Authored by Lance Roberts via RealInvestmentAdvice.com,
💰 The Dollar’s Funeral Nobody Attended
Open any finance corner of social media this week, and you will be hit with some version of the same obituary. ZeroHedge declared in December that the dollar’s death in 2026 is now a mainstream talking point, citing a WIRED piece arguing that this is the year “dollar dilution” truly accelerates. A widely circulated Dollar Collapse post this month warned that foreign demand for Treasuries is fading and that the greenback is losing its safe-haven status “in a generation.” WatcherGuru ran a headline last fall declaring rapid de-dollarization is happening right now, while YouTube personalities brandish century-long purchasing-power charts preaching gold and Bitcoin as salvation. The story writes itself: investors positioning capital today need to decide how much survives contact with the data.
To be fair, there is enough truth in the narrative to keep it alive. The DXY has retraced roughly 10% from its early-2025 peak near 103.5. The IMF’s COFER data shows the dollar’s share of global FX reserves has slipped from 73% in 2001 to around 58% today. Central banks purchased 863 tonnes of gold in 2025 — cooler than 2024’s 1,092-tonne haul, but still the fourth-largest annual reserve build on record, roughly double the 2010–2021 average of 473 tonnes, and the 15th consecutive year of net official buying. BRICS+ now includes Iran, Egypt, Ethiopia, the UAE, Saudi Arabia, and Indonesia, representing nearly half the world’s population. China has trimmed its headline Treasury holdings by more than 27% since 2022, and the bloc is actively building non-SWIFT payment infrastructure through BRICS Pay and CIPS.
Assemble those data points in the right order, and you can construct an apocalyptic narrative that plays extremely well in a three-minute video. The problem is that the dollar the narrative describes is not the dollar the flows describe. As is so often the case in markets, what “everybody knows” is precisely what is already priced in, and frequently wrong.
The Dollar’s Tape Disagrees
Start with the dollar itself. As of Tuesday, when I started writing this analysis, the DXY sits at roughly 98, essentially flat over the trailing 12 months and still well above its longer-term historical averages. For context, the index traded below 80 for most of 2011 through 2014. A dollar “near 100” is simply not consistent with the word “collapse.”
More importantly, look at what foreign investors are actually doing with their dollars. According to the latest Treasury International Capital (TIC) data, foreigners purchased a net $101 billion of long-term U.S. securities in February, following November’s blockbuster $222 billion print. Across the last five reporting months, net foreign inflows into long-term U.S. stocks and bonds totaled roughly $488 billion, a pace that rivals the post-COVID liquidity surge. If the world were truly abandoning the dollar, somebody forgot to tell the world’s money managers.
The BIS tells the same story from a different angle. The 2025 Triennial Central Bank Survey found that the US dollar accounted for 89.2% of all foreign exchange transactions in April 2025, up from 88.4% in 2022, across $9.6 trillion of daily turnover. The renminbi’s share climbed to 8.5%, a meaningful progress, but still a fraction of the dollar’s transactional footprint. Reserve share is drifting lower; actual dollar usage is not.
The China “Dumping” Illusion
If one chart carries the de-dollarization narrative more than any other, it is the headline decline in China’s reported U.S. Treasury holdings. Those holdings in “US Custody” declined from roughly $1.2 trillion at peak to about $683 billion today. That looks like a 50% purge, and it gets rolled out as Exhibit A in every “dollar is dying” thread. Pay attention to the highlight of “US Custody.”
As we detailed in our two recent pieces, “Is China Really Dumping US Treasuries?” (February 23) and “The Dollar’s Plumbing: Conspiracy Vs. Data” (March 20), that chart is genuinely misleading. The Treasury’s own TIC FAQ flags the problem: holdings are reported by the location of custody, not by who bears the economic risk. China has been quietly migrating that custody, not liquidating it.
The evidence is in the data for two very small countries. As of November 2025, Belgium reported $481 billion in Treasury holdings and Luxembourg $425 billion, enormous totals for nations not remotely building reserves at that scale. Belgium is home to Euroclear; Luxembourg hosts Clearstream, and both countries are global settlement hubs. Over the period, China’s reported holdings fell by roughly $600 billion, Belgium’s rose by roughly $500 billion. Over the last twelve months, the UK, Belgium, and Japan were each net Treasury buyers of more than $115 billion, with Belgium’s holdings up 26%, the largest percentage gain among major holders.
