The Big Question: Bubble fears - too far or not far enough?

00:00:25:14 - 00:00:27:24 Fears of a bubble have intensified this year 00:00:27:24 - 00:00:31:18 amid skyrocketing valuations of AI and tech stocks. 00:00:31:18 - 00:00:37:03 Major indices like the S&P 500 and Nasdaq have hit repeated all time highs 00:00:37:03 - 00:00:42:00 largely propelled by a handful of mega-cap growth stocks with an AI theme. 00:00:42:00 - 00:00:46:03 The major US and global indices are capitalisation weighted 00:00:46:03 - 00:00:49:23 which means that they own more of the bigger companies than the smaller ones. 00:00:49:23 - 00:00:53:10 and as a result these broad indices have significant 00:00:53:10 - 00:00:56:14 concentration risk in just a few stocks. 00:00:56:14 - 00:01:02:23 For the S&P 500, nearly 30% of the index is allocated to just five companies 00:01:02:23 - 00:01:06:12 and nearly 40% is allocated to just ten. 00:01:06:12 - 00:01:09:10 If the reality of AI does not live up to the hype 00:01:09:10 - 00:01:12:09 and some of those giant tech stocks fall in value, 00:01:12:09 - 00:01:16:24 they can have a major negative impact on the return of the entire index. 00:01:18:15 - 00:01:22:08 Well we've just passed the three year anniversary of the release of ChatGPT 00:01:22:08 - 00:01:25:16 and in that time every business associated 00:01:25:16 - 00:01:29:07 with building out the AI infrastructure has gone in one direction, 00:01:29:07 - 00:01:30:04 and that is up. 00:01:30:04 - 00:01:31:22 After several years of that 00:01:31:22 - 00:01:35:17 people are fairly asking how far can this run and how big can this be? 00:01:35:17 - 00:01:39:20 A common theme you see from prior market bubbles 00:01:39:20 - 00:01:41:14 is a narrowing of the market. 00:01:41:14 - 00:01:46:18 So narrower and narrower, smaller number of companies driving the returns 00:01:46:18 - 00:01:52:04 and the returns increasingly being driven by a single theme, in this case AI. 00:01:52:04 - 00:01:55:00 So it's a fair question at this point, three years in, 00:01:55:00 - 00:01:58:15 where most of the stock market has not done very much 00:01:58:15 - 00:02:01:21 but a narrow slice of the market has done incredibly well. 00:02:01:21 - 00:02:03:22 It's a fair question to ask are we in a bubble? 00:02:10:16 - 00:02:13:13 AI is an exciting new technology that should deliver 00:02:13:13 - 00:02:17:16 incredible productivity enhancements across the entire economy. 00:02:17:16 - 00:02:21:22 However, the investment case for most AI stocks is very questionable. 00:02:21:22 - 00:02:24:24 First, AI is massively capital intensive, 00:02:24:24 - 00:02:28:18 requiring hundreds of billions of dollars of investment in data centres. 00:02:28:18 - 00:02:30:14 Compared to that enormous spending 00:02:30:14 - 00:02:35:10 there is comparatively little revenue being generated by the AI service providers 00:02:35:10 - 00:02:39:04 Thus, AI companies are burning through cash at an incredible rate. 00:02:39:04 - 00:02:43:09 Furthermore, investors in AI companies have to pay high valuations 00:02:43:09 - 00:02:46:16 for the right to own these cash burning businesses. 00:02:46:16 - 00:02:49:05 It is more than possible that AI could be 00:02:49:05 - 00:02:52:17 a revolutionary technology and a smashing success 00:02:52:17 - 00:02:57:07 and still AI companies today could be terrible investments. 00:02:57:07 - 00:03:01:22 In fact, that's exactly what happened during the tech bubble of 1999. 00:03:01:22 - 00:03:06:03 The internet was a revolutionary technology and a smashing success 00:03:06:03 - 00:03:09:00 and yet all the internet darlings from the tech bubble 00:03:09:00 - 00:03:13:23 were disastrous investments for the next 5 to 10 years after that bubble burst. 00:03:15:17 - 00:03:18:24 We believe the long-term investment case for AI remains very strong 00:03:18:24 - 00:03:23:07 despite growing questions around whether or not the markets are currently in a bubble. 00:03:23:07 - 00:03:26:22 Ultimately, when there's a paradigm shift for any new technology, 00:03:26:22 - 00:03:31:19 you will see winners, losers, over-building, over-investment. 00:03:31:19 - 00:03:35:07 Our view though is that if the underlying technology 00:03:35:07 - 00:03:40:18 is something that brings real efficiency and real benefit to the end user 00:03:40:18 - 00:03:43:23 then the companies that are enabling that paradigm shift 00:03:43:23 - 00:03:46:08 should do quite well over time. 00:03:46:08 - 00:03:50:12 Our belief is that those are the types of businesses that we are invested in 00:03:50:12 - 00:03:53:15 and as long as the underlying value creation is there 00:03:53:15 - 00:03:56:15 then the underlying investment case remains. 