Market gaps present opening for AI-driven participants - GSIA

Structural weaknesses long viewed as constraints in the domestic equity market could become a source of competitive advantage for investment managers deploying artificial intelligence, according to Kisseih Antonio, president of the Ghana Securities Industry Association (GSIA) and Managing Director of Sentinel Asset Management.

Antonio made this case at the Global Access Forum organised by Ultima World on the theme ‘Artificial Intelligence as a Strategic Advantage in International Markets’, arguing that thin analyst coverage, limited liquidity and slow price discovery – features typically associated with underdeveloped markets – create conditions in which AI-powered strategies can generate excess returns that will be rapidly arbitraged away in more efficient markets.

“Pricing inefficiencies that would disappear within milliseconds on the New York Stock Exchange or London Stock Exchange persist for longer here. Managers deploying AI-based screening and signal generation are not competing against hundreds of quant funds running identical strategies. In many cases, they are operating in an uncrowded space,” he said.

His remarks challenge a long-standing narrative around the Ghana Stock Exchange (GSE), where limited depth and liquidity have traditionally been seen primarily as impediments to market development.

Instead, Antonio reasoned that these same features create exploitable opportunities for firms with the analytical capacity to identify and act on mispricing ahead of the market.

The argument arrives at a moment of considerable global momentum behind AI adoption in financial services. AI firms accounted for 61 percent of global venture capital investment in 2025, attracting US$258.7billion out of a total US$427.1billion – more than doubling AI’s share of venture capital since 2022, according to the Organisation for Economic Cooperation and Development (OECD).

Within the asset management industry specifically, a survey of 150 fund managers by the Alternative Investment Management Association (AIMA) found that 95 percent reported using generative AI in their work – up from 86 percent in 2023, with 58 percent expecting wider front-office integration over the next year.

Large asset management firms with annual revenues of at least US$1billion plan to invest an average of US$101million in AI in 2026, according to a KPMG quarterly survey.

Recent market performance lends weight to Mr. Antonio’s argument as the GSE was Africa’s best-performing equity market in 2025, delivering a 79.43 percent return in local currency terms.

The benchmark index crossed 15,000 points for the first time in March 2026 and has posted triple-digit gains over the past 12 months, despite operating within a relatively small and concentrated market.

At close of the first quarter, the GSE Composite Index had recorded a year-to-date gain of 48.91 percent while the Financial Stocks Index surged 71.86 percent.

By mid-April the index had pulled back to 14,024.22 on account of consolidations after its March highs, though it retained a year-to-date gain of 59.91 percent with market capitalisation standing at GH¢266.4billion – equivalent to approximately US$24.1billion.

Such performance has been accompanied by sharp dispersion in individual stock returns. Clydestone Ghana has risen 1,700 percent from a year ago while SIC Insurance (711 percent), Ecobank Ghana (629 percent) and GCB Bank (347 percent) also posted substantial gains over the same period.

Moves of that magnitude, in a market with historically limited research coverage, point to persistent information asymmetries; conditions under which quantitative models can identify patterns and signals more quickly than discretionary investors.

However, the same characteristics that create opportunity also impose constraints. Thin liquidity can limit the ability of asset managers to build or unwind positions without affecting prices, raising execution risks and capping the scale at which AI-driven strategies can be deployed.

Antonio identified data infrastructure as the most binding constraint on wider adoption. Ghana’s financial data ecosystem remains sparse, with gaps in historical datasets and limited access to real-time and alternative data streams. These shortcomings restrict the effectiveness of machine learning models, which depend on large and consistent datasets for training and validation.

A Q3 2025 global survey of 500 asset management executives conducted by ThoughtLab found that about half of firms lack basic processes to clean, normalise and tag internal data or source high-quality external data, citing data gaps and slow-moving cultures as the most significant barriers to AI adoption.

The challenge is more acute for managers operating in frontier markets, where data ecosystems are typically thinner and less standardised than those in developed market peers.

He also pointed to a dearth of domestic quantitative talent and the early-stage nature of regulatory frameworks governing AI in financial markets. The Securities and Exchange Commission (SEC) has yet to issue detailed guidance on the use of AI-driven strategies, consistent with the broader challenges faced by regulators in emerging markets.

Institutions such as the International Monetary Fund (IMF) have noted that supervisors in developing economies must adapt to AI adoption without the incremental transition experienced in advanced markets, increasing the risk of regulatory lag.

Despite these constraints, Antonio argued that delays in adoption could prove costly. Over a five to seven-year horizon, he said, the cost advantage of AI-enabled firms is likely to become structurally decisive, compressing margins for lagging managers and shifting competitive advantage toward early adopters.

A 2025 McKinsey analysis found that asset manager margins had already declined by three percentage points in North America and five percentage points in Europe over the preceding five years, a pressure that AI adoption is increasingly seen as a mechanism to relieve.

“The strategic question is not whether to act, but how quickly and with what level of commitment,” he said.