Macro
The US economy has demonstrated exceptional resilience in the face of aggressive monetary tightening. This is further underscored by the emergence of increased productivity and the early signs of an economic boost from AI.
While the US maintains its supremacy in the long run, driven by innovation, leadership in cutting-edge technologies, control of the world’s reserve currency, energetic independence, private property and the rule of law, and vibrant capital markets, its short-term outperformance is primarily attributed to temporary factors. These include unprecedented fiscal support and an increased propensity to spend by US consumers despite consumer sentiment hitting a 7-year low. A potential change in these factors could pose a critical risk to the US economic outlook, a concern for investors.
For the US to sustain its pre-eminence, it must address several long-term factors. These include reducing public debt, reforming social care in the face of an ageing population, working on inequality and environmental goals, and improving health and educational outcomes. As the November election approaches, these issues should be at the forefront of political discourse, shaping voter behaviour and future policy direction. However, the potential influence of existing social and political divides on voter priorities could shift attention to short-term issues that resonate more deeply with voters, posing a challenge to long-term planning.
Rates
The Fed is staying the course and is data-dependent. Given robust economic growth, close to full employment, and a bullish risk market, it sees no urgency in cutting rates. Chairman J. Powell confirmed this week that the current policy is restrictive and that, eventually, we will see a fundamental weakening in the economy. He is now calling for one rate cut this year, while the market expects two in September and December.
J. Powell’s remarks were priced in immediately (we moved 30 bps on the 10Y this week alone) and, as reflected by the ‘CME FedWatch Tool,’ removed one side of the probability distribution for future interest rate movement. The expected move is to the downside but with a softer path as the FED sees no urgency. This expectation prompts demand for Treasuries; this week, the US Treasury auction shows high demand, with 10Y and 30Y receiving bid-to-cover ratios of 2.67 and 2.49, respectively.
The FED would only expedite cuts if they saw a more significant weakening of the jobs market, a rollover in the real estate market, further weekends of the economic data, or increased stress in the risk markets. Critical indicators to watch relate to the job market. As job openings have recently plummeted, some FOMC members could be concerned about this pace and start worrying that this might roll over and affect employment.
In the first half of this year, 10Y mainly traded between 3.80% and 3.70%, and now we are in the middle of this range. The first half was more of an upward cycle of rates; now, we are in a downward cycle. The expectation is that yields will move closer to the lower part of this range and will be less volatile than in the first half.
Credit
Spreads are close to all-time tights; however, investors are looking to other metrics to judge the riskiness of the credit market. For example, the average duration of bonds in the HY is relatively short compared to the historical level, and the prices are at a discount, giving more investors favourable convexity.
US junk bonds have the best returns in the sixth week, and we see continuous demand. Buyers are setting in at any discount except for the lowest quality of the spectrum.
From the sector perspective, we observe more stress in the less cyclical parts of the economy, specifically in the areas requiring high capital replacement. Companies in those areas will see their free cash flow erosion due to the high-interest expense. On the other hand, there are more opportunities in the industries, materials, and chemicals sectors and more potential risks from the consumer, media, and telecoms sectors.
As investors expect rates to be cut, they will look to reduce their
exposure to floating-rate credit and increase duration-sensitive exposure. A higher-quality and longer-duration portfolio tends to work better in the late cycle, which many investors describe today. Avoid adding higher risk in the portfolio in the distressed part of the market.
Equities
US indexes hit a new all-time high, led by technology stocks, which now have the highest valuation in 23 years. Investors now have above-average equity exposure, and sentiment has new highs. There are also worrying signals, such as a narrower breadth of the rally, low and falling VIX Index, and low tail heading demand, all pointing to investors’ complacency. Investors’ bullishness is driven by one major technological shift—AI, which we are going to address this week.
So far, the AI trend has been all about building infrastructure and a ‘picks and shovels’ strategy. This strategy involves creating hardware and foundational tools necessary to develop, deploy or scale AI applications. The biggest winner of this strategy is Nvidia, as its high-performance GPUs are used extensively in training AI models. GPUs and Nvidia’s CUDA platform create a robust ecosystem for deep learning frameworks (such as TensorFlow and PyTorch), scientific simulations, and high-performance computing. These features provide the only feasible end-to-end solution for AI development and lock in developers and their applications in Nvidia’s ecosystem.
But now, we can observe the first successes of the beneficiaries of this infrastructure. Innovation occurs on every level of the tech stack – hardware, models, and applications. As AI moves up the application stack, we see many new promising software companies. Every SaaS (software as a service) is trying to incorporate AI into its product roadmap, and there are new entrants into enterprise generative AI. We see a global increase in AI applications, and they start being deployed by companies with a significant boost to the productivity of knowledgeable workers.
Application-layer startups are raising millions of dollars. However, investors are concerned about their product’s ability to stay relevant in the long run. At the same time, owners of the foundational large language model (such as Open AI with chat GPT) that underpin those applications are expanding their features and, over time, might directly compete with higher-level applications. Applications must focus on leveraging LLMs and harnessing the power of underlying models rather than competing with them.
It’s more challenging to target specific applications in the public markets. AI bets are oriented around mega-cap tech with foundational architecture or around cloud companies.