Guy Foster, Chief Strategist, discusses China’s new artificial intelligence (AI) lab and what this could mean for U.S. tech supremacy. Plus, Janet Mui, Head of Market Analysis, analyses developments related to President Trump’s trade tariff threats.
Last week was always likely to be a lively one, beginning, as it did, with the inauguration of the most unorthodox U.S. president in living history, Donald Trump. In contrast
to 2017, this version of a Trump presidency seems more organised and able to act faster and more decisively. Political events didn’t dramatically impact the market in either inauguration week, but the greater sense of calm may have been beneficial this time around.
When will Trump impose trade tariffs?
Most major markets performed well, but global investors will have experienced some movement in currencies. The U.S. dollar has lost ground, which probably reflects the fact that tariffs were not imposed on day one of the Trump presidency. U.S. tariffs on imports from Europe, for example, make European goods more expensive for U.S. consumers. As a result, those consumers buy fewer European goods, which means they need fewer euros. This is why, when a tariff is expected, the currency of the country imposing it tends to rise. The effect is to dilute the impact of the tariff, which is why there hasn’t been an improvement in the U.S. trade deficit since the return to the use of tariffs during President Trump’s first term.
Despite this, the threat of tariffs looms large over America’s trading partners. On his first day in office, President Trump claimed 25% tariffs would be imposed on Canada and Mexico by 1 February.
Perhaps acknowledging that tariffs won’t improve America’s trade balance, their use now seems to serve one of two purposes. In Trump’s first term, the threat of tariffs against Mexico and Canada achieved a renegotiation of the prevailing trade agreement. Tariffs against China were imposed and mostly remain, and the bilateral deficit between the U.S. and China has reduced since then. So,
tariffs can be threats, temporary measures, or structural impediments to trade, and can be used to reduce reliance on a trading partner.
A Fox news interview with President Trump saw him declare that he didn’t want to impose tariffs on China, affording some relief to a region bracing for them.
Trump was gracious in his praise for China’s President Xi Jinping and North Korea’s Supreme Leader Kim Jong Un, occasionally displaying his fondness for autocrats, which can seem somewhat disturbing. However, he was critical of Russian President Vladimir Putin, whom he’s previously seemed deferential to.
Addressing business leaders at the World Economic Forum in Davos, Switzerland, he appealed to Saudi Arabia to lower oil prices to end the war in Ukraine. A cut in oil prices would be achieved by shipping more oil, which is inconsistent with President Trump’s desire to see America pump more oil, and his desire for Europe to buy oil from the U.S. We have to be wary of placing too much emphasis on Trump’s wish list, which typically includes several mutually exclusive items.
President Trump announces major AI investment
The performance of the U.S. equity market has been broadening out since the election. Although 2024 was a strong year for the ‘Magnificent Seven’ technologyenabled mega-cap stocks (Amazon, Meta, Alphabet, Tesla, Nvidia, Microsoft and Apple), last month saw the group lose ground against the broader market.
That trend was interrupted when the Trump administration announced a major investment in AI infrastructure called Stargate. The Japanese firm Softbank, with support from Abu Dhabi’s investment fund MGX, will fund investment in AI infrastructure led by U.S. firm OpenAI, which it’s claimed could reach as much as $500bn.
The U.S. government isn’t funding the project but is clearing regulatory hurdles. Perhaps this announcement was the biggest surprise of the new administration.
Is U.S. technological supremacy under threat?
Hot on the heels of the Stargate announcement came the news that an upshot Chinese AI lab called DeepSeek appears to have produced an AI model that can rival the Western leaders such as OpenAI and Meta’s Llama. The new model employs a ‘mixture of experts’ system, which involves splitting tasks into distinct groups of similar sub-tasks. Separate models are trained for each of those groups, which echoes the way human tasks were made more efficient through specialisation during the 20th century.
The DeepSeek model was put together with a fraction of the expenditure that went into the current bestperforming models and caused ructions in a stock market that’s become very focused on how companies will monetise the benefits of AI.
DeepSeek’s approach implies that a similar quality of model can be achieved with a smaller or inferior set of chips – naturally, this would be very concerning for Nvidia, which is valued on the basis of high volumes of cutting-edge chips. The repercussions for the semiconductor value chain could be material.
A few notes of caution should be considered, though. DeepSeek hasn’t disclosed its training weights and some have questioned how confidently we can trust its development costs. It’s likely selling its product for lower margins than OpenAI. The DeepSeek model does fall short of OpenAI in some specialist processes.
