MacroVoices #552 David Rosenberg: Navigating The Noise

David Rosenberg · Rosenberg Research

Hyperscalers now borrow long-dated bonds to fund AI capex, competing with government and lifting real rates.

Hyperscalers are dramatically increasing capex commitments, and three of the four biggest are now net free cash flow negative. They can no longer fund the AI spending boom from their cash flows, so they go to the bond market and the long end of the curve in multiple currencies. Corporate credit demand is now competing with government credit demand, which is one reason real rates have risen. Rosenberg contrasts this with the debt narrative: the 10-year yield dipped below 4% in February, when the national debt was $39 trillion, and nobody talked about the debt rising to $40 trillion then.

“These hyperscalers dramatically increasing their CapEx commitments, and now three of the biggest of the four are now net free cash flow negative. So they're going to the bond market, and they're going to the long end of the bond market in multiple currencies to fund their CapEx boom.”at 15:17

David Rosenberg · Rosenberg Research

Reset in Fed rate expectations explains most of the Treasury yield rise since February lows.

Yields have risen more than 100 basis points across the Treasury curve. In February markets priced one or two rate cuts. The Fed then switched to a tightening bias, hiked, and markets now price three or four more hikes. That reset from an accommodative to a tighter Fed explains most of the run-up from the late-February lows and almost all of the rise since Warsh took over in June. The Fed has taken away the cost-to-carry. With such a flat curve, investors have no reason to take duration and hide in the front end and T-bills. Bond investors are not really vigilantes; they are looking at how the Fed will affect the cost of carry in Treasuries.

“The biggest change has been the reset of Fed expectations. That explains most of the run up in yields we've had from the lows in late February.”at 19:53

David Rosenberg · Rosenberg Research

Nov 4 Treasury funding announcement could cut long-end issuance and rally yields, as in late 2023.

Treasury Secretary Bessent may cut back on long-dated coupon issuance, as Janet Yellen did at the November 2023 funding announcement, when she dramatically cut bond and note supply and flooded the market with bills. The 10-year yield then fell about 100 basis points from late October to the end of December. Treasury supply has a huge influence on the yield curve, unlike the Fed. Rosenberg says the economy is priced off the 10-year note, so this matters more than the midterms. He dismisses the $6 billion buyback programme as having little impact on a market that turns over about $1 trillion a day.

“November 4 is a treasury [funding] announcement, and that could be a very big deal. Just remember how powerful that influence was on the yield curve back in the fall of twenty twenty three. We may well see a repeat of that.”at 29:22

David Rosenberg · Rosenberg Research

Has owned many two-year notes, betting the Fed won't deliver the four hikes markets now price.

He has already owned a lot of two-year notes because he doesn't think the Fed will raise rates four times. The market thinks the Fed will have to hike aggressively because inflation is a big problem, but the Fed did not signal that. Markets went from pricing a couple of cuts in February to pricing three or four more hikes. Rosenberg has been early, as usual, and is dipping his toes into more. He wants bond yields to come down, since he is one of the few who have not thrown in the towel. He expects two things: the November 3 midterms and an exit from fiscal excess.

“I've already owned a lot of two year notes because I don't think the Fed's gonna be raising rates four times.”at 33:36

David Rosenberg · Rosenberg Research

Treasury shifting issuance across the curve, not buybacks, is the real lever to cap long yields.

In fall 2023, after a sharp run-up in long-term rates, the Treasury's refunding announcement shifted issuance emphasis across the curve rather than intervening, and the 10-year rallied about 100 basis points in barely two months. Rosenberg expects Treasury Secretary Bessent to do something similar if yields run up again. He argues supply-side issuance changes are far more powerful than the bond buybacks, which he calls a drop in the bucket.

“It was just a shift in emphasis on where they're gonna issue across the curve. ... It's much more powerful than what he's doing on the bond buybacks.”at 35:46

David Rosenberg · Rosenberg Research

Bond market is under-owned and heavily shorted; contrarian buy as inflation fears are overdone.

Rosenberg has not given up on bonds despite the painful sell-off. He sees the bond market as under-owned, with many shorts on the Board of Trade and in futures and options pits, and he compares it to the capitulation signs in the stock market in early 2009 and the summer of 1982. He thinks the market is fixated on inflation, but he sees nothing that points to a pernicious inflationary environment. He notes that inflation is hard to forecast and depends on many variables.

