Dollar 2026 decline: more cyclical than structural

Not by long-term standards. For all the talk of the dollar being weak, it is still very strong by historical standards. One of the best ways to appreciate the overall level of the dollar is to look at a trade-weighted currency measure which is adjusted for relative consumer price levels – or what is called a ‘real’ rather than a nominal exchange rate.

Looking at the Fed’s measure of the dollar against 26 trading partners, last year’s sell-off has barely scratched the 45% rally since 2011. No wonder investors last year bought into the concept of a Mar-a-Lago accord to weaken the dollar, as Washington sought to level the global playing field for US manufacturers.

In short, the dollar is not particularly weak and could fall a lot further should it be warranted.

Yes, but only gradually. FX hedging, or the management of currency risk by portfolio managers, is a very important FX flow. The severity of the dollar’s sell-off last April, and the subsequent resilience of US asset markets, has lent much focus to the role of buy-side currency hedging. At the time, this was examined in a BIS study, where it looked like Asian investors had had a big say – even though Fed Chair Powell seems to disagree with the report’s conclusions.

We agreed with the report’s findings that high dollar hedging costs had kept investor dollar hedge ratios low, although it was notable – looking at EUR/USD hedging levels anyway – that investors were relatively underhedged early last year. The consensus view at the time was that the dollar would stay strong as tariffs took their toll on the currencies of US trading partners. That view was wide of the mark.

The latest data provided by the Danish central bank on the currency practices of local pension funds and asset managers – one of the very few data sets on buy-side FX hedge ratios – shows a 72% hedge ratio for the end of last year. January 2026 data will be released in early March. Our baseline assumes that the cyclical factor of a 50bp Fed cut versus unchanged ECB rates will see dollar hedging costs narrow further and should be consistent with dollar hedge ratios being raised to around 74% by the end of the year. In other words, more dollar selling will be coming through. A shift to an over-hedged position of 80/82% is not in our baseline and would probably require a much greater loss of confidence in the dollar.

Yes, the dollar has lost a big chunk of its safe haven value compared to 2024. In the chart, we identify a measure of the dollar’s safe haven value by calculating the difference between the Bloomberg dollar index three-month correlation with US stocks and US 10Y sovereign yields. The more negative the index, the more the dollar acts as a safe haven, as it responds positively to stock selloffs and higher long-dated Treasury yields. There are two key points here:

1. The dollar has lost some, but not all, of its safe-haven value. If we isolate the three-month correlation with the S&P500 only, that is -0.25: less negative than historical standards, but still statistically significant.
2. Correlations tend to be cyclical. The chart shows many instances of the dollar losing its appeal as a defensive currency. We need to be careful to conclude that this time the shift is structural, not cyclical.

No. Private investors – who account for more than 80% of foreign holdings in the US securities – remain highly engaged. Their net purchases rose from an average $1.0 trillion per year in 2022-24 ($88bn/month) to $1.5tr in 2025 ($128bn/month) – predominantly in equities, Treasuries, and corporate debt securities. As a result, foreign ownership of the total US securities market has continued to recover from its 2020 lows, reaching an estimated 20.2% as of Sep-25, the highest level in around a decade. Unless a major offloading in the foreign holdings of US assets takes place, the weakness in the USD would appear more cyclical.

By contrast, official foreign investors (central banks, governments, sovereign funds) have maintained their exposure to US assets, holding their USD positions broadly flat since 2020. While this still represents a cautious stance, it’s an improvement relative to the pre-2020 offloading period. Still, official investors are losing their visibility in the foreign investor pool.

If the pace accelerates materially…. China (together with Hong Kong) holds around $1.8tr of US securities, or only 5% of all foreign holdings. While China’s annual net sales have increased from $36bn in 2023 to $145bn in 2025, this has had a limited impact so far.

…but Europe matters more. Europe holds $17.1tr, or 47% of foreign holdings and has been a consistent net buyer, increasing annual net purchases from c.$0.6tr in both 2022 and 2023 to $0.8tr in 2024 and $0.9tr in 2025. These inflows have more than offset China’s gradual reductions, making European investor behaviour the key swing factor for US markets.

A further complication is the decline in transparency of ultimate investors: over the past four years, $0.3-0.8bn per year (up to 60% of net annual foreign inflows) has been attributable to international financial and custodial hubs from Europe (Belgium, Ireland, Luxembourg, Switzerland), Asia (Singapore), and the Caribbean. This complicates attribution and underscores that country‑level TIC flows can mask underlying investment sources.

