Back
The financial industry has spent the better part of a decade automating everything it can. AI, using algorithms and natural language processing, ingests, analyses and interprets large volumes of data, can draft memos, flag fraud, and process earnings transcripts at a speed no team of analysts can match. While AI has embedded itself across the financial industry faster than most people expected, there is one place it keeps coming unstuck, and it is the one that matters most – actually trading.
Failing the test
AI’s performance as a trader is now being assessed by live experiments. In one of the more high-profile competitions, eight frontier models, including Anthropic’s Claude, Google’s Gemini, OpenAI’s ChatGPT, and xAI’s Grok, were each given $10,000 and let loose to trade in US tech stocks for two weeks. The combined portfolio lost roughly a third of its capital. Out of 32 individual result sets, a model only finished in profit six times. A separate exercise tracking eleven markets-related competitions found the median model profitable in only two of them. The outcome of these tests shows that most AI trading systems will lose money when tested in real conditions.
Understanding AI limitations
How AI fails matters. The models size their positions badly, and their entry and exit timing is off. When different models receive identical instructions, they produce meaningfully different decisions. One model, for example, placed 1,418 trades using a prompt that saw another model only place 158 trades. Where one model may be stubbornly long, another will be comfortably short, and a third drawn to leverage. None of the actions reflect a reading of the market – they instead reflect whatever tendency got baked into the model during training.
The models do not lack data or processing power, they lack the ability to answer the forward-thinking questions that the markets demand. A large language model learns from what has already happened. It finds patterns in historical text and produces outputs consistent with those patterns. That works for tasks with a defined structure, like summarising a document, flagging an anomaly, or predicting whether an earnings estimate will be revised up or down. It breaks down, however, when the question is open-ended and the answer depends on a context that has never existed in quite this form before.
Trading dynamics
Markets produce that kind of question constantly. While there are mechanical elements like pricing relationships, flow dynamics, and rate arithmetic (where systematic tools perform well and always have), these are not usually the factors which move prices. Markets are about positioning, about whether a trade is crowded, about how participants respond when something they believed turns out to be wrong. Reading into those nuances requires reasoning capability in an environment of genuine uncertainty, where information is incomplete and the variables keep shifting depending on what else is happening. A model trained on historical data cannot reliably do this, and scaling the model does not fix it.
There is also a problem with the method of evaluating model performance. Any model tested today on how it would have handled March 2020 already knows what March 2020 looked like. The outcome is in the training data. Real markets do not provide that luxury. The only valid test is a live one, and live tests have not been encouraging. Firms that have found something workable are using these tools for contained, specific jobs – earnings direction prediction, data processing, and research synthesis – where historical patterns remain a reliable guide. That is genuinely useful work. It is, however, not what most people mean when they talk about AI replacing the fund manager.
Trading needs a human touch
Markets are not fully reducible to algorithms. The science behind them is real and rewards good tools. But underneath the models and the data is a system driven by human decisions – made under stress, with incomplete information, shaped by incentives and feedback loops that behave differently every time. Working through that does not come from a formula. The machines handle the structured part well. The rest still needs a person in the room.
MARKET MOVES
Bonds
The United States (US) 10-year Treasury yield has extended its slide to a third consecutive session, dropping to 4.32% – its lowest level in roughly two weeks. Cheaper oil is doing the heavy lifting, easing inflation expectations and reducing the case for a more restrictive US Federal Reserve (Fed). A US proposal to Iran outlining a gradual Hormuz reopening and the lifting of the port blockade is sitting with Tehran awaiting a response. Markets have the Fed putting rate cuts on hold through to year-end, with the odds of a September or October cut sitting at around 20%. Today’s payrolls report will be the next domestic focal point.
German bund yields have slipped further below 3% as a potential US-Iran resolution pushed Brent just under $100/barrel following Wednesday’s sharp 10% drop. Two European Central Bank (ECB) hikes, totalling 50 basis points, are now priced in for the year, with a 75% probability of the first move arriving in June. ECB executive board member Piero Cipollone flagged the risk of rising rates earlier this week, echoing a broader hawkish tone from the governing council, though easing energy prices may yet temper that view depending on how Tehran responds.
United Kingdom (UK) gilt yields continued their decline toward 4.90% as the prospect of lower energy costs prompted markets to scale back Bank of England (BoE) tightening expectations to around 50 basis points, over two hikes, by year-end, down from three anticipated just weeks ago. Washington’s one-page Hormuz memorandum is the catalyst, with investors watching for Tehran’s next move. Domestically, UK voters were at the polls yesterday in Scottish, Welsh, and English council elections, which will prove a significant test for UK Prime Minister Keir Starmer amid Labour’s recent difficulties. Nigel Farage’s Reform Party looks likely to emerge as the big winners.
