The shrinking upside in the dollar story

Dollar cycles are unusually long compared to most business or macro cycles because four mutually reinforcing forces underpin them. They only reverse when all four shift at once:

• Geoeconomic and geopolitical realignments
• Changes in growth and interest rate differentials
• Reversals in cross-border investment flows
• Deliberate FX adjustment or intervention

Our earlier research drew on historical turning points to show that the Trump shock could be sufficient to trigger such a shift, setting the stage for a structural dollar downcycle. This paper takes a different stance. Rather than revisiting the full-cycle argument, it focuses on risk. The upside risk for the dollar is diminishing, while the downside risk is becoming harder to ignore. It offers a near-term perspective on a long-cycle story and timely insight for investors reassessing their dollar exposure.

We think it is becoming less likely the dollar will go up for two reasons: one related to starting overvaluation and the second to normalising overweights. Meanwhile, the chance of the dollar falling is becoming harder to ignore as investors contemplate the growing (though still small) probability of displacing the dollar as a global asset. Overall, what is clear is that we have much more reason to expect a structural weakening in the currency, with downside risk outweighing the upside for three reasons:

1. The US dollar is now extremely expensive
2. Investors are unlikely to boost dollar demand
3. Downside risk to the dollar's position in the world is becoming increasingly salient

The US dollar has risen more than 40% from its 2011 low, giving overseas investors an extra lift on US assets but driving the currency to valuations that previously triggered domestic strain, most famously in 1985 (Plaza Accord) and 2001-02. Purchasing-power-parity (PPP) comparisons expose the distortion. Because PPP adjusts for local living costs, US GDP is identical in nominal and PPP terms. Yet, China’s PPP output is equivalent to US$40 trillion, double its nominal GDP in US dollar terms (US$20 trillion). India’s PPP output is four times larger than its nominal dollar size. Across emerging markets, the average PPP-to-US dollar GDP ratio is now 2.4 times, versus 1.3 times in developed markets.

Today’s ratios of PPP-based to US dollar-based GDP are at the highest levels in almost 25 years, coinciding with the last peak in the US dollar in 2001-2002. Considering developed markets only, the ratio is at the highest levels since the early eighties, when the extreme strength of the dollar ultimately led to Reagan’s Plaza Accord in 1985.

Figure 1: GDP in PPP terms relative to GDP in US dollars, compared to the trade-weighted dollar

Source: IMF, US Federal Reserve, Ninety One calculations. As of April 2025.

Figure 2: GDP in PPP terms relative to GDP in US dollars, compared to the trade-weighted dollar

Source: IMF, US Federal Reserve, Ninety One calculations. As of April 2025.

An over-strong dollar cheapens production abroad and erodes US manufacturing competitiveness, a growing political issue since the Trump era. Part of the gap reflects structural factors such as wages and industrial policy. Still, the currency is pivotal: when the dollar was 40% lower in 2011, the United States was among the cheapest developed manufacturing hubs, and the role of manufacturing activity in the economy (as measured by % of GDP) declined at a less precipitous rate.

Fully normalising the cost gap with emerging markets is not possible. Production cost differences between the US and EM are substantial, at more than 2x, and this is not only due to currency valuations. EMs have genuinely lower wages, to which the US should clearly not be aspiring. To fully compensate for those lower costs, the dollar would have to more than halve, a sharp move that would be destabilising for the rest of the world. However, a normalisation of the PPP ratio would see a more plausible 20-30% fall in the US dollar. This would align US costs more closely with Europe and Japan while leaving emerging market advantages largely intact. That may even create the conditions to partially redress the competitiveness imbalances that have resulted from years of strong US dollar performance against other currencies.

The idea that normalising investment flows will weigh against the dollar rests on three decisions allocators will make: first, institutional investors will trim overweight positions in US assets; second, institutional investors will reinstate currency-hedge ratios that were deliberately run down to ride the dollar boom, and third, over a longer horizon, as institutional investors sell dollar assets, benchmarks will rebalance with fewer dollar holdings. New flows will allocate less to dollar assets as institutional investors reduce holdings in line with the benchmark. This adjustment tends to take longer because it is easier to shift overweights and underweights within a single pool of capital than it is to alter benchmark weights across the global pool of capital.