As noted in that article:
“This is not a conspiracy. It is plumbing. One of the primary reasons that China uses Belgium for custodial purposes, besides avoiding geopolitical risk, is that the Euroclear Bank is based there and sits at the center of cross-border settlement and collateral mobility. Clearstream’s international depository is based in Luxembourg and serves the same global institutional client base. When a central bank or a state institution wants to hold a large Treasury portfolio with flexible settlement and collateral options, these hubs help address operational challenges.”
If we adjust China’s reported Treasuries to account for the custody migration to Euroclear and Clearstream, the total barely changes from its 2011 level. Meanwhile, total foreign holdings of U.S. Treasuries hit a record $9.4 trillion in November 2025. This is post-2022-sanctions risk management, not de-dollarization, and the dollar exposure is staying put. The plumbing changed. The balance did not.
The truly meaningful story is not about the dollar. It’s about who holds the debt and where they custody it.
Follow the Earnings, Follow the Flows
Why are foreigners still buying? The recent A16Z charts that are making the rounds this week tell you everything you need to know. Consensus earnings growth for the U.S. IT sector has been steadily revised higher all year, from 30.9% at the start of January to 37.1% in late February and 43.4% as of April. Info Tech is now expected to grow earnings more than 2x faster than the S&P 500 in 2026 (40% vs. 18%), with only Energy and Materials meaningfully outpacing the broader index.
This is the most important lesson to learn: Capital follows returns. Europe’s 2026 earnings growth runs in the mid-teens, while Japan is meaningfully lower, and Emerging markets approach U.S. IT growth rates but carry convertibility, governance, and geopolitical risks that most fiduciary capital will not underwrite at scale. Global pension funds, sovereign wealth vehicles, and private wealth allocators with fresh savings to deploy effectively have no choice but to route capital back into U.S. equities and the Treasuries that fund the dollar leg of those allocations. That is the mechanical underbelly of the AI capital cycle, and it is still early.
What to Recognize About the Dollar
The “dollar is dying” narrative does what every bear narrative does at cyclical inflection points: it trades a kernel of truth for a wholesale conclusion. Yes, the dollar has weakened, and the reserve share has drifted lower. Yes, central banks are buying gold, and China has rearranged its custody footprint. None of those observations is wrong. However, the leap from observation to apocalypse is exactly the leap investors need to consider very carefully before piling into.
The data simply does not cooperate with the “Dollar’s funeral” narrative. With net foreign inflows into U.S. stocks and bonds running near post-COVID highs, and total foreign holdings of U.S. Treasuries just setting a record of $9.4 trillion. The collapse narrative simply has no real support.
There are four things that matter more than headline-dollar print.
First, central bank gold buying is not “leaving the dollar.” Gold is priced in U.S. dollars, benchmarked to the LBMA and COMEX benchmarks, and converted back to U.S. dollars whenever it is mobilized for intervention, collateral, or settlement. Like Treasuries, agencies, or equities, gold on a central bank balance sheet is a dollar-linked reserve asset. Buying gold reduces exposure to U.S. Treasuries as a security type, but it does not reduce exposure to the dollar as the world’s unit of account. It is a portfolio rebalancing decision, not a currency defection.
Second, reserve share and transactional usage are not the same thing. Central banks can diversify into gold, euros, and yuan without meaningfully changing day-to-day dollar demand. One drifts slowly over decades; the other is set by trade invoicing and capital markets plumbing, and the dollar dominates both by wide margins.
Third, there is no viable alternative. The yuan is hamstrung by capital controls and limited convertibility. The euro lacks a unified fiscal backstop. Gold has no yield and no settlement rails. And BRICS itself is not politically unified: India signed a trade deal with the U.S. in February and halted Russian oil purchases weeks later.
Fourth, cyclical decline and structural decline are not the same thing. The dollar is in a cyclical downtrend that fits comfortably inside its roughly 7-to-10-year regimes. That is a trading pattern, not a funeral.
So what should investors actually focus on? Not whether the dollar survives, the flows have already answered that question. Instead, focus on the variables that genuinely move portfolios:
The earnings differential between U.S. and international equities,
Notably, the AI capital cycle, which will pull global savings back toward U.S. assets,
The Fed’s policy path, and
The cost of hedging dollar exposure relative to its realized volatility.
