The 2026Q3 ROA of -11.5% is broadly consistent with the -4.2% to -9.5% range posted through 2024-2026, indicating the underlying operating loss trajectory is continuing but not accelerating dramatically on a return-on-assets basis. However, the ROIC print of -99.6% is an artifact: when equity collapsed to approximately zero following the $598.7M to $28.6M total asset contraction, any ratio with equity or invested capital in the denominator becomes meaningless. The economically informative series remains the ROA trend, which shows GA returns averaging roughly -7% to -8% per quarter across ten quarters with no inflection toward breakeven, consistent with a company that has no mechanism to generate positive returns until Donlin reaches production. Reported ROE of -162.7% in late 2025 and the -101.7% print in 2025Q2 similarly reflect denominator instability rather than episodic loss events, and should not be extrapolated.

Interest coverage has deteriorated from -2.54x in 2024Q2 to -8.34x in 2026Q3, with extreme volatility along the way (-13.95x in 2025Q2, +5.91x in 2025Q4). The occasional positive print appears to reflect swings in non-operating items — consistent with the prior finding that equity-method JV accounting introduces significant net-line volatility — rather than any genuine coverage capacity. For a pre-revenue developer, the meaningful profitability lens is cash cost per corporate dollar spent: SG&A rose 58% over the period while operating losses more than quadrupled, suggesting project-level spend is compounding faster than overhead and that incremental capital deployed is not yet translating into measurable de-risking value on the income statement.

Total debt held between $145M and $174M for nine consecutive quarters against zero operating revenue, meaning coverage was structurally impossible from operations regardless of the D/E print. The negative interest coverage across most quarters — and its deterioration to -8.34x in 2026Q3 — confirms that debt service has been funded entirely from cash reserves or capital markets access. The subsequent drop in reported D/E to 0.38-0.41 appears more consistent with debt being extinguished, reclassified, or derecognized in conjunction with the massive asset contraction than with a healthy refinancing. Investors should not read the lower D/E as a de-risking signal without a corresponding explanation of the $570M single-quarter asset decline and what happened to the associated obligations.

A current ratio above 80x is only superficially comforting for a company burning roughly $36M per quarter at the latest operating loss run-rate; at 2026Q3's $26.2M cash level, the standalone runway appears to be well under a year before requiring fresh financing, absent the ability to draw on the Barrick JV structure. Elevated and rising DPO — from 16,941 in 2024Q3 to 43,663 in 2026Q2 — indicates payables are being stretched, which may represent either deliberate working capital management by a cash-constrained developer or a delay signal that warrants monitoring against supplier and contractor payment terms. Quick ratio equals current ratio exactly across all reported quarters, confirming there is no inventory dependence and that liquidity is a pure cash question.

Every peer in the comp set generates positive ROE (12.4%-37.5%) and net margins (13.1%-37.5%), while NG posts negative returns across all ten quarters, confirming these are not true economic comparables but valuation context only. The appropriate peer set for NG remains development-stage optionality names — Seabridge Gold, Northern Dynasty, Skeena Resources — which share the zero-revenue, pre-FID profile, though the provided comp table does not include them, limiting direct assessment. The 14.32x book multiple is particularly fragile given that NG's book value itself has collapsed toward zero in 2026Q3; if equity remains near zero, the P/B ratio becomes undefined and the entire valuation anchor shifts to EV or NAV-based frameworks, which are not reported here.