The chances of normalising investment flows are higher than usual as long as European allocators are losing on both their US equity allocations against domestic alternatives and their US dollar overweights against their base currencies. That has certainly been the case this year, where an unhedged allocation YTD (as of 30 June 2025) to MSCI USA has lost European allocators 13% in euros and Japanese allocators 5% in yen when compared against their domestic allocations.

How big will the first two categories of flows be? Foreign investors now hold roughly US$32 trillion of US securities, circa US$18.6 trillion of equities, US$7.5 trillion of Treasuries and the balance in agency and corporate debt. About half of that stock is owned by European (US$13.9 trillion) and Japanese institutions (US$2.6 trillion), whose combined allocation and hedging decisions matter most for the dollar’s direction from the investment channel.¹

Figure 3: Estimates of foreign investor holdings in unhedged US equities and fixed income

Source: TIC, Haver, Bloomberg, UBS, Ninety One calculations. The calculation assumes a 20% FX hedge ratio for equities and 50% for fixed income.

A piece by Deutsche Bank’s FX Research team triangulating the custodial and survey data collected by various central banks and treasuries, along with estimated portfolio data from SIFMA (Securities Industry and Financial Markets Association) and WFE (World Federation of Exchanges) shows that European equity portfolios reduced their domestic bias by five percentage points in favour of the United States over the 15-year period from 2011 to 2024. 'Normalising' this five-point overweight in US equities would mean European investors selling roughly US$450 billion of stock, an amount comparable to the net flows into US equities from Europe that often occur over just three to six months.

Figure 4: Over the past 15 years, European investors have reduced their domestic equity bias in favour of US allocations

Source: Deutsche Bank, Ninety One calculations. June 2025.

Bond allocations have fallen as Euro-area governments issued more paper after the pandemic. Ergo, normalisation to historic weights relative to the benchmark would require adding US bond exposure and not selling US bonds.

For Japanese portfolios, it is a slightly different story according to the same research. The overall weight of US equities has risen as valuations surged from 7% to 17%, however arguably, Japanese pension funds are still relatively underweight US equities. This makes it less likely that there will be material rebalancing from Japanese allocators.

In short, trimming overweights to the dollar is likely to precipitate some selling, mainly in Europe, but either way, it is probably not enough to move the needle.

What about hedge ratios? Work by Wenxin Du and Amy Huber shows that global investors raised their dollar hedges by about 15 percentage points in the immediate post-GFC period as they hedged against US economic instability. Subsequently, they let them stagnate or even fall, especially as the dollar rallied from 2016 onward. Since 2018, hedge ratios on US bonds appear to have slipped by about five points and on equities by one to two points, letting global owners enjoy unhedged currency gains. Simply rewinding those modest moves would create roughly US$1 trillion of dollar-selling FX trades, easily eclipsing the equity-sale channel but probably not especially large in the context of the vast global FX market, where average daily turnover as of 2022 was US$7.5 trillion a day.²

Ultimately, the larger changes in dollar demand will likely come when US equity weights in global benchmarks fall from historic heights, prompting further selling. A near 70% weight in the ACWI is probably not sustainable in an economy forming 25% of global GDP. As that adjusts lower, investors will own fewer dollar assets, and new flows will buy fewer dollars. For that to happen, however, assets in other economies need to increase in value, which requires their earnings to rise. Overseas investors rarely lighten US exposure while American firms dominate global earnings and innovation. While US equities may continue to see higher ROEs, faster growth, and more innovation, investors could shift allocations if compelling opportunities emerge outside the US. Whether that happens depends on thematic and macro trends, such as the progress of the capex cycle or the next phase of the AI trade.