No – at least not yet. Across the main structural metrics of US dollar usage – in global assets, liabilities, market turnover, and transactions – there has been no broad deterioration since 2024. In several segments, 2025 even showed signs of re-dollarisation. The latest IMF COFER release, now covering 100% of global FX reserves in 3Q25, shows a USD share of 56.9%, which is slightly higher than the FX-adjusted YE24 level (change in the headline shares is heavily affected by the exchange rate swings and may not reflect actual central bank behaviour).

The share of the USD in the OTC FX was reported at 86.8% by mid-2025, which is also an FX-adjusted recovery compared to 2024. The increase in USD share in other metrics is not consistent with the claim that last year’s dollar weakness reflected a fundamental loss in confidence in the dollar.

Looking ahead, changes in the dollar’s role as a reserve, liability or invoicing currency would be slow-moving and would require parallel shifts in private sector usage, not just central banks’ reserves. With global central bank reserves expected to grow by around $0.7tr in 2026 (IMF WEO), half of it from EM Asia, and official investors cautious about the dollar, headline USD shares may continue to decline, in line with the long-term trends — but this does not automatically imply structural erosion.

Among anecdotal evidence of de-dollarisation and an important factor to watch, we note the quadrupling of the CNY share in SWIFT’s trade finance section over the last four years to 8.3%, accompanied by China’s strengthening of its trade ties with various partners and developing its own payment infrastructure.

No. The concept of US exceptionalism is a two-sided coin. It is a trade-off between US attractiveness and the rest of the world. The theme of a ‘resilient’ global economy has been building since the second half of last year as investors positioned for the first synchronised expansion since the rolling shock of the Covid pandemic, Russian invasion, and tariffs on the global economy.

That global optimism has encouraged portfolio flows into procyclical opportunities such as emerging markets. Tracking flows into two of the largest ETFs targeting the EM equity space shows a rate of cumulative flows at its strongest in over a decade. Those flows into EM have typically been associated with a weaker dollar. It’s not that the US outlook is deteriorating. It’s just that, for the first time in a long time, there are some more attractive opportunities overseas. This theme is consistent with a benign dollar decline.

Beyond the actual flow data, four things to look at:

1. A widening of swap spreads, on the theory that net selling of Treasuries causes yields to trade higher relative to SOFR rates. It would be difficult to disentangle this from the generic outright selling of course, but it would be a signal.
2. Significant steepening in the curve, either the 5/10yr or the 5/30yr. The idea here is that the front end is protected by where the funds rate is pitched, while longer tenors are de-anchored, and thus more prone to outright selling pressure. Again, a disentanglement issue here.
3. There is auction data that can be interesting. The so-called “indirect bid” is the bucket where the official sector would come in, and this is typically central banks and the like. However, it’s not the best measure, as central banks have other ways to participate. Still, we should watch for any collapse in the indirect bid all the same, at least as a symptom of (foreign) official sector selling pressure.
4. Should there be destabilising foreign selling, it would likely show up in the functioning of the market. Wider bids/offers and/or smaller sizes could be the typical outcome. A chunk of this should be reflected in the likes of the Bloomberg US Govt Securities Index.

What could trigger a much more structural sell-off in the dollar? We are not subscribers to the view that European investors would pull their US assets en masse without good reason. This community is driven by prospective returns, not politics.

Barring the outlandish notion that the US could introduce capital controls, it is probably the area of Fed independence in the spotlight this year. The independence of the Fed is the cornerstone of global financial stability and were the Fed to be seen cutting rates inappropriately, we could see a run on the dollar. This would be embodied in US real interest rates turning negative again.

Another risk could lie in the US fiscal story. A combination of persistent large deficits and heavier Treasury issuance could further erode the perceived ‘safety’ pillar of the US’s structural appeal. If it happens alongside softer foreign participation in the US securities market, it would meaningfully increase the odds of a more structural USD repricing.

Our baseline view for the dollar is a bearish one for the remainder of 2026. USD hedging should keep up at a good pace thanks to lower front-end rates (we expect two Fed cuts this year), and a slowdown in US growth in the second half of the year will, in our view, coincide with upbeat eurozone figures, lifting EUR/USD.

We don’t expect this year’s dollar decline to match 2025’s in magnitude, but the concentration of risks in the US – from equity valuations to fiscal and political risks ahead of the midterm elections – means the risks remain on the downside for the greenback. We target 1.22 in EUR/USD by year-end.