“I'm looking at the bond market right now the same way. I mean, it's been terrible, but it's under owned. There's so many short positions on the board of trade and futures and options pits.”at 39:26

Patrick Ceresna · Big Picture Trading

Bond selloff won't stabilize without a central-bank catalyst; none has come yet.

The 10-year yield is above 5.30% and the 30-year above 5.60%, and the rate of change has accelerated, which is causing anxiety. Many expected an oversold bounce, but the roughly $30 trillion bond market needs a catalyst to stabilize. Ceresna sees no central bank messaging that provides one yet, and he says the stress runs across the whole curve and into all risk pricing.

“...a lot of people are speculating that the oversold condition of these bond markets was gonna warrant a bounce. But when you have a $30,000,000,000,000 freight train moving this way, it needs some form of a catalyst at some juncture here in order to stabilize.”at 55:00

David Rosenberg · Rosenberg Research

Tariff shock never passed through to persistent core goods inflation; core goods CPI is at a four-year low.

Media and pundits warned of a mountain of imported inflation from the Trump tariffs, but it never materialised. The net effective tariff rate rose by the biggest amount since the early-to-mid 1930s, yet there was never a sustained pass-through to final inflation in core goods, where it should have shown up most. A year ago core goods CPI was running at 1.5% year over year; it is now below 0.7%, a four-year low. Rosenberg concludes that someone absorbed the tariff increase (exporters, importers or margins), and inflation is not the big threat it is made out to be.

“a year ago, the year over year trend in the core good CPI was running at one and a half percent. Today, it's running below 0.7%. It's at a four year low.”at 5:06

David Rosenberg · Rosenberg Research

Energy shock won't become sustained inflation unless it feeds into wages, and wages are decelerating.

A diesel and fuel price shock is not self-sustaining unless it passes into wages. Labour is about 30% of retail cost, six times the weight of energy, and wages are what turn a one-off price shock into persistent inflation. Without that feed-through, the shock acts like a tax increase: negative real wages cut real consumer demand, and demand destruction in cyclical goods and services pushes the other 80-85% of the price basket down. Nominal wages are decelerating and the labour market lacks the 2021-22 surge in quits, hiring and openings. NY Fed's John Williams said on September 29 he sees no evidence price shocks are entering wages. Rosenberg will wait for collective bargaining agreements and average hourly earnings to escalate before calling it sustained inflation. He attributes the 2021-23 wage-price spiral to COVID stimulus checks and paying people not to work.

“This is what's critical. Prices have to feed into wages. Because if the price shock doesn't feed into wages, it hits the wall in the labor market.”at 10:45

David Rosenberg · Rosenberg Research

COVID-era stimulus checks accounted for over half the 2020-22 inflation surge from 0 to 9%.

The Biden-followed-by-Trump giveaways, including $2 trillion in fiscal stimulus checks, accounted for more than half the inflation seen going from zero to 9% between 2020 and 2022. Rosenberg ties this to the wage-price spiral, where the government paid people not to work and the checks were spent just as the economy reopened. He also argues that a proposed $5,000 dividend to voting citizens, costing about $1 trillion, cannot solve inflation and repeats the same kind of giveaway.

“the $2,000,000,000,000 fiscal stimulus checks, that accounted for more than half the inflation we saw going from zero to 9% from, say, call it 2020 to 2022.”at 24:49

David Rosenberg · Rosenberg Research

Split government has historically been bad for equities, while bonds like gridlock; 2024 sweep rallied stocks.

After the November 2024 election, the stock market surged and Treasuries sold off. That was not because Trump won but because of a red wave clean sweep, which signalled the one big beautiful bill, with depreciation allowances, tax refunds and no tax on tips. Stocks rallied on the implications for profits, and bond yields soared. The 2016 pattern was similar: stocks surged after a tough year, and bonds sold off. The midterms that bring a split government are not typically good news for equities, while bonds tend to like it because of the gridlock. The 1994 Gingrich split government produced a bond rally.

“And then when you go to the midterms and you get the split government, not typically good news for equities, and bonds tend to like it because of the gridlock.”at 32:54

David Rosenberg · Rosenberg Research

Narrow market: most stocks in correction; rate-sensitive small caps and cyclicals hurting.

