portfolios, as we recognise that much of the alpha going ahead is likely to be derived via stock selection. 

Potential for sentiment to pivot in data-driven environment

As we move through the second half of the year, most assets appear fairly priced, with some selected opportunities within equities and fixed income, as described above. We maintain a cautious stance overall as the current flow of mixed economic data continues to provide investors with fertile ground for an abrupt change in narrative and swing in market sentiment (and prices!).

²⁵ Source: UBS S&P 500 Ex-Tech basket tracking the performance of the S&P 500 ex Information Technology, Communication Services, TSLA and AMZN.

The information provided should not be considered a recommendation to purchase or sell any particular security. The views expressed in this document should not be taken as a recommendation, advice or forecast. Past performance is not a guide to future performance. The value of investments will fluctuate, which will cause prices to fall as well as rise and you may not get back the original amount you invested.