Andrew Pyle
February 16, 2024
Crowding out risk
There are many positive factors attributed to the Artificial Intelligence (AI) craze. It is going to make companies and people more efficient, cutting down the time required to perform tasks. This productivity boost is welcome at a time when the population age wave is cresting, and labour supply is becoming more constrained. If true, then the coming years will see earnings growth that will not only validate current stock market valuations but higher ones yet.
There is just one not so little snag and one that we have highlighted many times before. AI is an expensive capital outlay. For companies with strong balance sheets and net free cash flow growth, this shouldn’t be an issue. Other companies may struggle to fully embrace those AI technologies that will ultimately keep them in a competitive position with their peers. Challenges are a normal part of corporate life and some companies do fail to remain viable. The risk that is common to both the strong and not so strong companies is that spending on AI may divert capital away from other investments, whether in equipment, non-AI systems and people, to name a few.
This is what we refer to as crowding out. Many of you will recall this term from back in the 80s and 90s, when governments were racking up hefty deficits, resulting in increased borrowing. As the supply of new government debt rose, this squeezed out available demand for corporate bonds. As they say, there are only so many dollars to go around. In this case, we aren’t necessarily talking about crowding out investment dollars in the bonds space, but this is one source of capital available to companies, in addition to public equity, private equity and private credit.
Indeed, we have seen a surge in new public debt offerings in recent weeks, especially in the U.S., and we are on track in February to beat the record for investment grade bond offerings. Some of this increased activity can be put down to opportunistic funding by companies taking advantage of strong institutional and retail demand for bonds. While interest rates are poised to decline at some point, which you would think would lead companies to delay issuing new debt in order to get a lower cost of capital, the demand conditions at that point may not be as robust as they are now. In other words, spreads on corporate bonds could be wider. As the above chart shows, the spread between the 10-year U.S. triple-B corporate bond yield (aggregate) and the 10-year treasury yield is not far from the tightest levels over the past ten years. Even if we saw only a modest slowing in the economy (like 2018-2019), the spread could easily widen out back towards 2%. If long-term yields didn’t fall by a commensurate amount, all-in funding costs would increase.
Still, some of this increase may also reflect the need for funding as it pertains to spending on new technologies, such as AI. The difficulty we have is in breaking out where this spending is going. It’s not like companies are just buying ChatGPT. There is a software aspect to integrating AI into a company’s business model, but there is also a hardware and personnel component as well. Let’s start with the macro view. U.S. business fixed investment is made up of three main components – structures, equipment and intellectual property products (software and research). In the fourth quarter, total business investment was about $3.3 billion in real 2017-dollar terms. Of this, structures accounted for $641 billion, equipment was $1.25 trillion and intellectual property spending was $1.4 trillion.
The chart above shows these three main components over the past 20 years (indexed to 2003). Spending on structures has grown by a total of 40% over this period, but for all intents and purposes has been flat. That growth is just under half that of equipment and not even a fifth of what we have seen from intellectual property. Prior to 2015, the growth in both equipment and intellectual property was fairly even, but the two have diverged sharply since then. In the past ten years, company spending on software and research has grown by 84%, compared to about 13% for equipment. Again, not all intellectual property investment goes to AI but, then again, there is a decent proportion of equipment spending that is in computers. In fact, at around $500 billion last quarter, it is the largest component of equipment spending.
Considering the speed at which systems need to be upgraded, there is almost a built-in guaranteed faster obsoletion rate for computers and peripheral equipment, as adoption of AI builds. It’s not just computers that wear out, however. Manufacturing equipment needs to be replaced and upgraded and the same holds true for physical structures. Failure to do so can impact productivity and a company’s competitive position and financial health going forward. This would be true even in an environment where there was a clear and predictable path to positive implementation of AI in the business model and where the outlook for economic growth was also positive.
What if the economy doesn’t support the growth in revenues required to pay for both AI and non-AI spending? The jury has been deliberating regarding the health of the global economy for quite some time and numbers out of North America have painted a picture of resilience in the face of higher rates. The picture became a little blurrier this week. Japan and the UK reported that their respective economies contracted for a second time in a row last quarter, which suggests recession. Europe grew by a tenth of a percent in Q4 but has essentially stalled. On this side of the pond, we saw U.S. retail sales fall by 0.8% in January and industrial production declined by 0.1% in the same month. Next week, we get the January leading indicator report but, as of December, the index had fallen for 22 straight months. This doesn’t mean a recession is coming, but it does point to a moderation in growth.
Investment in AI has never been an overnight thing, even though market participants sometimes act as though it is. It will be a long-term development and the technology is going to become more pervasive, though not all companies are going to succeed in implementing this technology. Not because of a lack of desire or need, but perhaps because of a lack of ability. For investors, this means sticking with companies that have the balance sheets and revenue growth to be able to integrate AI, while maintaining the health of the rest of their overall business.
On behalf of the Pyle Wealth Advisory team, have a wonderful weekend.
Andrew Pyle
CIBC Private Wealth consists of services provided by CIBC and certain of its subsidiaries, including CIBC Wood Gundy, a division of CIBC World Markets Inc. The CIBC logo and “CIBC Private Wealth” are trademarks of CIBC, used under license. “Wood Gundy” is a registered trademark of CIBC World Markets Inc.
Andrew Pyle is an Investment Advisor with CIBC Wood Gundy in Peterborough. The views of Andrew Pyle do not necessarily reflect those of CIBC World Markets Inc.
This information, including any opinion, is based on various sources believed to be reliable, but its accuracy cannot be guaranteed and is subject to change. CIBC and CIBC World Markets Inc., their affiliates, directors, officers and employees may buy, sell, or hold a position in securities of a company mentioned herein, its affiliates or subsidiaries, and may also perform financial advisory services, investment banking or other services for, or have lending or other credit relationships with the same. CIBC World Markets Inc. and its representatives will receive sales commissions and/or a spread between bid and ask prices if you purchase, sell or hold the securities referred to above. © CIBC World Markets Inc. 2024.