Naturally, many global potential users would be nervous about using a Chinese model due to privacy or censorship concerns. But Western providers should be able to replicate DeepSeek’s open-source model. It might not matter too much for chip demand, though. In the 19th century, William Stanley Jevons observed that greater efficiency in steam engines caused coal demand to rise rather than decline because it drives greater adoption. The modern world is full of examples of humans doing more of something on the back of technology making it more efficient, rather than taking the savings. Washing clothes, refrigerating (first food, then buildings) or lighting the Las Vegas Sphere are some examples.
So, having more efficient models seems likely to accelerate and expand adoption rather than reduce the demand for chips.
The U.S. has risen to its position of technological supremacy through being open and competitive. One school of thought suggests that massive statefacilitated investment in infrastructure plays to
the observed AI scaling laws, whereby increasing parameters and computing capacity will derive better models. But Michael Porter, the authority on competitive advantage, notes that it stems from “pressure, challenge and adversity”.
DeepSeek claims to have achieved this breakthrough despite being limited by U.S. technology export restrictions, which raises questions about whether the United States’ industrial strategy is as foolproof as originally hoped.
How will President Trump impact immigration?
Immigration was a major policy priority, and the new administration took several steps to reduce immigration and prepare to deport unauthorised migrants. Much of this has been achieved by executive order, but legislation requiring the detention of undocumented immigrants suspected of crimes, even if they have not been charged, was passed by the Senate on Monday 20 January.
The legislation was able to pass so swiftly because it was supported by some Democrats. Ultimately, President Trump has discussed deporting millions of unauthorised migrants (estimates assume there are over 11 million in total). Doing so would be costly and logistically challenging. The greatest concern would be that a substantial reduction in workers risks causing inflation. The risk is particularly stark in sectors with high immigrant workforces, such as agriculture. Allowing immigration policy to drive a substantial increase in food prices due to a shortage of workers would be a substantial own goal, which is just one reason the administration is likely to fall short of its target to deport millions of unauthorised migrants.
How is the UK economy faring?
Away from the U.S., the UK’s public finance numbers were released, showing a higher-than-expected borrowing figure of £17.8 billion. However, this includes a one-off investment in military equipment. Excluding this, the borrowing figure would have been significantly lower.
The UK’s bond yields have been underperforming compared to other regions, and the market is pricing in a high probability of an interest rate cut in February.
Although the quality of the UK’s employment data has been poor, it seems there’s a clear trend emerging of slowing jobs growth, which may even have turned negative in the last two months (the data can be heavily revised, so it’s hard to say with certainty).
Retail is a sector that’s particularly under pressure. Retailers are large UK employers and so feel the brunt of the increase in employer’s National Insurance contributions, which was announced in last October’s Autumn Budget.
Sainsbury’s announced last week that it would cut 3,000 jobs and stagger a planned pay increase to reduce costs. Job losses wouldn’t just be on the shop floor; the supermarket plans to cut 20% of senior management positions.
This week: Earnings season, interest rate decisions and confirmation hearings
This week will be an important one for the fourth quarter earnings season. So far, earnings season has gone well, largely supported by the banks, with decent results from companies like Fastenal, Amphenol and Schwab. The current percentage of beats to misses (the number of companies that have beaten or missed analysts’ earnings expectations) is running slightly ahead of its historical average. This week will include the first of the ‘Magnificent Seven’ (Apple, Microsoft, Meta and Tesla) and broaden out the sectors reporting.
On Wednesday, the U.S. will be setting interest rates. So far, all forecasts are for rates to remain on hold, despite the new president saying he would “demand that interest rates drop immediately”. Analysis of the options market reveals a small but growing expectation that the Federal Reserve’s next move could be to increase rates, rather than cut them. To be clear, the majority view is still that the next move will be a rate cut, but uncertainty surrounding the new administration’s policies has led to a cohort of bond traders betting on an interest rate increase.
President Trump may be happier with Thursday’s European Central Bank meeting, in which interest rates are almost certain to be cut. These meetings come after Japan raised interest rates by 0.25% to 0.5% Friday morning. This was widely expected, so focus was on the Bank of Japan forecasts, which saw a sharp upgrade to expected inflation – so the risks to Japanese interest rates seem skewed to the upside.
There will be more action on the political front too, with confirmation hearings for the new U.S. Secretary of Commerce, Howard Lutnick, and one of the most controversial selections, Robert F Kennedy Jr. as Secretary of Health and Human Services.