Equities
US futures are edging higher this morning after Thursday’s session saw the Dow, S&P 500, and Nasdaq fall 0.63%, 0.38%, and 0.13% respectively – nine of the 11 S&P sectors closed in the red, with materials, energy, and industrials leading losses. Fresh Hormuz skirmishes between US Navy destroyers and Iranian forces unsettled markets, though Trump maintains that the ceasefire remains intact. Today’s April payrolls print is the key data event, with consensus that it will reflect around 62,000 new jobs, a sharp step down from March’s 178,000.
Europe’s STOXX 50 fell 0.9% to 5,973 and the STOXX 600 lost 1% to 617, as corporate earnings disappointed alongside continued Hormuz uncertainty. Technology group Rheinmetall, dropped 7% on a first quarter profit miss and power utility Enel fell 2% post-results. Iran has yet to respond to Washington’s Hormuz memorandum, keeping the diplomatic picture unresolved.
The UK’s FTSE 100 lost more than 1% on Thursday as sliding crude dragged energy giants Shell and BP down 3.1% and 2.6% respectively. Energy services group Centrica warned on profit guidance, and defence and security company BAE Systems fell over 5% despite holding its sales outlook. Travel and leisure names including InterContinental Hotels, EasyJet, and IAG outperformed as the hope of lower fuel costs lifted sentiment in the sector.
Commodities
A week of heavy losses has partially unwound, with Brent back above $101/barrel after overnight military exchanges in the Strait of Hormuz rattled an already fragile diplomatic process. US destroyers transiting the waterway came under Iranian attack and responded with defensive strikes. Washington was quick to frame the action as contained rather than escalatory. A formal US proposal to end the nearly 10-week conflict and reopen the strait is in Tehran’s hands, with Pakistan serving as the conduit for a response expected within days. The International Energy Agency put the supply damage in stark terms this week stating that 14 million barrels per day of global output has been disrupted, with a slow and uneven recovery the most likely scenario even if a deal materialises.
Above $4,700 today, gold has steadied after a turbulent week, though the bigger picture remains sobering. The metal has shed more than 10% since the start of the war. The dynamic in this war has been counterintuitive. Usually, conflict will drive safe-haven buying but the energy price shock from this war has been severe enough to stoke inflation fears and push rate hike expectations higher, conditions that work against non-yielding assets, like gold.
Currencies
The US Dollar Index is holding above 98 today, drawing support from the Hormuz military flare-up that has kept geopolitical risk firmly in the price. Today’s April payrolls print will be the other key dollar driver. The greenback is on track to end the week broadly flat against most majors.
The euro is pushing toward $1.18/€ as dollar safe-haven demand softens on peace deal optimism. Two ECB hikes totalling 50 basis points are now priced in for the year, with a 75% probability of the first arriving in June. The ECB’s Cipollone flagged the risk of rate hikes this week, a view gaining traction across the governing council.
Sterling has climbed above $1.36/£, its highest level since mid-February, lifted by a softer dollar and receding rate hike expectations. Two BoE hikes are now priced in for the year, down from three. The UK is awaiting the results from yesterday’s council elections in Scotland, Wales and England, which may destabilise the labour government.
*Please note that all information is at the time of writing.
Key indicators:
GBP/USD: 1.3610
GBP/EUR: 1.1560
GBP/ZAR: 22.37
BRENT CRUDE: $101.24
GOLD: $4,709.90
Sources: Bloomberg, Financial Times, Flat Circle Blog, Intelligent Alpha, Investing.com, Nof1, NX1 Capital, Trading Economics and Refinitiv.
Written by Citadel Advisory Partner and Citadel Global Managing Director, Bianca Botes.
© Peregrine Wealth Ltd
This publication has been compiled for information purposes only and does not take into account the needs or circumstances of any person or constitute advice of any kind. It is not an offer to sell or an invitation to invest. The information and opinions in this publication have been recorded by Peregrine Wealth Ltd in good faith from sources believed to be reliable, but no representation or warranty, expressed or implied, is made as to their accuracy, completeness or correctness. Peregrine Wealth Ltd accepts no liability whatsoever for any direct, indirect or consequential loss arising from the use of this publication or its contents. Peregrine Wealth Ltd (registration number 39538) is licensed by the Guernsey Financial Services Commission.