Those are the inputs that change returns. Whether the dollar prints 96 or 102 next quarter will not meaningfully alter the investment case for a diversified, dollar-denominated portfolio. However, the dollar is not collapsing or being replaced; it is simply being repriced. There is a very large difference between the two, and that difference is where investor attention belongs.
Narratives make headlines. Flows make markets. Right now, the flows are still pointing home — and the AI cycle means they likely will for some time.
🔑 Key Catalysts This Week
This is the most consequential week of the year. The FOMC decision, the Q1 GDP advance estimate, and earnings from five Magnificent 7 names: Meta, Microsoft, Alphabet, Amazon, and Apple. It all lands in a five-day window that will determine the market’s direction for the summer. Nothing else comes close.
The FOMC meeting Tuesday-Wednesday is the first catalyst. The rate decision itself is a foregone conclusion, a hold at 3.50–3.75%, but this is a statement-only meeting with no Summary of Economic Projections or dot plot, which makes the language and Powell’s press conference carry outsized weight. The key question is: Does the committee acknowledge that the labor market is deteriorating faster than expected, or does it lean into the inflation-first framing that dominated the March meeting? If Powell uses the word “patient” with respect to cuts, markets read that as “no action until at least September.” If he signals that the balance of risks has shifted toward employment, rate-cut expectations reprice immediately. This may be Powell’s final press conference as Chair before his term expires May 23, making every word a potential legacy statement, with Warsh waiting in the wings.
Tuesday is a collision day unlike anything we’ve seen this cycle. Consumer Confidence at 10:00 AM, which has been deteriorating sharply, with the Expectations component flirting with the sub-80 recession threshold. That number drops just four hours before the FOMC decision at 2:00 PM, followed by Powell’s presser at 2:30 PM. Then, post-market is Meta’s (META) Q1 earnings after the close. Meta’s report is the first read on whether digital advertising spend held up through the March oil shock and tariff escalation, and the $115–135 billion capex guidance for 2026 remains the single largest AI infrastructure commitment in the world.
However, Wednesday is the mega-cap technology “trifecta.” Microsoft (MSFT), Alphabet (GOOG), and Amazon (AMZN) all report after the close. Microsoft’s fiscal Q3 may be the most important tech earnings event of the year. Azure cloud growth guided at 37–38% in constant currency is the AI infrastructure monetization proof point, and Copilot adoption data will tell us whether enterprise AI spend is translating into revenue or stalling at the pilot stage. The stock is down over 8% year-to-date and trading at the cheapest forward multiple since 2017. Alphabet’s Google Cloud (which grew 48% last quarter) and Amazon’s AWS (24% growth, $200 billion in 2026 capex guidance) round out the cloud trilogy. If all three miss on cloud growth, the AI capex cycle narrative cracks; if not, the rally continues.
Thursday is Q1 GDP at 8:30 AM. Q4 was revised down to just 0.5% annualized. The Atlanta Fed’s GDPNow sits at 1.2%, while the New York Fed’s Nowcast is at 2.3%, a historically wide spread that reflects genuine uncertainty about whether the economy is decelerating or stalling. A sub-1% print would ignite recession fears and put immediate pressure on the Fed to cut rates, regardless of where inflation sits. Apple (AAPL) reports after the close on Thursday and caps the Mag-7 wave, with iPhone 17 cycle data, China tariff exposure, and the Tim Cook-to-John Ternus CEO transition all in play.
Here is what to watch. The Fed will tell us what it thinks, GDP will tell us what happened, and the Magnificent-7 will tell us whether the growth premium that justifies their collective $14 trillion market cap. Any two of these could move markets 3%+ in a session. All three in the same week is a vol event. Hedge accordingly.
For now, the market continues to climb a wall of worry, and the technicals say respect the trend. RSI is elevated but not grossly overbought, and while breadth is improving, all moving averages remain green. Our March 200-DMA analysis continues to play out textbook. But Thursday’s reversal on the Tehran headlines was a shot across the bow; this market remains one oil headline away from a 2–3% air pocket. The pullback we flagged last week hasn’t materialized so far, which is making the setup increasingly stretched. New money should wait for a retest of 7,000 or the 50-DMA (~6,979). Stay long but trail stops and take partial profits into BofA’s 7,168–7,206 target.
Trade accordingly.