Navigating AI investments: diversification and valuation within the AI theme
As Q2 earnings announcements approach, it's crucial for investors to revisit growth expectations and re-evaluate concentration risk when considering exposure to the artificial intelligence (AI) theme. With the market breadth remaining quite narrow year-to-date, ensuring proper diversification and being mindful of valuations are key to building a well-rounded portfolio within the AI industry.
Avoid concentration risk
One standout performer in the AI industry is Nvidia, which has experienced explosive price appreciation since the latest earnings announcement when they forecasted $11 billion in sales for the upcoming quarter - over 50% higher than analyst expectations. Consequently, valuation metrics such as price-to-sales soared. If Nvidia seemed expensive six months ago, it has become an even more challenging ‘buy’ to justify for value-oriented investors.
However, relying heavily on Nvidia as an AI exposure can lead to concentration risk. Many investors seeking exposure to the AI megatrend opt for a broad tech approach, sometimes assuming an allocation to the Nasdaq-100 or MSCI Information Technology sector suffices for AI theme coverage, resulting in mega cap tech firms comprising a significant percentage of their portfolios. This concentrated and expensive approach may not be ideal for those looking to mitigate risk and diversify their investments while seeking to gain exposure to the AI theme. Luckily, there are other avenues to gain exposure to this exciting technology.
Source: WisdomTree, Factset as of 30/06/2023.
Historical performance is not an indication of future performance and any investments may go down in value.
Consider other AI exposures
Companies like MongoDB and Elastic are creating vector search databases, enabling improved efficiencies in querying large language models (LLMs) like OpenAI’s ChatGPT to be used for inference. These innovations power various use cases, such as similarity search, recommendation engines, Q&A systems, dynamic personalisation, and long-term memory for language models –enabling businesses to harness proprietary data with LLMs to increase their efficiency.
Firms like Autodesk, with its computer-aided-design software, stand to benefit from generative learning algorithms which can be used to increase design efficiency by generating potential solutions tuned to the input parameters of the user. This allows the user to augment their creative abilities, using AI-generated outputs to streamline the process from idea generation to finished product.
At WisdomTree, we have partnered with NASDAQ and the Consumer Technology Association (CTA) to develop a bespoke index that identifies and classifies AI-focussed companies across three classifications: Enablers, Enhancers, and Engagers. Through strategic allocations among these groups, investors can capture many different aspects of the AI theme while providing a pure AI investment exposure. For example, gaining exposure to AI computing chips within Enablers, generative AI solutions within Engagers, as well as innovative research from companies within Enhancers, allows for more comprehensive exposure to the breadth of the AI theme.
Be mindful of growth and valuations
When seeking exposure to high-growth companies, valuation often comes at a cost. The significant run-up in prices of large-cap tech names at the beginning of the year has led to higher valuations and increased concentration within some of the broad market baskets. Through the robust investment process of the NASDAQ CTA Artificial Intelligence Index, the strategy leverages CTA’s AI domain expertise to identify pure play AI companies with strong potential. From a fundamental perspective, this results in a blend of strong forward looking and historical growth with reasonable valuation statistics such as price-to-sales and provides one example of a strategy that may be more diversified than the traditional view of AI.
Growth and valuation metrics
Source: WisdomTree, Factset. Price to sales and 2Q23 consensus sales growth as of 30/06/2023. Historical growth rates as of 30/06/2023.
Historical performance is not an indication of future performance and any investments may go down in value.
Breaking things down further on the growth side, it becomes evident that sales and earnings volatility persists in the AI industry, as demonstrated by Nvidia. When a large company like Nvidia forecasts a 50% growth in sales, the market reaction is significant, and the single stock impact on the broader portfolio can take effect – both in terms of allocation and aggregate statistics like growth rates. In the moment, for the holder the price appreciation can be great, but as valuations steepen and another quarter nears, investors may anticipate impacts that the next earnings announcement may bring and question trimming the allocation to reduce risk.
Source: WisdomTree, Factset as of 30/06/2023.
Historical performance is not an indication of future performance and any investments may go down in value.
Source: WisdomTree, Factset as of 30/06/2023.
Historical performance is not an indication of future performance and any investments may go down in value.
As you can see above, the NASDAQ CTA Artificial Intelligence Index growth profile is much more diversified than more concentrated alternatives. Those with large weights to mega-cap tech stars may show reasonable numbers, but are much more reliant on these heavy allocations. The same can be said for their returns and volatility profile.
Investing in artificial intelligence appears to require a well-balanced portfolio that prioritises diversification across the AI ecosystem. By diversifying across different AI classifications and adopting a strategy that emphasises use cases as well as growth and valuation considerations, investors could mitigate concentration risk while gaining exposure to artificial intelligence in a portfolio.
Related blogs
+ Tap into the AI revolution with WisdomTree
+ Differentiating your AI exposure from the Nasdaq-100
+ Why the Nasdaq-100 may not capture the full growth potential of AI…and where to look instead