Hartnett: A Democratic Sweep Will Trigger A Stock Market Rout, And Pop The AI Bubble

The biggest story last week was not the unexpectedly hot jobs report which, unfortunately, will be revised sharply lower next month as the labor market reverts to its deteriorating, AI-enhanced, trendline: Instead, what everyone was - or should have been focusing on - was the bottom falling out of the bond market with global yields jumping to the highest level in 2 decades, to wit: 

• 10Y Treasury yields jumping to 4.81%, near 2008 crisis levels
• 30Y Treasury yields jumping to 5.31%, highest since 2007
   
• Japan 10Y JGB  >3.0%    First time since 1996
• Japan 30Y JGB  4.2%, or 4x the BoJ policy rate
   
• German 10Y Bund  3.38%, post-2011 high
• France OAT-Bund spread 88bps, 2012 crisis highs
• Italy BTP-Bund spreads, 84bps, 2012 highs

A Bloomberg index of global bond yields just rose to the highest since 2007, and is just 1% away from the highest levels this century.

Appropriately, the topic of soaring bond yields is also the kick-off theme of the latest weekly Flow Show (available to pro subs) from BofA's Michael Hartnett, who writes that with a 99% probability the ECB hikes Sept 10th, 53% Fed hikes on 16th, 98% BoJ hikes 18th (per Bloomberg futures pricing), the hikes are coming fast and furious as central banks try to restore credibility to ward off surge in bond yields (which, as we have discussed extensively, is now the biggest threat to AI capex and the K-shaped consumer booms). In light of this, Hartnett says that if the Fed does hike despite stalled payrolls...

... then it will restore credibility and make sure the current "peak yields" don't go higher, it's also why to Hartnett, duration (RTY, XBI, KRE, REIT) keeps working despite surging yields and why "nouveau-leveraged" Mag7s are on the cusp of upside breakout. On the other hand, if the Fed does not hike - as Trump made painfully clear he will not approve - or even merely keeps rates on hold, then all bets are off, as is the Fed's credibility because for all his rhetoric, Warsh will prove to be "just one more of the guys."

Of course, it's not just the Fed: with Trump approval ratings the lowest on record...

... as a plurality of Americans say the most important problem facing the country is "the economy, unemployment and jobs" (followed in distant second place by those who said "threats to democratic values and norms"), Hartnett says that the White House is realizing that $4/gallon gas, 160 dollar-yen, 5% Treasury bond yields are "Maginot Lines" for the US admin, hence policy interventions via FX, bond buybacks, monetary policy (pressure on BoJ to raise policy rate that’s averaged 0.1% this century)...

... and why the policy panic working for now (see the surging Japan yen); or, as Hartnett described a month ago, global markets are subject to “whatever it takes” policies to maintain nominal macro boom and asset price bull...  and why Hartnett says to stay long commodities and debasement hedges, e.g. gold.

To be sure, this observation doesn't exist in a vacuum, and sits neatly inside a coherent set of themes Hartnett has been pushing over  the past several weeks: 

• "Bonds boss the bubble." His view is that long-dated yields - not equity stories - now dictate the AI trade, captured in his line from a week ago that "bonds trade information, equities trade ideas." He argues AI spenders and builders will keep underperforming AI adopters until global 30-year yields fall below 5%, and that the market is currently priced for a "perfect consensus": no landing, no Fed hike, no AI capex cut, and no Democratic sweep (which will inevitably disappoint).
• Stay long commodities and gold. With "whatever-it-takes" fiscal intervention holding down long-end yields, Hartnett has kept commodities and gold as the core inflation/geopolitical hedge. 
• The midterms are the contrarian flip. His base stance is long equities / short bonds, but he carves out a tactical exception: if Democrats look poised to sweep both chambers, a 10%+ equity selloff becomes likely, making bonds the contrarian Q4 buy. Investors have largely shrugged off election risk so far, which is exactly why he sees the asymmetry.
• The AI bubble is "fit to burst." In related commentary he laid out a post-bubble playbook — "long humiliation, short hubris" — favoring long bonds plus defensives (consumer staples, mining/materials, healthcare) over the crowded AI-buildout names, noting hyperscaler free cash flow has turned negative under buildout commitments.

As Hartnett continues to hammer the rising bond yield theme, he next takes a somewhat contrarian view, and notes that the 10-year rolling return from US stocks is 15%, commodities 11%... while Treasuries are -2%, the worst of the past 100 years.

