When we think of artificial intelligence (AI), our minds often jump to Hollywood blockbusters, imagining futuristic machines from I, Robot or The Terminator. Though the idea of intelligent robots in public assistance roles is still a few years away, AI algorithms already rule the global financial markets as of now.

With intelligent algorithms executing global trade of over $21 billion, human judgement no longer drives price action. Intelligent bots rip through order books in microseconds – exploiting inefficiencies invisible to the naked eye. These machines don’t debate value; they weaponise speed. A single line of code misfiring can erase billions before anyone blinks. Pakistan, meanwhile, is still writing its rulebook. The Securities and Exchange Commission of Pakistan (SECP) has finally realised the importance to regulate algorithmic trade space and is rushing to catch up with a technological revolution that has transformed global finance over the past two decades.

Proponents claim that algorithmic trading removes fallible human emotion from trading decisions and makes markets run more smoothly. However, many algorithms feed off trends from X (formerly Twitter), news headlines and other feeds, which implies they can still be affected by human herd mentality. This often results in amplification of price movements – with a possibility of market crash. In 2012, the US Knight Capital Group went bankrupt after it lost more than $450 million when its AI systems made erroneous orders that couldn’t be undone. Similarly, in 2014, a small over-the-counter trade in Japanese stock market led to an avalanche effect of trades amounting to $617 billion. Similarly, we had pound’s flash crash in 2016 due to rogue computer trades, misjudging market sentiments during Brexit negotiations.

Learning lessons from history, it would be wrong to assume that the Pakistan Stock Exchange (PSX) can handle the surge from algorithmic trading volumes and speeds. We need to forecast potential algorithmic trade scenarios and determine minimum technical requirements for microsecond trading algorithms. This is necessary to avoid system crashes and trading halts. Moreover, it is important to see if we possess adequate oversight capabilities and technical expertise to monitor microsecond trading patterns effectively.

The SECP has recommended to initially restrict algorithmic trading to institutional investors, which while prudent, creates a two-tier market that may disadvantage smaller players permanently. India’s experience shows that retail traders eventually demand access, and often get it through less regulated channels. Similarly, the proposed concept paper by SECP mandates “Initial Conformance Tests” and periodic testing, it lacks detail on stress-testing scenarios. While mandating kill switches, the proposed framework doesn’t specify response times or coordination mechanisms during market-wide disruptions. The 2010 Flash Crash in the US demonstrated that individual kill switches may be insufficient during systemic events.

Before approving any algorithms, the PSX must demonstrate that its systems can handle microsecond trading and massive order flows. The framework should mandate minimum technical standards and redundancies. Secondly, instead of a binary institutional/retail split, it may create a graduated system based on financial sophistication, capital requirements, and risk management capabilities. This could include certified retail traders and smaller institutional players.

Thirdly, the PSX needs to implement comprehensive market surveillance systems capable of detecting manipulation patterns across multiple algorithms and timeframes. The current framework’s emphasis on post-trade analysis may be insufficient.

Pakistan stands at a critical juncture. The global trend towards algorithmic trading is irreversible, and the country cannot afford to remain analogue while regional competitors leverage technological advantages.

India’s messy experience with retail algorithmic trading offers important lessons, but Pakistan’s delayed entry also provides opportunities to learn from others’ mistakes.

The writer is a Cambridge graduate and is working as a strategy consultant