Mean reversion: powerful until the regime shifts

If the current US administration were replaced tomorrow, could markets and institutions revert back to what investors had grown used to before 2024? I doubt it. That isn’t a political statement, it’s an economic one. Behaviour changes. Long-held alliances fray. Policy starts moving in a more transactional direction, less bounded by the old guardrails. When that happens, the earlier setting won’t just sit there waiting to be restored. When policy begins breaking away from the old rules, markets start to price possibilities that previously sat outside the system. That uncertainty doesn’t disappear just because the occupant changes. That’s why the old rules deserve a harder look.

There’s been a lot of discussion lately encouraging investors not to react to the noise. That instinct has history behind it. Markets have a long habit of overshooting during wars, geopolitical shocks and policy scares, then settling once the fog lifts. Buy on the cannon, sell on the trumpet, or as Warren Buffett used to say: “buy fear, sell greed.” There’s real wisdom in Buffett’s mantra. Patient investors have often been rewarded while others panicked. Tactical advice may still work. The strategic assumptions underneath it, however, may have changed.

The issue may sit somewhere else. This piece isn’t about whether to ‘buy the dip.’ It’s about what we are buying back into. A tactical rule can work very well inside a stable strategic setting; it tells us far less, however, when the setting itself may be shifting. Historical examples cited to support mean reversion came from a different base. They came from a world that, for all its shocks and conflicts, still operated inside a more settled institutional frame. That stability formed the foundation from which such recoveries occurred. When that base begins to shift, the comparison becomes less straightforward.

The base from which we’re operating today, I believe, isn’t the same base as the historical examples cited. Debt burdens are historically unprecedented. Valuations remain stretched, even after looking past the ‘Magnificent 7.’ Policy signals today seem to arrive with shorter notice and wider dispersion. More broadly, the world has been moving away from the more rules-based multilateral setting to which investors had grown accustomed and toward something more transactional, more bilateral and less anchored in the institutional frameworks that shaped the previous regime. None of that tells us where the market trades next quarter. It does suggest, however, that the old comparisons deserve more caution than they usually get.

For close to four decades investors in developed markets operated inside something like an economic detente. Left and right still fought over policy, often hard, but both sides understood that pushing too hard outside the rules-based system, and the institutional guardrails around it, was self-defeating. Driving-up the uncertainty premium raised borrowing costs for everyone. That restraint was understood and broadly upheld.

That’s where the language of ‘regime shift’ becomes useful, used carefully. Scientists studying complex systems in physics, ecology and biology have observed the same broad pattern for years. In physics, regime shifts are referred to as ‘phase transitions’. As systems approach structural change, they often grow noisier before they settle into something new. Financial markets are obviously not ecosystems or weather systems, but the parallel is still worth noticing. Volatility won’t always show up as a passing mood. More often it bunches together. Relationships inside the system begin to wobble. The system becomes noisier before it becomes different. When that happens, the signals analysts watch start to appear together. Volatility clusters, while relationships can often drift.

Volatility may not simply be noise: it may be the market trying to price a changing environment. Diversification rests on assumptions about how assets behave relative to one another, and those assumptions sit on correlation structures that do not hold by divine right. Correlations hold until they don’t. When the regime shifts, those relationships can move with it. Mean reversion may still occur, and it often does. The harder question, however, is whether mean reversion instincts rely on a degree of regime stability investors have become too comfortable assuming will persist into perpetuity.

That’s why I’m not persuaded by the easy reassurance that every shock becomes a buying opportunity. More often than not, it does. The tactical trade is usually the right call. But saying markets overshoot isn’t quite the same thing as asking from what they’re overshooting. A market can rebound from a dislocation while the structure underneath it is still shifting. A dislocation can bounce-back quickly. The base underneath it is a whole different question. Even where prices do recover, the setting into which they recover may be carrying a different set of assumptions, constraints and risks from what those investors remember.

It’s also worth flagging how most warnings of structural break turn out to be too early or plain wrong. Investors resist the phrase ‘regime shift’ for good reason. History is full of pundits declaring a new era just before mean reversion reasserted itself. That scepticism has earned its stripes. Yet when a regime does shift, the early signals look identical to noise. When regimes start to shift, time horizons begin to impact more than usual. Some investors can wait for the system to settle. Older cohorts do not have that luxury. That’s the difficulty. The evidence doesn’t arrive carrying a label. It turns up looking like another wobble, another exaggerated move, another scare investors are told to look through.

So this isn’t a case for panic, or timing the market into cash. Discipline and patience still count. What deserves closer examination, however, is the assumption underneath the reflex. Buy-the-dip can be good tactical advice. It’s not, by itself, a strategic framework. Markets still bounce. The $4 trillion question is whether the floor from which they are bouncing has shifted.

Rob Prugue

Rob Prugue is an investment consultant and the former long-time CEO of Lazard Asset Management, Asia Pacific.