For bond bulls (if any are still left, now that even career bond bull Lacy Hunt turned bearish) this is a good sign: as the next two charts show, negative long-run returns have been a great entry points for stocks in 1939, 1974, 2009...

... and commodities in 1933, 2018.

And while the US midterms are not a “regime change” election like Thatcher/Reagan in 1980, or BREXIT/Trump 2016, a Fed hike, TSY buybacks, signal a rising risk the midterms show the biggest voter priority is “affordability” not lower taxes, faster AI data center expansion... which is why to Hartnett lower Q4 yields remain a very good contrarian play.

Hartnett's latest Flow Show then pivots away from bond yields, and to the main topic of the week, namely the upcoming midterms (appropriately just as we penned "Democrat Sweep? Here Are JPMorgan's Midterm Trades - And Why Gridlock Pays"). The BofA strategist believes that for all the posturing, the midterms are not a “regime change” election, e.g. Thatcher/Reagan in 1980 = end of inflation/start of bond bull, BREXIT/Trump in 2016 = end of globalization = start of commodity bull;

Alas the coming midterms are unlikely to change the trajectory of US government spending (which will keep rising until it is forced to stop); Hartnett views that 2020s as a decade of political populism as MAGA (Reform party in UK) and Democratic Socialists of America (Greens in UK) represent the culmination of post-GFC Tea Party and Occupy Wall St insurgents. 

More importantly, the populists (right or left) are spending a lot to stay popular... which is why 2020s is a decade of fiscal excess, nominal GDP boom (past six years up 63% in US from $20tn to $32tn) and “Anything But Bonds” strategic asset allocations (TSYs up 74% in past six years, from $23TN to $40TN). Meanwhile, as the latest BofA Fund Managers Survey shows, investors are not fearful of midterms saying POTUS governs through Executive Orders not Congress (277 thus far, on track for most since Truman), and say a Democrat sweep is unlikely given tough Senate “map”; when asked about the most likely outcome from midterms in August BofA Fund Manager Survey (see report), 47% said GOP Senate & DEM House, 23% said DEM sweep, 9% said GOP sweep/maintains control of Congress (current GOP Senate majority is 53-47, in House 218-212).

To be sure, the Senate map is tough for Dems: they must flip 4 of 6 most vulnerable GOP seats in North Carolina (current probability of DEM flip = 92%), Maine (69%), Alaska (64%), Ohio (55%), Texas (51%), Iowa (37%); and DEMs must defend vulnerable seats in Georgia (94% = current prob of DEM hold), New Hampshire (84%), Michigan (65%); the key battleground states for investors to watch are Ohio, Texas, Iowa, Michigan.

Note that Wall Street is already focused on Texas Governor race between GOP incumbent Abbott (currently polling 49% according to Real Clear Politics) and his Democrat challenger Hinojosa (45%); the clash is seen as big referendum on AI data center expansion (Abbott was recently forced to announce a data center moratorium to arrest decline in polling numbers).

But as Hartnett's next chart shows, the Democratic sweep likelihood is rising, with Trump's Presidential approval number ranges from 35-40%, significantly below historical average 2 months ahead of midterms (53% as shown below).

Furthermore, the BofA strategist points to the latest Polymarket probabilities, which show odds of a Democratic sweep at 50% (vs., GOP Senate/DEM House at 35%, and a GOP sweep at just 10%).

This matters because for Hartnett, a Democrat sweep is a threat to asset prices: an electoral shift from populist capitalism to populist socialism, means the next big direction in tax & regulation is up not down (and EPS negative), and would be accompanied by policies to lower inflation, healthcare, improve affordability challenge K-shape wealth boom, AI capex boom, stocks "too big to fail" Wall Street zeitgeist.

Additionally, loss of political capital = less ability for Trump to coerce resources, corporations, foreign governments into support for policy priorities of AI war with China, resource monopolization.

Putting all this together, Hartnett says a Democrat sweep = big risk-off: it would lead to a slump in i) stocks (more than 10%), ii) the dollar, and iii) bond yields into year-end, while international stocks outperform on less trade & military wars... but Europe outperforms Asia (loses Trump AI friend); the BofA strategist says the best hedge for a Democrat sweep is short financials & US dollar. In contrast, a surprise GOP sweep (maintain House/Senate) control = big risk-on, and more importantly a green light for AI bubble and positive US dollar (“exceptionalism returns").

Finally, the largely priced-in scenario of a “GOP Senate/ DEM House” translates into more of the same: modest risk-on... “gridlock = goldilocks”.

More in the full BofA Flow Show note available to pro subs.