Mean Reversion vs. Momentum: What the Evidence Actually Says

Ask the internet how markets work and you will get two confident, contradictory answers. One camp says trends persist; the other says extremes snap back. The remarkable thing is that the academic record supports both — at different horizons.

Over three-to-twelve-month windows, past winners tend to keep winning: the momentum effect^[1]. Over multi-year windows, past extreme losers tend to outperform — long-horizon reversal, the signature of overreaction^[2]. Both have survived decades of out-of-sample scrutiny and sit inside the modern factor canon^[3].

The market is not always one thing. The edge is in matching the strategy to the regime — not in believing one rule forever.

Why this is fatal to influencer “systems”

A single rule sold as universally true ignores regime, and edges decay as they are discovered and crowded^[4]. The honest version is adaptive and evidence-based: test the effect, size it, bound the risk, and retire it when the data turns.

Niro builds on research-grounded, regime-aware structures — the logic institutions use — and runs every one through defined-risk controls and significance gating. That is the opposite of a one-size-fits-all rule shouted from a feed.