19 March 2025
For years, we were told that technology would make financial markets more efficient. AI would predict trends with machine-like precision, high-frequency trading (HFT) would smooth out price discrepancies, and big data would eliminate the uncertainty that fuels irrational investing. And yet markets are as chaotic as ever. How does that work?
It’s time to ask a different question: What if these tools don’t eliminate market anomalies but instead magnify them?
The Efficient Market Hypothesis (EMH) is the academic equivalent of a comforting bedtime story. It tells us that markets process all available information instantly, making it impossible to consistently outperform them. Sure, AI and HFT make trading faster and crunch more data, but that doesn’t mean they create a fairer or more predictable market. If anything, they can supercharge existing inefficiencies, distorting prices faster and with greater intensity than ever before.
Consider this: If markets were truly efficient, why do flash crashes still happen? Why do meme stocks defy fundamental valuations? Why do hedge funds with teams of PhDs still get burned?
The answer is simple - machines don’t erase irrationality; they accelerate it.
AI is supposed to remove human emotion from investing. The problem? AI learns from us and we’re not rational creatures. Instead of eliminating biases, AI models often reinforce them.
Take sentiment analysis. Hedge funds and institutions now use AI to scan social media, news reports, and earnings calls for signals. But what happens when everyone uses the same model? The market overreacts to every scrap of information.
A single misleading tweet, a misinterpreted earnings statement, or a fabricated rumour can send stocks soaring or crashing in minutes. AI doesn’t just reflect human biases - it magnifies them at unprecedented speed, reinforcing flawed decision-making on a global scale.
The Counterpoint: AI also plays a critical role in identifying investment opportunities that human traders might overlook. By processing vast amounts of data, AI-driven strategies can detect early signals of market trends, helping investors make more informed decisions. In areas like risk assessment, fraud detection, and portfolio optimisation, AI has undeniably improved the investment landscape.
HFT was supposed to eliminate price anomalies by arbitraging away inefficiencies in real-time. Instead, it has created an entirely new class of problems.
Let’s talk about the 2010 Flash Crash - a trillion-dollar market wipe-out which happened in minutes. Algorithms trading at lightning speed started feeding off each other's signals, triggering a death spiral. No humans caused this and no human could stop it in time. HFT traders might profit off short-term price gaps, but when things go wrong, they can go wrong spectacularly fast.
HFT doesn’t smooth out markets - it exaggerates volatility, creating sharp, short-lived price swings that make it even harder for long-term investors to navigate the chaos.
The Counterpoint: While HFT has been blamed for market instability, it has also provided significant benefits. By increasing liquidity and narrowing bid-ask spreads, HFT makes markets more efficient in normal conditions. The ability to quickly match buyers and sellers also reduces transaction costs, particularly for retail investors. Many argue that without HFT, markets would be even less efficient and more prone to large price discrepancies.
Access to more data should make investing more precise. However, the edge disappears when everyone has the same data, and information overload becomes the real problem.
Markets are no longer reacting to fundamentals but to each other. If AI models all digest the same dataset and come to the same conclusions, does that make the market efficient? No. It makes the market fragile; one unexpected event, one outlier that doesn’t fit the model and everything collapses.
Investors today aren’t necessarily making better decisions. They’re just making faster, crowd-driven ones and when that crowd moves in the wrong direction? Catastrophe follows.
The Counterpoint: On the other hand, big data has transformed investment analysis. AI-driven quant funds have outperformed traditional models by leveraging alternative data sources - everything from satellite imagery to credit card transactions - to predict company performance more accurately. While data-driven trading does introduce risks, it has also helped investors’ process market signals more efficiently than ever before.
AI, HFT, and big data were supposed to be antidotes to market inefficiencies, but they’ve become multipliers of them instead. They amplify the irrational, the reactionary, and the unforeseen.
Markets don’t just behave unpredictably with more technology; they can fail because of it, faster, harder, and with wider consequences. This isn’t to say that AI, HFT, and big data never improve efficiency - at times, they do. But it’s dangerous and naive to assume that more technology automatically equals more efficient markets. In many cases, these tools create new risks that investors are only beginning to understand.
So, the next time you hear someone say, “Markets are more efficient now”, ask them this: Are they more efficient, or just more unpredictable?
The future of finance isn’t about eliminating anomalies. It’s about outlasting them and the smartest investors will be the ones who recognise this first.
Back to News
+44 (0)20 7048 9400
Email Us
Client login
