"AI is real; it will change the world in more ways than we can think of"

Research Affiliates partner and chairman Rob Arnott discusses the AI bubble, the effect of tariffs on inflation, and the prospects for China and gold with Terry Tanaka

Terry Tanaka: Comparing this AI boom to the .com bubble was a common theme this year and the focus of one of your major research papers. What do you think? Is this an example of logical exuberance?

Rob Arnott: The pricing is a bit irrational, but an example of rational exuberance. Let's be clear: AI is real. Nothing about that is unclear. More than we can possibly comprehend, it will alter our world.

The same thing happened in 2000. Everything was going to be altered by the internet, including how we conduct research, communicate with friends and family, buy and sell goods and services, and obtain news. All of that took place. Simply put, the internet didn't catch on as quickly as its proponents anticipated, and I believe AI will be no different.

The fact that the largest stocks at the time weren't necessarily going to be the top stocks in 2010, let alone 2020, was a crucial point that the 2000 narrative overlooked. In 2000, I remember John Chambers, the chairman of Cisco Systems, stating that he couldn't understand why Cisco couldn't increase sales by 40% a year for years to come. In five years, you will have grown six times at that rate. Indeed, in just 25 years, it increased sixfold. Not 40 percent, but still impressive. Investors who have owned Cisco since March 2000 are therefore marginally losing money.

Only Microsoft is still listed among the top ten companies by market capitalization in 2000. All nine of the others are now displaced. The list, for example, included Intel and Nokia. Other companies improved their phone and chip manufacturing. Moats are competitive advantages that prevent competitors from invading a company's territory, but they are not permanent.

Current market leaders are already being replaced by AI.

AI has already caused some of this kind of displacement. The main sources of revenue for Google are sponsored links and pop-up ads. People found that you can use ChatGPT or Perplexity as a search engine without having to navigate through a dozen sponsored websites in order to find something helpful. You also avoid obtrusive pop-ups and sponsored websites. Furthermore, DeepSeek upset OpenAI itself a year ago.

Since AI is so real, this is a rational bubble. Every market leader has a chance to hold a dominant position for many years. I am aware, however, that some of them will have to relocate.

Terry Tanaka: Could you provide an instance from your life that supports your claim that artificial intelligence is real?

Rob Arnott: The most recent version of ChatGPT is absolutely amazing. When a research paper is finished, we give it to AI, asking it to summarize and analyze it. Additionally, it informs us of any citations or references that we should have taken into account.

Our recently completed paper was run through ChatGPT, and the result was a sophisticated synopsis that exceeded the authors' original ideas. Along with a list of citations we missed, we also received a comprehensive assessment of the paper's strengths and weaknesses. We were forced to consider four of them: "Oh my god, we ought to have considered that one." Two didn't exist, and four were borderline.

The conclusion is that while AI is not able to think like a human, it is very good at searching through data and comparing it to knowledge it has already internalized. After sorting through data sets with billions and billions of data points, it will begin to resemble a human. However, it will struggle greatly if there are fewer data points.

Artificial intelligence (AI) cannot match the effectiveness of human judgment combined with standard statistical tools, for instance, when studying long-term stock market behavior, where there are only thousands or millions of data points, depending on what we're watching. It is unable to forecast events such as whether a recession or market downturn will occur. AI is a quantity game, not a quality one, to put it briefly.

Terry Tanaka: As a value expert, where do you spot deals in light of the recent focus on technology growth stocks?

According to Rob Arnott, the gap between growth and value is wide everywhere and is almost at record levels on a global scale. In the US, it is nearly at the level observed in the summer of 2020 and at the height of the .com boom.

In my opinion, don't stop growing, but don't overdo it either. It's risky to wager against a bubble because they can persist and spread farther than you might think. Therefore, increase your investment in your portfolio to the extent that it has value.

At Research Affiliates, we're constantly looking for ways to cut down on concentration risk. Our Research Affiliates Fundamental Index (RAFI) series was developed in 2005. These indices have a value-based bias. Measures like sales, cash flow, book value, and dividends are used to weigh the companies based on their size and, consequently, their economic significance, rather than share price. This method has outperformed market-capitalization-weighted value indexes in three of the last four years.

