Betting against "AI".

Linton

aroominyork said:

masonic said:

Eyeful said:

gm0:
So your using the "Looking for the pin in the haystack" approach.
Many bight fund managers with large recourses behind them, over many years use that method.
After changes, they do not consistently beat a simple major global index.
So I wish you the best of luck, when trying to do so yourself.

If you are just concerned with achieving the best possible long term return, come what may, then I'd tend to agree. But that isn't always the objective. I don't think it can be said that those trying to preserve capital in priority to growing it cannot consistently beat a simple major global index in metrics such as peak-to-trough losses, volatility, etc. Adding bonds to serve this purpose is not a panacea. Gold is not even remotely tempting at its current price. So other means of hedging risks, often at no additional cost in terms of fees, is worth exploring at least.

That’s fair enough… but how do you generate an X axis of performance against a Y axis of other metrics to assess whether you “consistently beat a simple major global index”? Is 10% less volatility at the cost of 2% worse performance a win?

We recently discussed US smaller companies and mitigating reliance on the Mag7 (which I recognise is not equivalent to AI). I am increasingly shaping my portfolio so the tilts from a global index will be nice if they profit me but, if they do not, I will not feel I did anything silly. That is why, masonic, although like you I am underweight US I decided not to buy smaller companies because that felt like a double tilt and, if the Mag7 continue to outperform, that could trigger the ‘silly’ button.

From looking at my various real and potential portfolios there is no obvious correlation between amount of deviation from a global index allocation and long term performance as long as you are well diversified. All well diversified portfolios appear to behave in a very similar way.  Tilting can however affect volatility in the shorter term and hence one's ability to sleep at night.  It is also useful for improving diversification metrics.

Perhaps it would be worth your while experimenting with trustnet charting and morningstar..

ChilliBob

Whilst I don't share your Ai views entirely I have been considering similar very recent (as I'm sure a lot of investors have). I was toying with an equal weight s&p tracker combined with euro/Asia trackers. Not something I've done yet, but mulling over.

I'd want to do it unwrapped using OEICs though, and there seems to be limited choices there on II. I may wait till ISA allowances come up again in April and do it then with etfs instead I guess. 

Linton

coyrls said:

I asked Chat GPT: "How can I invest in a way that will avoid exposure to AI companies?"  Its answer was:

1. Avoid AI-Heavy Sectors

• Stay Away from Tech: Avoid investing in technology sectors, as many companies in this space heavily integrate AI.
• Watch for Cross-Sector AI Use: Be mindful that many non-tech industries like finance, healthcare, and retail are also increasingly adopting AI.

---

2. Focus on AI-Free Sectors

Look for industries that are traditionally less affected by AI advancements:

• Agriculture (Traditional): Small-scale farming or organic farming operations that rely on human labor and natural methods.
• Construction: Smaller, locally-focused construction companies.
• Hospitality (Non-Tech-Driven): Small businesses like bed-and-breakfasts or boutique hotels.
• Arts and Crafts: Businesses that focus on handmade goods or artisan production.

---

3. Invest in Actively Managed Funds

• Work with fund managers who can actively screen and exclude AI-involved companies.
• Look for thematic ETFs or mutual funds that focus on non-tech industries and explicitly avoid AI exposure.

---

4. Screen for Non-AI Investments

Use tools like ESG (Environmental, Social, and Governance) filters or specialized platforms to exclude companies investing in AI:

• Ethical Investing Platforms: Some platforms allow you to filter companies or funds based on specific exclusions like AI.
• Custom Screening: Research the holdings of funds or ETFs to identify and exclude those with AI exposure.

---

5. Invest in Local or Private Ventures

• Consider local businesses or private equity opportunities in sectors like local retail, food services, or manual trades.
• Direct Investments: Invest in non-AI startups or co-ops that align with your values.

---

6. Use ESG and Exclusionary Criteria

• Look for funds that prioritize social impact, local economies, and manual labor, as these are less likely to involve AI.
• Focus on SRI (Socially Responsible Investing) strategies, and specify your avoidance of AI.

---

7. Monitor and Research

• Keep up with company reports and developments to ensure they aren't pivoting toward AI.
• Use online tools like MSCI ESG Ratings or Sustainalytics to track AI involvement.

Deepseek, China's recently announced AI offers similar advice though it seems to me rather more helpful.  It is much faster than chatgpt:

Investing in a way that avoids exposure to AI companies requires a deliberate approach to portfolio construction. Here are some steps you can take:

### 1. **Screen Out AI-Related Companies**

   - **Avoid Tech and AI-Focused Sectors**: Many AI companies are concentrated in the technology sector, particularly in areas like software, semiconductors, and cloud computing. Avoid investing in companies that develop AI technologies or heavily rely on AI for their operations.

   - **Use ESG (Environmental, Social, Governance) Screens**: Some ESG funds exclude companies involved in controversial technologies, including AI. Look for funds that explicitly avoid AI or advanced technology companies.

   - **Research Individual Companies**: If you're investing in individual stocks, research companies to ensure they are not involved in AI development or applications.

### 2. **Invest in Non-Tech Sectors**

   - Focus on industries that are less likely to be involved in AI, such as:

     - **Utilities**: Companies providing essential services like water, electricity, and gas.

     - **Consumer Staples**: Companies producing everyday goods like food, beverages, and household products.

     - **Healthcare (Non-Tech)**: Traditional healthcare providers, pharmaceuticals, and medical equipment companies that do not rely on AI.

     - **Real Estate**: Real estate investment trusts (REITs) or property management companies.

