Rather than the US net debt threatening the dollar's role, it is the dollar's global role that has driven the deficit: US financial assets — treasuries plus private securitized assets — are uniquely attractive due to the economy's size, perceived military and legal safety, and the creativity of US financial markets, which sustains the dollar's reserve role and skews the trade deficit toward the US.

Under heroic assumptions, a 15pp tariff increase with the dollar at current levels would shrink the goods-and-services trade deficit by $300-400 billion (about 1% of GDP), but since the current account deficit is 4% of GDP, this is a material change rather than elimination — and the effect on the current account may be even smaller because the investment income deficit is likely to widen.

Basic economic theory implies a 10% broad tariff should push the dollar up about 5%, offsetting the tariff's effect on import costs and equally reducing both exports and imports, leaving the trade deficit unchanged in that simple framework.

There was a massive amount of currency manipulation in 2003-2013 (now persisting at a much lower level) that, because the dollar is the main currency it flows into, tends to skew the trade deficit toward the US.

The US current account has been in deficit continuously for over 40 years, with net liabilities to the rest of the world reported at 90% of US GDP at end of last year (perhaps ~67% adjusted for measurement issues); no single country in human history has ever borrowed so much as a share of the world economy.

The 'exorbitant privilege' — the US paying a lower return on its borrowing than it earns on its lending — is not unique to the US but shared by most advanced economies, while developing economies have the reverse; if the US retains this privilege, stabilizing its net liability/GDP ratio requires the goods-and-services deficit to shrink by 2% of GDP (implying a 15-20% real depreciation), but if it loses the privilege the deficit must shrink 3.5% of GDP (implying a 25-30% real depreciation).

Because about 80% of US bond and bank-loan claims on foreigners are in dollars, the exorbitant privilege is largely driven by the spread between what the US charges foreign dollar borrowers and what it pays on dollar liabilities (tied to treasury spreads); those spreads are now at their lowest in over 25 years, with a risk of going lower that would reduce the privilege.

Customs duties have quadrupled this year but, as of August, were only about 8 percentage points of goods imports higher — roughly half of estimates like Chad Bown's 21pp, CBO's 18pp, or Yale Budget Lab's 15pp — suggesting either loopholes/exemptions, evasion, or more increases to come.

The dollar appreciated ~5% in real broad terms from October 2024 to January 2025 (consistent with markets pricing a 10% tariff), but from late January it gave up all those gains and returned to the starting point; Gagnon interprets this reversal as foreign investors recoiling from erratic policies, out-of-control fiscal policy, and anti-foreigner grievance rhetoric.

If all the factors (large fiscal deficit, AI investment boom, clean household balance sheets) were truly pushing up the trade deficit, the dollar should be stronger; since markets are not pricing that, the trade deficit cannot widen unless the dollar rises — the saving-investment imbalance reconciles with a depreciated dollar through inflation and timing rather than being a genuine conundrum.