The Irrelevance of Relevance

In October 1973, Israeli military intelligence operated under a framework it called the Konzeptziya: Egypt would not attack until it could neutralise the Israeli Air Force, therefore Egyptian troop movements were irrelevant. The assessment held until Egyptian forces crossed the Suez Canal. Fifty years later, a different generation of analysts applied an updated version of the same filter to Hamas in Gaza — contained, degraded, not the priority. The word had changed. The architecture had not. Both failures were catastrophic, and neither was unique to Israel.

Somewhere between the grant application and the trading screen, “relevance” became the word that ate the world. Universities now demand it of every research proposal. Fund managers invoke it to justify chasing whatever narrative is moving markets this quarter. Foreign policy analysts deploy it to triage which crises deserve column inches and which can be safely ignored. In each arena, the demand for relevance produces the same perverse result: it systematically filters out the things that matter most.

Start with the academy. The modern research funding apparatus operates like a Herfindahl–Hirschman Index in reverse: instead of measuring dangerous concentration, it enforces it. Panel after panel channels money toward whatever the policy establishment has already decided is important — artificial intelligence, net zero, pandemic preparedness — while defunding the basic science from which every one of those applied fields originally sprang. Quantum mechanics was gloriously irrelevant for decades before it underwrote the semiconductor industry. Ramanujan’s number theory gathered dust in Cambridge notebooks until it resurfaced in string theory and, later, in the elliptic curve cryptography that now secures global payments infrastructure. Relevance, as a filter, would have killed both at birth. The academy’s obsession with demonstrable impact is Le Chatelier’s principle made institutional: the harder the system pushes toward predetermined equilibrium, the more it forecloses the disruptive perturbations from which genuine breakthroughs emerge.

The same pathology infects financial markets, where it wears a slightly different mask. Here, relevance goes by the name of “narrative.” A stock is relevant if it fits this quarter’s story — generative AI, reshoring, the energy transition. A sector is irrelevant if no analyst on television is talking about it. The result is a market structure that systematically overprices the fashionable and underprices the overlooked, generating precisely the fat-tailed return distributions that modern portfolio theory was never built to handle.

Consider shipping. For most of the 2010s, dry bulk carriers were the definition of irrelevant — ugly balance sheets, cyclical earnings, zero narrative appeal. Analysts who covered them apologised for it. Then pandemic supply-chain disruptions and a decade of underinvestment in fleet capacity converged, and freight rates did what neglected variables do: the Baltic Dry Index rose fourteenfold in eighteen months. The investors who had been paying attention to the irrelevant did not need to predict the catalyst. They had bought optionality cheaply because no one else wanted it.

That is the deeper market logic. The investor who chases relevance is, in the language of options pricing, buying an at-the-money call on consensus at its peak implied volatility — the most expensive position available. The contrarian studying the out-of-favour balance sheet is effectively writing puts at deep out-of-the-money strikes, collecting premium while everyone else pays it. Put-call parity tells us that the price of protection must equal the price of opportunity. When the entire market is crowded into one side of that equation, the expected shortfall on the other side becomes catastrophic. Relevance, in markets, is not merely unhelpful; it is the mechanism by which capital is most efficiently destroyed.

And then there is geopolitics, where the cost of relevance-filtering is measured not in basis points but in blindsided governments and humanitarian crises that were somebody else’s problem until, suddenly, they were not. The Western foreign policy establishment runs on a triage model: events are sorted into “relevant to our interests” and “peripheral.” The 1997 Thai baht devaluation was peripheral — until it cascaded through every emerging market on earth and arrived, via Long-Term Capital Management, on the desks of the Federal Reserve. The Houthi militia’s slow accumulation of anti-ship missile capability was peripheral — until container shipping rates through the Red Sea tripled and Europe’s path back to its inflation target was quietly complicated. The demographic shift unfolding across Central Asia is, by current consensus, peripheral. Uzbekistan’s birth rate has been falling for a decade, driven by a shrinking cohort of childbearing age born during the post-independence economic collapse. Kazakhstan sits on rare-earth and uranium deposits that three great powers need. The Aral Sea basin’s water politics pit five nations against each other in a region where every river is somebody else’s irrigation system. None of this is on the briefing agenda. All of it will be.

Each of these episodes is a case study in fat-tail contagion: a variable deemed irrelevant by the prevailing model turns out to sit in the tail of the distribution, where the real variance lives. The expected shortfall of ignoring the “irrelevant” is not a modest underperformance. It is regime change — in portfolios, in policy, and occasionally in governments. Israel’s intelligence community has a word for the architecture that produces these surprises. Twice in fifty years, a Konzeptziya told the relevant authorities that the irrelevant variable would not move. The rest of the world lacks the word but not the habit.

What connects these three arenas is a shared epistemological error. In each case, relevance functions as a filter that optimises for legibility: the grant panel can explain why it funded pandemic research; the portfolio manager can explain why she bought Nvidia; the intelligence analyst can explain why he briefed on China and not on Tajikistan. Legibility is comfortable. It satisfies committees. It survives audits. But it is not the same thing as understanding, and the gap between the two is exactly where risk accumulates. It also herds. The professional cost of being wrong alone always exceeds the cost of being wrong together, so the relevance filter is self-reinforcing: once enough grant panels, enough analysts, enough briefing desks converge on the same triage, deviating from it becomes a career risk in itself. The contrarian researcher, the deep-value fund manager, the analyst who insists on monitoring Tajikistan — each faces an incentive structure that punishes solitary error far more harshly than collective blindness. The herd does not follow the signal; it becomes the signal, and anything outside its field of vision ceases, institutionally, to exist.

Worse, the error is path-dependent. Each funding cycle that bypasses basic science shrinks the pool of researchers who could have spotted the next breakthrough. Each quarter that ignores an unfashionable sector thins the analyst coverage that might have flagged the turning point. Each intelligence briefing that leaves the peripheral item off the agenda erodes the institutional knowledge needed to interpret it when it finally forces its way on. The relevance filter does not just miss once; it creates a blind spot that widens with each repetition, so that the eventual surprise is larger precisely because of the duration of neglect.

Real options theory offers a useful corrective. An option has value precisely because the future is uncertain. The optionality embedded in basic research, in contrarian investment, in monitoring peripheral geopolitical theatres, derives its worth from the possibility that the world will not conform to the consensus model. But theory alone does not explain why the option stays so cheap. Contrarianism does. The contrarian pays a visible cost — career risk, ridicule, years of underperformance against a benchmark the herd has defined — in exchange for optionality that is invisible until the moment it pays off. The premium is not financial; it is reputational, and most institutions are not structured to bear it. That is why the irrelevant remains systematically underpriced: not because no one can see its potential value, but because the social cost of acting on that value exceeds what most professionals are willing to pay. To demand relevance upfront is to exercise every option at inception — converting the portfolio’s embedded volatility into a single, fragile bet on the present configuration of the world.

The relevant, almost by definition, is already priced in. It is already known, already modelled, already crowded. The irrelevant is where the unpriced risk and the unpriced opportunity coexist, waiting for the perturbation that no committee anticipated. The most dangerous sentence in any institution is “that’s not relevant to us.” It is the sound of a briefing room that has emptied itself of every voice that might have said otherwise — and is about to discover, at considerable expense, that the territory does not consult the map.