The two sides of sovereign balance sheets: how much do government financial assets tell us about r - g?

Governments hold large stocks of financial assets, yet debt rules and sustainability assessments look almost only at what they owe. This paper asks whether the asset side of public balance sheets tells us something about the interest–growth differential, r - g, that debt alone does not. We follow 17 advanced economies from 2000 to 2026 with sparse dynamic factor models, which let the states governing r - g emerge from the data instead of being imposed through debt ratios. The answer is more uncertain than either side of the debate assumes. On whether the asset side adds information beyond debt, a pre-specified forecast test is inconclusive: it finds no evidence that it does for r - g a year ahead, though it could not reliably have detected information large enough to move a one-year debt projection by the smallest annual adjustment the EU debt safeguard demands. A second comparison, on a consistently measured r - g, can neither confirm nor rule out information of that size. Nor is the asset side shown to matter differently when debt is high: the link between asset accumulation and long-term yields is not established to change with the debt burden. Countries respond to common forces in ways a weak test cannot tell apart, yet their coefficients differ for every variable. Applied to fiscal anchors, net and gross debt measures disagree in about one country-quarter in seven, showing where each would bind. The paper maps how much the asset side can, and cannot yet, tell us about debt sustainability.