Almost sure null bankruptcy of testing-by-betting strategies

The bounded mean betting procedure serves as a crucial interface between the domains of (1) sequential, anytime-valid statistical inference, and (2) online learning and portfolio selection algorithms. While recent work in both domains has established the exponential wealth growth of numerous betting strategies under any alternative distribution, the tightness of the inverted confidence sets, and the pathwise minimax regret bounds, little has been studied regarding the asymptotics of these strategies under the null hypothesis. Under the null, a strategy induces a wealth martingale converging to some random variable that can be zero (bankrupt) or non-zero (non-bankrupt, e.g. when it eventually stops betting). In this paper, we show the conceptually intuitive but technically nontrivial fact that these strategies (universal portfolio, Krichevsky-Trofimov, GRAPA, hedging, etc.) all go bankrupt with probability one, under any non-degenerate null distribution. Part of our analysis is based on the subtle almost sure divergence of various sums of $\sum O_p(n^{-1})$ type, a result of independent interest. We also demonstrate the necessity of null bankruptcy by showing that non-bankrupt strategies are all improvable in some sense. Our results significantly deepen our understanding of these betting strategies as they qualify their behavior on “almost all paths”, whereas previous results are usually on “all paths” (e.g. regret bounds) or “most paths” (e.g. concentration inequalities and confidence sets).

💡 Research Summary

The paper investigates the asymptotic behavior of a broad class of testing‑by‑betting strategies when the null hypothesis that the mean of a bounded i.i.d. sequence equals a specified value holds. The authors focus on strategies that are known to be “power‑one” under any alternative distribution—i.e., their wealth processes grow to infinity almost surely, often at an exponential rate. Examples include the Krichevsky‑Trofimov (KT) bettor, the GRAPA (Growth‑Rate Adaptive to the Particular Alternative) bettor, mixture portfolios (continuous mixtures over fixed‑fraction bets), and predictable hedging schemes. While these methods have been celebrated for their strong regret guarantees, optimal confidence sequences, and exponential wealth growth under alternatives, their fate under the null has been largely unexplored beyond the trivial observation that the wealth process is a non‑negative martingale and therefore converges almost surely to some limit.

The central contribution is a rigorous proof that all of the aforementioned “good” betting strategies go bankrupt with probability one under any non‑degenerate null distribution (i.e., any distribution on (