The combination of limited European de-risking and a potential US$1 trillion hedge-ratio reset provides a meaningful, but not overwhelming, headwind for the greenback. A full-blown dollar bear market probably requires a visible, durable turn favouring European, Japanese or other ex-US equities, perhaps catalysed by cheaper valuations abroad, shifting AI-supply-chain dynamics, or rising political uncertainty in Washington. Over a multi-year horizon, however, it’s clear the debate around investment flows is between degrees of less demand for the dollar. The actual direction of travel seems clear due to the alignment of larger hedge rebuilding, gradual overweight unwinds, and eventual benchmark weight lowering.

¹ Foreign Portfolio Holdings of U.S. Securities. Department of the Treasury.
² The global foreign exchange market in a higher-volatility environment. BIS.

This is an area we did not explore in our earlier paper, which focused on cyclical dollar bear markets. Predicting a cycle that occurs once every 18 years is already challenging; anticipating structural shifts that may happen only once every century or two is even harder. Yet in today’s multipolar world, the persistence of a single dominant currency at the centre of the global financial system looks increasingly anomalous. That raises the risk that change is coming. In this section, we examine several interconnected risks to the dollar’s position in the world.

The erosion of dollar dominance is distinct from a conventional dollar bear market. We have had three dollar-bear markets since the Nixon shock (1971, 1985, and 2002), and the US dollar was still the dominant currency throughout this entire period.

Centrality is hard to measure, but economist Ken Rogoff’s shorthand method of looking at which central banks use the US dollar as their exchange rate anchor or reference currency is a good measure of what he calls a “portmanteau measure of dominance.”³ That’s because national central banks have a thorough knowledge of their own economies and the global financial system. By this logic, the dollar’s central role in the financial world began to shift around 2015, when the People's Bank of China (PBoC) started to make its currency more flexible. This shift was solidified when Russia invaded Ukraine, and US$300 billion worth of Russian central bank reserves were frozen, a warning for ‘sovereign fiduciaries’ everywhere. The rise of gold since 2021, driven by central bank buying and defying its typical correlation with real interest rates, suggests that gold has risen in the hierarchy of safe assets for sovereigns. In fact, by some measures, gold has overtaken the euro as a reserve asset.⁴

One way dollar centrality could erode is through China loosening its ties to the dollar. As the renminbi dollar exchange rate becomes more volatile, other Asian countries may also choose to strike a new balance, given China's importance in global trade.

Figure 5: The dollar’s vast footprint as an anchor or reference currency, 2019

Source: Updated from Ilzetzki, Reinhart and Rogoff, Quarterly Journal of Economics, 2019. Ninety One, July 2025.

Domestic policy shifts

But perhaps the greatest threats to the dollar's position in the world come from within, not from without. The selection of Stephen Miran as head of Trump’s Council of Economic Advisors gave ballast to those who worried about dollar centrality because of explicit policy positions he once advocated publicly. In his notable November 2024 paper, A User’s Guide to Restructuring the Global Trading System, Miran advocated taxing dollar inflows, essentially imposing a levy on foreign capital entering the United States. The rationale was rooted in the argument that the dollar’s status as the world’s primary reserve currency creates persistent, inelastic demand for US assets, which keeps the dollar overvalued and forces the US to run chronic trade deficits. This overvaluation undermines American manufacturing and the tradable sector by making US exports less competitive. In other words, Miran explicitly highlighted dollar centrality as a negative for the economy.

Miran’s thinking did not come from nowhere. He drew on the work of economist Michael Pettis, who has long argued that global imbalances are driven by surplus countries’ policies and the reserve currency role. Rebalancing requires managing capital flows, not just trade. The idea of taxing capital inflows has also appeared in US policy debates: in 2019, Senators Josh Hawley and Tammy Baldwin proposed measures to tax foreign purchases of US assets for the same reason. While the bill did not advance in Congress, it marked a significant shift in US policy thinking toward managing capital flows away from a laissez-faire approach.

