Art vs Science: How Political Violence Underwriters Are Re‑Balancing Instinct and Intelligence

Political violence and terrorism (PVT) underwriting has never been simple but today, the balance between traditional underwriting instinct and modern data science is shifting faster than ever. In a recent episode of the Political Risk Podcast, Fidelis Partnership’s Head of Crisis Management, Billy Ayres and AdvantageGo’s Global Head of Pre-Sales, Simon Fagg unpacked how underwriters are adapting to a world where social unrest is more frequent, terrorism continues to evolve, and the demand for faster, more reliable insights keeps growing.

Across the discussion, three themes stood out:

• * the ongoing tension between art and science
• * the challenges of incomplete or unreliable data, and
• * the rising need for real‑time exposure intelligence in an increasingly volatile world.

1. The Art–Science Tension Is Still Alive But the Scales Are Shifting

Underwriting has always been a blend of qualitative judgement and quantitative inputs. But in PVT, where historical claims data is scarce and risks don’t follow the “laws of physics” of cat perils, that blend is even more acute.

Ayres summed it up succinctly, “From a very high-level perspective, the art is somewhat still winning the battle against the science.”

The lack of robust stochastic models, the bread and butter of property underwriting, forces PVT underwriters to rely heavily on experience, context and geopolitical understanding. While deterministic modelling is widely used, the limited depth of historical claims data creates inherent uncertainty.

Even so, underwriters want more from the scientific side of the equation. Better signals on political instability, more predictive analytics on elections or protest movements, and a more consistent view of country‑level risk. As Ayres put it, the dream is “uniformity and consistency within our data,” though he joked it’s about “as likely as my morning train turning up on time.”

The future? A world where richer data sets enhance rather than replace specialist underwriting instinct.

2. When Data Is Poor, Underwriting Gets Even Harder

In a line of business shaped by geopolitics, social movements and the sheer unpredictability of human behaviour, data quality varies wildly.

Sometimes it’s as simple as receiving a poor‑quality schedule that doesn’t allow the most basic understanding of a risk. Other times, the problem is far bigger, for example, social media producing vast amounts of information, much of it unreliable.

Ayres highlighted the double‑edged sword clearly, “There’s a readily available amount of data on social media that might not always be as accurate as you’d like it to be… We want more data, but we want to make sure it is reliable.”

And when the data simply isn’t there?

Underwriters either decline to quote or lean even more heavily on expert judgement, local insight and deterministic modelling. At Fidelis, drawing on internal regional expertise, including professionals who’ve lived through local escalation cycles has become an increasingly important part of their “art” toolkit.

The market feels this gap sharply when trying to model SRCC (strikes, riots, civil commotion), now emerging as the most loss‑active peril in the portfolio. Without historical loss curves or standard industry footprints, the potential for accumulation is hard to quantify and even harder to price accurately under intense time pressure.

3. Real‑Time Exposure Intelligence Is Becoming Essential

Perhaps the clearest theme running through the conversation is the growing demand for real‑time visibility. Not just on a single risk, but on how that risk interacts with the portfolio as global events unfold.

The rise in large‑scale SRCC events from Chile to South Africa to New Caledonia has forced underwriters to rethink traditional aggregation models. Country‑level views are no longer sufficient. City‑centre clusters and protest‑footprint‑based scenarios matter far more.

Tools like Fidelis’s in‑house FireAnt platform, and platforms like AdvantageGo’s Exact, are giving underwriters the ability to drop dynamic ‘nets’ over cities, test scenarios, and instantly see exposure spikes.

As Fagg described it, the future is about rapid point‑of-bind assessment… and the ability to ascertain what’s going on in more or less real time with trusted data.

With social mobilisation accelerating, election cycles heating up globally, and geopolitical volatility now a constant rather than an exception, this real-time situational awareness is becoming a competitive necessity.

A Line of Business in Flux And Reinvention

Political violence underwriting is evolving quickly. Terrorism risk is shifting towards low‑tech, high‑frequency attacks such as vehicle ramming or lone‑actor assaults. SRCC has become more common, more coordinated and more destructive in aggregate. And clients, now far more aware of these exposures, expect flexible, responsive cover.

For underwriters, the answer isn’t to choose between art and science. It’s to fuse them.

The podcast makes it clear: the future belongs to underwriters who combine deep geopolitical awareness, trusted real‑time analytics, and smarter scenario-based modelling – all supported by tools capable of bringing disparate data sets together at speed.