Rosenberg says the S&P 500 is misleading because about 20 mega-cap stocks make up 50% of its market cap. The average and median US stock is down 15% from its 52-week high and in correction. Fed-sensitive Russell 2000 stocks are down 10% from their recent peak. Homebuilders, retailers, regional banks, restaurants, hotels, cruise lines and theme parks are all hurting. Only the large caps are impervious to rates, so he rejects the idea that the Fed has no impact and says investors must dig beneath the surface.

“Nobody talks about the fact that the average and median stock in The United States is down 15% from the fifty two week highs. They're at a correction mode.”at 43:23

Patrick Ceresna · Big Picture Trading

Index hides weak breadth; broad correction risk justifies downside protection through Q4.

The S&P 500 is a poor guide to what is happening underneath. The median stock is 16% below its 52-week high, and only about 20% of S&P 500 stocks are above their 50-day moving average, so four in five have broken trend. Equal-weight S&P and Russell 2000 selling is sustained, while the cap-weighted index is held up by mega-caps above key sell triggers. A drop of about 150 S&P points could trigger a gamma flip, rising implied volatility, vol-target and CTA selling worth tens of billions. Ceresna suggests buying the January 2027 SPX 7,430 put, about 3% out of the money and costing just under 2% of notional, as a defined-premium hedge through a consequential fourth quarter with Fed meetings, midterms and Hormuz uncertainty.

“Let's be clear that the S and P 500 is not actually a very good indication of what is actually happening under the surface. We have actually seen a substantial correction happening in risk assets.”at 56:10

David Rosenberg · Rosenberg Research

Energy prices will stay elevated rather than reverse sharply like 1990; inflation momentum stops once energy peaks.

He disagrees with the thesis that refinery damage and a diesel shortage will keep pushing refined product prices higher. In Gulf War one, WTI went from 20 to 40 and fell straight back to 20 once the war ended, but this time he expects prices to stay at an elevated level. Hormuz traffic is higher than at any point since the war began, and Gulf exports, especially via the Saudi East-West pipeline, are ramping significantly, so relief won't come from a big fall in prices. Inflation is a rate of change, not a level. Even if prices stay where they are, the inflation rate falls out of the data over twelve months. All that's needed to stop the momentum is for energy prices to peak, which he thinks will happen.

“I think that the relief won't come from prices coming down significantly. They'll stay elevated.”at 8:31

Patrick Ceresna · Big Picture Trading

Oil may stay range-bound through October; politics favors stability before midterms.

Crude traded above $100 in mid-September, then gave back about half of that advance and settled in the $90s. The market is waiting for a catalyst, such as a peace deal, more flows through the Strait of Hormuz, or escalation. Ceresna argues that the administration has a political motive to keep the situation stable going into the midterms, so the trading range may hold through October, though he is unsure whether an accelerated move comes first.

“Obviously, there's a political driver for trying to keep this situation stable going into the midterms.”at 1:01:19

David Rosenberg · Rosenberg Research

Gold's secular bull intact; current drop is corrective; central-bank diversification still has room.

Rosenberg sees the recent slump as a corrective phase. Real rates have risen and the US dollar has strengthened, and gold is priced in dollars, so this year has been tough after a strong 2025. He cut gold in his model portfolio earlier this year and then began adding back during the correction. The fundamental driver is central banks shifting reserves out of US dollars into gold, which has continued. Gold's share of global FX reserves is about 30%, versus 70% in 1980 and 10% in 1999, so full mean reversion has not happened. The secular bull ends only when central banks stop diversifying.

“The day I see that the central banks say we're done in this diversification, that'll be the day that the fundamental bull market ends, and we're not there yet.”at 47:41

Patrick Ceresna · Big Picture Trading

Gold likely range-bound into elections, retesting $4,000; long-term bull case intact.

Gold gave up the $4,200 area as higher real rates and a rising dollar weighed on it. The hoped-for Fed operation twist and easing messaging from July and August never materialised, and gold has fallen back to near its summer lows. A retest of $4,000 support looks likely, and tight liquidity could push it down to about $3,800. Gold must sustain prices above $4,500 to turn the technical trend back up. Ceresna expects range-bound consolidation into the elections, but he says the long-term bull case is still intact.