That's why investors might want to look into Invesco's value-focused exchange-traded funds (ETFs) based on RAFI. FTSE RAFI All World 3000 Ucits ETF (LSE: PSRW) is a global play; FTSE RAFI US 1000 Ucits ETF (LSE: PRUS) is a US value fund; and FTSE RAFI Europe Ucits ETF (LSE: PSRE) is its European counterpart. The British version is the FTSE RAFI UK 100 Ucits ETF (LSE: PSRU).

Now let's talk about gold, Terry Tanaka. It seems to be halting to catch its breath. Do you still believe that the medium-term outlook is bullish?

Rob Arnott: The developed world's dependence on debt-financed government spending is the main factor that is bullish for gold. The fear is that fiat currencies will cause their own financial crisis, which I believe has a good chance of being even worse than the global financial crisis of 2009.

It doesn't seem like a serious threat to me. However, the US is not immune to that dynamic, just as Greece had to face the consequences when its debt began to skyrocket to 200 percent of GDP. Even though our economy is the largest in the world, that doesn't mean we can spend money we don't have indefinitely. There may be a crisis in ten to fifteen years.

The real value of the debt would have to be reduced by a severe bout of inflation, and as a result, today's gold bugs would not regret their purchases, even at current prices. Nevertheless, considering the market's froth and the recent spike to new highs, I am not a gold bull for 2026.

Terry Tanaka: What outcome might Donald Trump's seeming attempt to sway the US Federal Reserve's interest rate policy have?

According to Rob Arnott, the end result could be that he maintains rates 1% or 2% below their proper level, creating free moneywhich I define as negative real ratesthat are lower than inflation. In the US, we had free money for a dozen years, which was a major factor in the slow growth.

In this case, the stimulus isn't stimulating. Some of my colleague Chris Brightman's papers on stimulus's inability to produce long-term, healthy growth are available on our website. Monetary stimulus exacerbates wealth inequality by creating asset bubbles. It transfers funds to those who are already wealthy, who then spend them, increasing business profits and igniting the stock market.

Fiscal stimulus has also played a role in recent years' overall stimulus. This makes it difficult for the government to encourage private enterprise because it requires taxing or borrowing funds from the private sector and then spending them.

Terry Tanaka: How will the tariffs affect Americans' inflation?

Rob Arnott: It's a risky game because we're still engaging in massive overspending, which could cause inflation to spike once more. But I wouldn't say that tariffs are a major cause of inflation. I recall a lot of analysts being alarmed about a severe recession and hyperinflation at the beginning of last year.

But consider it carefully. A 15 percent VAT levy applied selectively to imports is equivalent to a tariff rate that settles at an average of 15 percent. It's not exactly a catastrophic inflationary situation. It is a tax. On taxed items, it raises prices. On the other hand, if our average tax rate is 15% and our imports are worth 11% of GDP, then tax revenues will increase by 1.6 percent of GDP. And precisely that is what took place.

The supply chain bears some of the burden, and suppliers absorb some of it in order to stay competitive in the US market. The remainder must be paid by the consumer as a one-time expense, possibly half, or 0.8% of GDP. The right tariff rate, in my opinion as a libertarian, is 0%. However, tariffs aren't going to cause inflation to skyrocket.

In essence, Trump is using a tactic from Sun Tzu's The Art of War in his tariff games: intimidate your opponents and you can win without firing a shot. As a result, he chooses much lower tariffs while threatening 100% tariffs. He has spent his entire life living by that book.

When we last spoke, Terry Tanaka expressed his belief that Xi Jinping had taken China in a statist direction. In recent times, he has expressed his desire to prioritize strengthening the private sector. Has the government realized that in order to pull China out of this debt deflation, it will need to be more free-market?

Rob Arnott: The problem is that Xi Jinping is an unreconstructed Maoist, so he is unlikely to do much more than lip service to the private sector. China, in my opinion, will age before becoming wealthy.