For decades, success in quantitative trading was defined by alpha. The ability to extract consistent, risk-adjusted returns from inefficiencies in the market separated elite firms from the rest. Faster execution, better models, and sharper statistical frameworks created an arms race where marginal gains translated into meaningful profit.
That paradigm is now under pressure.
Markets have become more efficient, data has become more commoditized, and traditional signals decay faster than ever. In this environment, a growing number of quantitative professionals are beginning to question whether the pursuit of incremental alpha is hitting diminishing returns. Among them is Kelvin Leung, a Cambridge-educated mathematician and former Optiver quantitative researcher, whose career trajectory reflects a broader shift already underway: the move from extracting alpha to understanding intelligence itself as the new structural edge.
The Limits of Traditional Alpha
The modern quant ecosystem is extraordinarily sophisticated. Strategies that once generated outsized returns are now widely understood, replicated, and arbitraged away. The result is a landscape where edge is fleeting, and competition is relentless.
In such an environment, success increasingly depends on speed and scale rather than insight alone. Firms invest heavily in infrastructure, data pipelines, and execution systems to maintain even the slightest advantage. But this model has natural limits. When everyone has access to similar tools and information, differentiation becomes harder to sustain.
Kelvin experienced this firsthand at Optiver’s Amsterdam office, where he spearheaded the development of a US single-stock options strategy, building its framework, signals, and infrastructure from the ground up. The work was technically rewarding, but the half-life of any given edge was getting shorter. A more fundamental question was forming: what happens when the very structure of markets begins to change?
Intelligence as the New Structural Edge
Artificial intelligence is often misclassified as a mere investment theme, but it is actually a general-purpose technology reshaping the fundamental mechanics of information and markets. This shift carries two profound implications:
- The Democratization of Building: AI has lowered the barriers to entry, making it easier than ever for specialized professionals to launch their own funds or companies if they possess the necessary expertise.
- Rapid Model Obsolescence: Because the market is evolving so quickly, any strategy based on historical data decays faster than traditional models can adapt.
Leung argues that these mean the edge is no longer in having a model. It is in how quickly you can build, test, and replace one. And doing that well increasingly requires genuine fluency in understanding both the technology and the markets. The quant who understands AI at a technical level has a structural advantage over the one who treats it as a black box, because they can adapt faster when the environment changes again.
From Model Builders to Hybrid Thinkers
Markets are changing faster than ever, and a new class of traders is emerging to meet them: hybrid thinkers who operate beyond the boundaries of traditional quantitative analysis. They understand markets, and they know how to build: constructing AI-powered tools to capture opportunities, and identifying which previously intractable problems AI now makes solvable.
Kelvin Leung’s shift toward an AI-focused investment approach reflects this evolution. Rather than refining existing strategies within well-defined frameworks, he is focused on a harder question: what does advanced AI make possible that was not possible before, and how does that change the way you trade? When something new arrives, the edge goes to people who can recognise what it enables and build the tools to exploit it.
This requires a rare combination. Deep market experience to know where the opportunities are, as well as the technical fluency to build the systems that capture them.
The Psychology of Strategic Asymmetry
The transition from institutional trading to AI entrepreneurship is as much a psychological shift as a technical one. Traditional trading environments often reward “lower-variance” outcomes, focusing heavily on known edges and incremental improvements to existing models. While this discipline is essential for market making, it can lead to a default preference for predictable, yet capped, trajectories.
Leung describes the challenge as “rewiring the utility function”. In practical terms, this means moving away from the comfort of a stable research role to embrace a wider distribution of possible outcomes. By leaving a clear path of progress at Optiver to build at the intersection of AI and markets, Leung is applying a core trading principle to his own career: asymmetry.
In both markets and careers, outsized returns rarely come from refining the known; they come from taking calculated risks in domains with high upside potential. While many professionals choose the security of incremental gains, Leung’s move reflects the conviction that the most significant opportunities now lie in the less-defined, rapidly evolving AI space.
It is a high-stakes trade that few are willing to make, distinguishing those who optimize within the system from those who seek to build the next one.
Rethinking Edge in an AI-Driven Market
If AI continues to advance at its current pace, the traditional definition of “edge” in financial markets must be fundamentally reconsidered. Conventional advantages are rapidly becoming commoditized and less differentiated.
Kelvin believes a new form of structural advantage is emerging in its place: the accelerated buildout of alpha. Rather than relying on static models, AI enables the rapid development and deployment of new strategies as market structures evolve in real time. This shift moves the competitive frontier from maintaining legacy signals to the ability to iterate and build at the same speed as the market itself.
This does not mean that quantitative methods will become obsolete. On the contrary, they remain essential. But they are no longer sufficient on their own. The next generation of successful investors will likely be those who can integrate quantitative rigor with a deeper understanding of technological change.
The Road Ahead
In a world where AI increasingly shapes how information is processed and decisions are made, understanding intelligence itself becomes a critical skill. Traders and investors who can navigate this transition will be better positioned to identify opportunities that others may overlook.
For individuals like Kelvin Leung, the decision to move in this direction reflects both a strategic calculation and a broader view of where the world is heading. It is a recognition that the most valuable edge may no longer come from competing within the system, but from understanding how the system itself is being transformed.
That perspective may ultimately define the next era of market leadership.