While Miran, once in the administration, disavowed his paper as a guide to policy, the idea of taxing foreign capital recurred in May 2025. At the time, a draft US fiscal reconciliation bill was sent from the House to the Senate with a provision known as Section 899. This would have allowed the US to impose higher taxes on foreign investors and companies from countries deemed to have ‘unfair’ or ‘punitive’ tax policies towards US firms, such as the so-called digital services taxes. The policy would have progressively raised taxes on dividends and interest from US stocks and some corporate bonds held by affected foreign investors by five percentage points annually, up to a maximum of 20 percentage points. It would also have ended current tax exemptions for sovereign wealth funds.

In the end, Section 899 was stripped from the final version of the reconciliation bill. But its inclusion, however temporary, underscores how quickly anti-inflow proposals can move from fringe policy papers into near-term legislative drafts.

If the United States wanted less foreign ownership of US assets, it could hardly do better than to pursue a policy like Section 899. Such a move would increase US term premia and give ballast to the ‘sell dollar’ trade for years. Yes, Section 899 was partly about giving President Trump more leverage against trading partners that are seeking to tax US corporations and individuals, which itself would throw sand in the wheels of the global economy. Nevertheless, a Section 899-type policy would add new risks to foreign holdings in the US (adding those risks to foreign holdings is, in fact, the point). The proposal may return in a new form or under different legislative cover. The bigger takeaway is that US policymakers are increasingly willing to contemplate leaning against the dollar’s global status rather than shoring it up.

Fed independence

In addition to proposals to tax inflows, there are growing risks to Fed independence. The willingness of both Democrats and Republicans to challenge the Fed’s independence, with the former tilting the Fed towards a greater focus on the environment, inequality and social justice, and the latter tilting the institution towards lower interest rates, is already contributing to questions around institutional direction.

The Supreme Court’s Wilcox case in the first half of 2025 was a narrow miss and exemplified the risks involved. Wilcox and Harris were dismissed officials who argued that if the president’s removal power extended to them, it could be used to dismiss top officials at the Fed. They reasoned that Wilcox’s National Labor Relations Board (NLRB) and Harris’s Merit Systems Protection Board (MSPB) were independent federal bodies with statutory protections.

The Supreme Court’s ruling drew a sharp distinction between the Fed and other agencies, allowing the president to remove officials from agencies such as the NLRB and the MSPB but explicitly exempting the Federal Reserve, describing it as a ‘uniquely structured, quasi-private entity’ with a distinct historical tradition in dicta, or passing remarks, thereby signalling to the executive that it would not sit idly by if Fed independence were to be challenged. That has diminished the threat to Fed independence but not entirely removed it.

America’s debt trajectory

Finally, America’s growing debt trajectory increases the salience of downside tail risks for the US dollar. US debt dynamics were already under strain from the interest rate reset following the period of very low long-term real interest rates. The willingness of the Biden administration and now the Trump administration to run large fiscal deficits is decreasing confidence that the US will be able to stabilise debt/GDP. The scale of the fiscal consolidation required is immense. A debt scenario analysis we ran in January suggests that stabilising debt/GDP would need something like 3% real GDP growth, 10-year yields under 4%, inflation north of 2.5%, and a primary budget consolidation of about 1%. If growth is lower, then the other factors must compensate. Given there is no net fiscal consolidation in the budget reconciliation bill this year, even including tariff revenues, fiscal risks are growing, not shrinking.

³ This time really is different for the dollar, writes Kenneth Rogoff. The Economist.
⁴ Gold overtakes euro as global reserve asset, ECB says. Financial Times.

The case for upside tail risk in the dollar is weakening while downside tail risk is becoming more salient. In fact, we can be more emphatic. The dollar has reached levels, both in terms of overall valuation and in terms of global investment portfolios, that were previously shown to be unsustainable, thereby capping upside. Meanwhile, the US seems prepared to reconsider the value of dollar dominance for its own economy via proposals to tax capital inflows, eroding Fed independence, or maintaining higher fiscal spending. As a result, there is much more reason to expect a structural weakening in the dollar. The balance of probabilities favours sustained dollar weakness over the coming cycle.