“At this point, we were bullish gold, and still long term, the bull case is quite intact.”at 1:02:54

David Rosenberg · Rosenberg Research

One big beautiful bill's stimulus was one-off; no further fiscal stimulus after the November 3 midterms.

The one big beautiful bill was huge stimulus this year through the depreciation allowance and large income tax refunds. Together with the equity wealth effect on high-end luxury spending and the AI spending boom, it helped the economy do fairly well this year. But these were one-offs that cannot be extrapolated. The bill would never have passed without a Republican-dominated Congress. With a likely Democratic House after November 3, spending bills will stall, so there is no more fiscal stimulus in the economy after that date. The end of this fiscal support is expected to cool growth.

“The one big beautiful bill was huge stimulus for the economy this year. ... These were one offs, and they come to an end, basically. There's no more fiscal stimulus in the economy after November 3.”at 26:33

David Rosenberg · Rosenberg Research

After midterm gridlock, historically 80% odds of slowdown, lower inflation and lower bond yields over two years.

Historically, when one-party power after a national election swings to a two-party system with gridlock, the economy slows 80% of the time, inflation goes down 80% of the time, and bond yields go down 80% of the time over the two years to the next election. He expects the November 3 midterms to put Democrats in control of the House, with prediction polls giving more than a 50% chance they take the Senate. That would mean policy gridlock, because spending bills are passed in the House. Gridlock would end six years of fiscal stimulus and, since government is part of the economy, incrementally weigh on aggregate demand. He says he will take those 80% odds.

“80% of the time, historically, when a one party power after a national election swings to a two party system with gridlock, 80% of the time, the economy slows, 80% of the time inflation goes down, and 80% of the time bond yields go down in that two year period”at 26:54

David Rosenberg · Rosenberg Research

Chinese open-weight models show no barriers to entry, adding global competition and disinflation.

Erik Townsend cited Matt Berry's point that cheap, free Chinese open-source models now match frontier models from a few months earlier, letting firms run AI on their own hardware and keep data away from OpenAI and Anthropic clouds. Rosenberg calls this global competition and says there are no real barriers to entry. He expects China to act as a source of global disinflation, as it has in the past, and argues this counters the view that AI has pushed up memory chip prices. He adds that AI productivity gains will be disinflationary too, though further in the future.

“I think that it's a great point, otherwise known as global competition. ... we are seeing in real time that there are no real barriers to entry.”at 37:35

David Rosenberg · Rosenberg Research

US housing is deflating: new home prices down 8.5% and rents falling, pulling inflation lower.

Homebuilders are discounting heavily, and there are still 8.5 months of new-home inventory. New home prices are down 8.5% year over year. Residential rents are still falling because the multifamily pipeline was built on demographic assumptions that no longer hold, since population growth has stalled due to outmigration. Rosenberg says that if rents and home prices were plugged into CPI in real time, shelter would push headline inflation near zero and core well below 2%. He expects this to show up with lags, since shelter matters more than energy in CPI.

“In this $50,000,000,000,000 asset class called residential real estate, it's deflating. ... residential rents are still deflating.”at 41:12

David Rosenberg · Rosenberg Research

Hiring is firm but wage growth isn't accelerating, a disinflationary signal.

The ADP report showed a headline gain of 90,000 jobs, but Rosenberg sees no acceleration in wage growth. He calls the mix of a hot labor market and no wage acceleration an interesting dichotomy. He reads it as disinflationary, and he thinks the market is ignoring it alongside the housing deflation story.

“No acceleration on wage growth whatsoever. It was a nice headline number up 90,000. ... That's very interesting to me that we have a hot labor market apparently, but no acceleration of wage growth.”at 42:21

Patrick Ceresna · Big Picture Trading

Dollar strength vs euro could extend; EUR/USD could fall to $1.10 or $1.08.

The dollar index has broken to 52-week highs while the euro has fallen to 52-week lows. Rate differentials have widened in recent weeks. Europe is hit disproportionately by terms-of-trade stress from the Strait of Hormuz energy shock. With the euro below $1.13, Ceresna sees room for it to fall to $1.10 or $1.08, which would accelerate dollar strength. He links this to funding stresses in global rates and dollar markets ahead of the midterms.

“Now with the euro breaking down below the $1.13 handle, we have a situation where the euro could head down to $1.10 or even [$1.08], which would only accelerate this US Dollar strength.”at 1:00:12