     - **Agriculture**: Companies involved in farming, food production, or agricultural equipment.

### 3. **Choose Thematic or Sector-Specific Funds**

   - Invest in mutual funds or ETFs that focus on sectors unrelated to AI. For example:

     - **Utilities ETFs**: Funds that track utility companies.

     - **Consumer Staples ETFs**: Funds focused on companies producing essential goods.

     - **Healthcare ETFs**: Funds that exclude tech-driven healthcare companies.

   - Avoid funds with significant exposure to technology or innovation themes.

### 4. **Consider Value or Dividend-Focused Investments**

   - Value stocks and dividend-paying companies are often found in traditional industries like manufacturing, energy, and financials, which are less likely to be involved in AI.

   - Look for value-oriented or dividend-focused ETFs and mutual funds.

### 5. **Avoid Innovation and Growth Funds**

   - Many innovation and growth funds invest heavily in AI and technology companies. Avoid funds with terms like "innovation," "disruption," "technology," or "growth" in their names or descriptions.

### 6. **Invest in International or Emerging Markets**

   - Some international or emerging market companies may have less exposure to AI, especially in regions where technology adoption is slower. However, be cautious, as AI is becoming increasingly global.

### 7. **Work with a Financial Advisor**

   - A financial advisor can help you build a customized portfolio that aligns with your goal of avoiding AI exposure. They can also help you identify funds and companies that meet your criteria.

### 8. **Monitor Your Portfolio Regularly**

   - Even if you initially avoid AI companies, industries evolve, and companies may adopt AI over time. Regularly review your investments to ensure they remain aligned with your goals.

### Example Portfolio Allocation:

- **Utilities**: 20%

- **Consumer Staples**: 20%

- **Healthcare (Non-Tech)**: 15%

- **Real Estate**: 15%

- **Energy**: 10%

- **Financials**: 10%

- **Agriculture**: 10%

By focusing on traditional industries and avoiding tech-heavy sectors, you can reduce your exposure to AI companies. However, keep in mind that AI is becoming increasingly pervasive, so complete avoidance may be challenging over the long term.

gravel_2

Without having to be too specific with what you target possibly a global equal weight (these seem pretty rare, seeing only MWEP which seems to have very little scale or available information and available in $ only).

Alternatively a world ex-USA (also rare, possibly only EXUS) then some combination you are comfortable with between normal SPX, Russell or Nasdaq100 funds paired with an equivalent equal weight fund to take the tech focus off. You also then have control over US level of representation vs ex-USA.

All this is missing is developing world and low caps.

valiant24

aroominyork said:

On the point of a value tilt, and given the high level of crossover between AI and the Mag7, it's worth noting that the value-tilting PSRW includes Apple, Microsoft and Amazon in its top ten holdings. XDEV, which has a stronger value tilt, does not includes any of the Mag7 in its top ten although does have other tech firms (I have no idea now much Cisco, Qualcomm, Intel etc. are engaged in AI) .

Thanks XDEV looks quite interesting, and diversified.  Top 10 holding doesn't seem to many companies with AI aspirations, apart from IBM which IMHO is a complete dog.  I'll take a closer look.

valiant24

prowla said:

Calling the current pattern matching and rule-based systems "AI" is a bit like calling cruise control a driverless car.

I'm wondering when the bubble will burst.

prowla said:

Calling the current pattern matching and rule-based systems "AI" is a bit like calling cruise control a driverless car.

I'm wondering when the bubble will burst.

prowla said:

Calling the current pattern matching and rule-based systems "AI" is a bit like calling cruise control a driverless car.

I'm wondering when the bubble will burst.

I said I wasn't going to get into a debate on AI!   But this is exactly my view.

valiant24

Eyeful said:

gm0:
So your using the "Looking for the pin in the haystack" approach.
Many bight fund managers with large recourses behind them, over many years use that method.
After changes, they do not consistently beat a simple major global index.
So I wish you the best of luck, when trying to do so yourself.

Absolutely fair comment.

AlanP_2

I prefer not to bet against or for anything as I have no idea what will work and what won't across the global economy.

What I do is set a target for what I want our overall portfolio to return (CPI + 3%) over a rolling 3 year period. In my mind this will ensure that we can withdraw 3-5% a year for the next 30 years (hopefully) without running out. For context this is to supplement good CPI linked DB pensions and 2 full SPs so not needed for essentials.

Whether that beats or loses out to a 100% equity, 10% equity or whatever portfolio is irrelevant to me.

Linton

I don’t think the issue is betting for or against AI. Rather that the % allocation to the mag7 given by global index trackers is far too high for safety and is crowding out other options.

I believe that AI, like databases or even the wider internet now, will be of major importance. But  I am not going to “bet” on the current early developers making serious profits from it. There are plenty of other things to also invest in.

prowla

valiant24 said:

prowla said:

Calling the current pattern matching and rule-based systems "AI" is a bit like calling cruise control a driverless car.

I'm wondering when the bubble will burst.

prowla said:

Calling the current pattern matching and rule-based systems "AI" is a bit like calling cruise control a driverless car.

I'm wondering when the bubble will burst.

prowla said:

Calling the current pattern matching and rule-based systems "AI" is a bit like calling cruise control a driverless car.

I'm wondering when the bubble will burst.

I said I wasn't going to get into a debate on AI!   But this is exactly my view.

Sure - my point re. investing was the "I wonder when the bubble will burst".

Then again, the tech industry does like its booms & busts, so who am I to say!