In Money Matters

Matthew Weller

Bot vs. bot: we really don’t know what drives the markets and hope for the best

Bot vs. bot: we really don’t know what drives the markets and hope for the best

You know that unnerving feeling when the market moves in ways that just don’t add up? When a obscure stock suddenly rockets for no clear reason, or a flash crash ripples through the system? Often, the culprit isn’t in a boardroom or an economic report, but in the chaotic chatter of social media – now amplified and distorted by AI. This isn’t just noise; it’s a direct threat to the algorithms that drive most modern trading. What follows is a breakdown of a stark warning from the IMF about this very vulnerability, and why investors everywhere should be paying close attention.

In its most recent analysis of financial stability, the International Monetary Fund has issued a direct call to national regulators, urging them to actively surveil social media platforms for content capable of distorting AI-powered trading algorithms. The report underscores a critical vulnerability: with research from the London School of Economics indicating that 60% to 70% of global trading volume is now executed by automated systems, the market’s susceptibility to digitally propagated signals is unprecedented. The IMF explicitly warns that the very algorithms designed to capitalise on market sentiment and news flow can be weaponised, leading to cascading failures such as flash crashes, endemic algorithmic bias, and sophisticated market manipulation.

The institution identifies a particular threat in emerging markets, where regulatory frameworks often lag behind technological adoption. In these jurisdictions, small-capitalisation stocks and new listings are frequently the targets of orchestrated campaigns, exploiting thinner liquidity and less rigorous oversight. The IMF advises that even fundamental surveillance tools, deployed to track aberrant trading volumes, price spikes, and correlations with social media trends, could provide a crucial defence. It recommends that authorities implement monitoring focused on major platforms, tracking predefined keywords and hashtags to decipher commentary patterns among influential accounts, thereby establishing market-specific behavioural blueprints for manipulation.

A main worry is the return of ‘pump-and-dump’ tricks, now driven harder by digital tools. In such tricks, crooks push up the price of an asset with a wave of hype then sell the blocks they bought earlier while the price peaks leaving small investors with heavy losses. The IMF says that watching social media is vital to stop those tricks and the wider lies that feed them. Real regulators already act on the same warning – the UK Financial Conduct Authority ran a targeted awareness drive in December after it spotted a worrying new pattern. The FCA warned that AI-made deepfake videos, which copy trusted money experts or celebrities to praise an investment, are fast becoming a favourite fraud tool. The videos are built to give a false sense of trust and urgency pushing people into hasty choices. The blunt advice is that trying to beat the trick is a losing bet, because guessing the exact moment of the dump is almost impossible and the money put in can disappear completely.

The report ends with a plain verdict on the split between regulators around the world. Rich countries have the skill and the systems to adjust but many poorer markets lack the tools to handle the risks that come from swift AI use in finance. Unless they spend heavily on technology and update their laws, those markets become more fragile. For an investor, the result is a patchwork of risk – money sent to places with lighter surveillance might earn more but it also carries an extra price for the chance that market rules will fail. The IMF’s analysis ultimately frames social media monitoring not as a discretionary upgrade but as a necessary component of modern market infrastructure, essential for preserving stability in an era where trading algorithms and digital sentiment are inextricably, and perilously, linked.