00:04:02:07 - 00:04:07:14 China potentially poses a significant and growing threat to US AI dominance. 00:04:07:14 - 00:04:11:18 but it does not seem like an immediate or existential threat yet. 00:04:11:18 - 00:04:15:07 Currently, the US holds the lead in AI innovation. 00:04:15:07 - 00:04:18:15 China seems to be excelling at cost efficiency. 00:04:18:15 - 00:04:21:23 The US benefits from superior private investment, 00:04:21:23 - 00:04:25:16 top-tier compute access and breakthrough research. 00:04:25:16 - 00:04:28:18 But export controls have accelerated China's efforts, 00:04:28:18 - 00:04:31:14 and they have narrowed the gap in model performance. 00:04:31:14 - 00:04:35:08 As with any technology, the situation is very fluid 00:04:35:08 - 00:04:37:18 and could change with each new breakthrough in AI. 00:04:39:06 - 00:04:43:10 China and the United States are clearly engaged in a battle for economic supremacy 00:04:43:10 - 00:04:46:03 and AI is simply the next front in that war. 00:04:46:03 - 00:04:49:03 Both countries view it as a matter of national security. 00:04:49:03 - 00:04:53:02 Our view is a little bit more nuanced and that we don't see it as a zero sum game. 00:04:53:02 - 00:04:55:22 Progress in the United States and progress in China does not 00:04:55:22 - 00:04:59:07 necessarily have to have a negative impact on the other. 00:04:59:07 - 00:05:03:07 So ultimately, we think that AI will go down a similar path 00:05:03:07 - 00:05:07:18 to many other areas of tech which is what we would call a two stack world, 00:05:07:18 - 00:05:10:06 a Chinese version and an American version. 00:05:10:06 - 00:05:14:20 One issue that the Chinese face that is structural and is long term in nature 00:05:14:20 - 00:05:18:14 is that some of the underlying capital equipment that is required 00:05:18:14 - 00:05:22:16 to actually manufacture advanced semiconductors 00:05:22:16 - 00:05:25:22 only comes from Western based businesses, 00:05:25:22 - 00:05:30:13 so that will continue to be a structural headwind to Chinese AI development. 00:05:30:13 - 00:05:32:02 But it's one that they've shown 00:05:32:02 - 00:05:34:06 that they're quite adept at overcoming with innovation. 00:05:41:23 - 00:05:46:09 At Lyrical we look to invest in stocks that have three key traits, 00:05:46:09 - 00:05:49:11 value, quality and analysability. 00:05:49:11 - 00:05:54:15 We do not hold any of the hyped AI stocks because they do not possess these traits. 00:05:54:15 - 00:05:58:06 First, we believe AI stocks are incredibly expensive 00:05:58:06 - 00:06:02:04 and so they do not meet our valuation criteria by a wide margin. 00:06:02:04 - 00:06:08:06 Second, AI companies are incredibly capital intensive and do not meet our quality criteria 00:06:08:06 - 00:06:11:03 that prefers capital light business models. 00:06:11:03 - 00:06:17:12 Lastly, AI technology is rapidly evolving and difficult, if not impossible, to predict. 00:06:17:12 - 00:06:21:12 and so AI stocks do not meet our analysability criteria. 00:06:21:12 - 00:06:25:08 That said, while we do not own any of the hyped AI stocks, we do own 00:06:25:08 - 00:06:29:02 many other stocks that are benefiting from AI spending, 00:06:29:02 - 00:06:32:08 including power producers and data centre suppliers. 00:06:32:08 - 00:06:33:13 We also own companies 00:06:33:13 - 00:06:37:07 that are benefiting from AI technology to improve their businesses. 00:06:37:07 - 00:06:39:20 If AI stocks truly are in a bubble, 00:06:39:20 - 00:06:44:09 our portfolio, we believe, is well positioned to avoid the painful losses 00:06:44:09 - 00:06:46:18 that would occur if that bubble were to pop. 00:06:48:02 - 00:06:52:10 Given the duration and the narrowness of the bull market that we’re in 00:06:52:10 - 00:06:58:03 It’s a fair question to ask how are we managing the tech-related risks in our portfolio? 00:06:58:03 - 00:07:01:11 And really we look at those kinds of risks through two lenses. 00:07:01:11 - 00:07:05:08 One is a criteria lens at the individual company level. 00:07:05:08 - 00:07:07:21 We, by design, want to own businesses 00:07:07:21 - 00:07:11:00 that we believe are leaders at what they do 00:07:11:00 - 00:07:15:24 and that we believe have durable, competitive moats around the franchises. 00:07:15:24 - 00:07:20:07 We think that that is one of the most powerful risk management tools at our disposal. 00:07:20:07 - 00:07:24:04 Secondly, the second lens is portfolio construction 00:07:24:04 - 00:07:27:15 and that's about managing the aggregate 00:07:27:15 - 00:07:31:03 weight and exposure to different industries and in this case, 00:07:31:03 - 00:07:35:14 technology and balancing out that exposure appropriately with exposure 00:07:35:14 - 00:07:40:00 in other areas such as consumer, life sciences and industrial businesses.