From Bayesian Physics to Causal Finance

What particle physics taught me about making the invisible visible in financial data

Seeing what other’s don’t means digging deep underneath the surface. Image generated with Leonardo AI

When I began my career in particle physics, my job was to search for things no one had ever seen. Invisible Higgs decays, elusive dark matter particles — signals so faint they were drowned in oceans of background noise. Colleagues of mine built vast detectors and generated mountains of data, but the data never spoke for itself. To extract meaning, we needed a way of asking: what hidden reality could have generated these traces?

That’s where Bayesian likelihoods came in. We built models of possible worlds — different particle masses, interactions, decays — and asked: given the messy data we observed, how likely was each world to be true? Instead of declaring certainty, we lived in probabilities, constantly updating our beliefs as new evidence emerged.

At the time, I didn’t think of it as “causality.” I thought of it as physics. But looking back now, I see that what we were really doing was learning to connect evidence with causes. Not chasing correlations, but weighing hidden drivers. And it turns out, that mindset — honed in the search for invisible particles — is exactly what finance needs today.

My Dwellings in Bayesian Physics

In particle physics, nothing was ever simple. You couldn’t just point a telescope and see a new particle. Instead, you had to build statistical models that encoded what the world might look like if a Higgs boson or a dark matter particle existed — and then compare those models against real experimental traces.

The workhorse of this approach was the Bayesian likelihood. For each set of parameters — say a Higgs mass of 125 GeV or a new interaction strength — we asked: given the observed data, how likely is this scenario to be true? It wasn’t about proving or disproving a single theory. It was about constantly updating our beliefs, weighing uncertainty, and refining our picture of reality.

This probabilistic way of thinking gave us resilience. A single data point rarely mattered on its own. What mattered was how the whole picture shifted when evidence accumulated. By the end of my PhD, I realized that this mindset was just as valuable as any single equation. It was a way of seeing the world: less about certainties, more about likelihoods — and about inferring the invisible from the visible.

The Shift to Causality

Years later, when I found myself working in finance, I was struck by a strange déjà vu. Analysts were swimming in oceans of data — stock returns, ESG disclosures, credit spreads — but too often, the analysis stopped at correlation. If A moved with B, the assumption was that A explained B.

In physics, we never made that mistake. A signal in the detector might look like new physics, but until you asked what underlying parameters could have generated it, you had nothing but coincidence. Bayesian reasoning, it turns out, was a proto-form of causal reasoning: instead of “does this correlate?” the real question was “what hidden driver explains this pattern?”

That’s what causality gives us today: a structured framework for connecting evidence with causes. In finance, this means going beyond the temptation of spurious ESG correlations — “companies with more women in management outperform” — to test whether the relationship is causal, or merely coincidental. It means treating financial data with the same humility physicists bring to experimental data: messy, noisy, but full of hidden structure waiting to be revealed.

Why Finance Needs This

Finance today is where physics was decades ago: overwhelmed by data, dazzled by correlations, but still struggling to make the invisible visible. ESG and the proliferation of all kinds of unstructured data add an extra layer of complexity — environmental risks, social dynamics, governance structures — all interacting with financial performance in ways that correlation can’t untangle.

This is exactly the kind of problem Bayesian and causal thinking were built for. You start with priors — informed beliefs about what might matter. You update as evidence comes in. You distinguish genuine drivers from red herrings. And you don’t mistake coincidence for cause.

The payoff is enormous. Investors who think causally see further and act sooner. They can tell when a carbon price will truly affect valuations, when board diversity signals deeper governance strength, or when water stress is a material risk rather than a passing headline. They build strategies on explanation, not just association.

In short: finance doesn’t need more data, it needs better reasoning. And that reasoning starts with the same mindset I learned chasing invisible particles: a Bayesian lens sharpened into causal clarity.

The Bottom Line: Making the Invisible Visible

Physics taught me that the most important forces are often invisible. You don’t see the Higgs boson directly. You don’t see dark matter in your telescope. But their fingerprints ripple through the data, and if you know how to ask the right Bayesian questions, you can infer the hidden drivers shaping the visible world.

Finance is no different. Asset prices, corporate disclosures, ESG metrics — they are surface traces. The temptation is to stop at correlation, to treat patterns in the data as final truths. But behind every chart is a deeper question: what causes this?

When investors can distinguish cause from coincidence, they gain a decisive edge. They avoid false signals. They see risks before they erupt. They uncover opportunities that others dismiss.

This is the lesson I carry from particle physics into finance: make the invisible visible. Bayesian reasoning gave me the tools to weigh uncertainty; causality gives us the framework to turn those weights into explanations. Together, they offer a path beyond the noise — toward clarity, resilience, and smarter decisions in a world that refuses to sit still.

---

Reads of the Week

• This candid reflection by physics PhD student Adria Kerr offers a raw and relatable look at the emotional terrain of academic research. Instead of being a journey paved with clear guidance, Kerr reveals how much of research is about learning to navigate ambiguity, make your own opportunities, and grow from the discomfort of independence.

• This deep dive into the engineering history of the Manhattan Project challenges the usual narrative of brilliant scientists unlocking atomic secrets in isolation. Brian Potter shows that building the bomb was as much about industrial grit as it was about theoretical physics — with thousands of engineers, technicians, and factory workers inventing processes and tools from scratch.

• What if the universe isn’t a machine, but a language — and every choice, from a quantum event to a spoken word, is a sentence in progress? In this philosophically rich essay, Mark Davey blends quantum physics and linguistics to argue that uncertainty isn’t chaos but creation itself. For Wangari Digest readers drawn to deep questions about science, meaning, and how we shape reality through perception, this is an invitation to see ambiguity not as confusion but as the cosmos waiting to speak through us.