The fastest way to stop the Fed from raising rates: A U.S. stock market "threat of self-destruction"

date
14:21 22/11/2026
avatar
GMT Eight
Bank of America Chief Investment Strategist Michael Hartnett noted in this week's latest Flow Show that the fastest way to stop the Fed from continuing to raise rates is to make the stock market give back all of its 2026 gains.
Bank of America Chief Investment Strategist Michael Hartnett noted in this week's latest Flow Show that the fastest way to stop the Fed from continuing to raise rates is to make the stock market give back all of its 2026 gains. That target level corresponds to 6,845 on the S&P 500, and the index closed at 6,925 last Friday, leaving only 80 points, or about 1.2%, before it is triggered. After maintaining a "risk-off" stance for six consecutive weeks, Hartnett's framework is shifting from prediction to market reality. Last week, U.S. stocks posted their worst weekly performance since "Liberation Day." The S&P 500 fell 8.4% for the week, declined for a seventh consecutive trading day and broke below its 200-day moving average, entering a technical correction; the Nasdaq 100 fell 19.8% from its October record high, nearing the edge of a bear market; the VIX briefly broke above 39. According to Hartnett's statistics, the S&P 500's year-to-date gain has narrowed from 10.9% a week ago to 1.2%. The market has already begun to "do the work" for the Fed. During last Friday's session, the 50-basis-point "shock" rate hike advocated by Fed Governor Waller was removed from the options, the probability of a January hike fell below 60%, and the 10-year Treasury yield fell back below 5%. But Fed officials' rhetoric has not softened: Governor Cook said financial conditions remain too loose, the Philadelphia Fed's Paulson is open to consecutive rate hikes, and the meeting minutes showed that almost all officials believe further tightening is needed. Hartnett's judgment is: if the Fed remains Wall Street's "good friend," then widening credit spreads and falling stocks will end the hiking cycle early, with the trigger conditions being the S&P falling below 6,845 and IG CDX rising above 80 basis points; if the Fed turns into a true inflation fighter after the midterm elections, hard-landing risk will rise, and the market will not be able to price "peak rates" until the policy meetings on March 18 or April 28 of next year at the earliest. "The fastest way to stop the Fed": two numbers and two thresholds Hartnett summarizes this round's logic in one sentence: "The fastest way to stop the Fed is for the stock market to give back all of its year-to-date gains." The S&P 500 closed 2025 at 6,845, a level that becomes the first threshold in his judgment of a policy pivot. The other threshold is in the credit market. Investment-grade CDX broke above 80 basis points last week, and cash investment-grade spreads widened by about 10 basis points in a single week, the widest since May 2025. Pressure on both the equity and bond sides forms the transmission chain he describes as "the market doing the Fed's work." Goldman Sachs gives the same judgment on the mechanism. According to Goldman's view, a stock market decline is the only factor that can stop rate hikes; Goldman's Tony Pasquariello asked last Friday whether the Fed is about to "get serious." The disagreement lies in magnitude. Hartnett believes 6,845 is enough to make the Fed stop; skeptics point out that a Fed whose inflation target has been missed for 67 consecutive months, with three hotter-than-desired CPI readings on its desk, and which has just seen voters punish the ruling party over prices, needs far more than a flat S&P 500 for the year. The 2018 experience was a 20% decline, and the Fed at that time did not have an inflation problem. Dollar outperforms the S&P 500, gold and bitcoin reduced to "positions" Hartnett's year-to-date performance leaderboard shows a clear contraction in risk assets: crude oil up 44.9%, the dollar up 5.7%, U.S. stocks up 4.8%, global stocks up 4.4%, cash up 3.3%; high-yield bonds down 1.2%, investment-grade bonds down 3.0%, government bonds down 3.9%, gold down 15.1%, bitcoin down 29.5%. Within one week, the U.S. stock market's year-to-date gain fell from 10.9% to 4.8% (as of last Thursday), and the Nasdaq turned negative for the year. Crude oil's lead narrowed for a third consecutive week, from 54.0% to 44.9%; the two major "currency debasement hedges" went from "ugly" to "uglier": gold from -10.4% to -15.1%, and bitcoin from -18.9% to -29.5%. The only gainer was the dollar, rising from 4.9% to 5.7%, meaning the dollar has outperformed the S&P 500 in 2026, and the "buy everything except the dollar" trade has been declared dead. As of last Friday's close, conditions had deteriorated further: the S&P is up only 1.2% year-to-date, gold probed $3,500, and bitcoin fell below $60,000, a new low since September 2024. As The Market Ear previously noted, gold deleveraging is far from over. Hartnett uses a chart comparing the dollar with Fed policy inflection points to explain why he cares about "which Fed" will show up. Every policy swing over the past year has corresponded to a high in some asset class: the dovish rate cut on October 29 last year corresponded to bitcoin's high, Trump's nomination of Warsh on January 30 this year corresponded to gold's high, Warsh's dovish misstep at the June 17 FOMC corresponded to highs in semiconductors and South Korea's KOSPI, and the hawkish rate hike on September 16 corresponded to copper's high, the low in investment-grade spreads, and the starting point of the dollar's rally. Conversely, the signal that rate-hike panic has peaked is the mirror image of those moves: the dollar peaks, and bitcoin, gold and the KOSPI bottom. The current situation is that all four have failed to materializethe dollar recorded its strongest weekly close since April 2025, bitcoin fell for a seventh consecutive week, gold returned to September 2025 levels, and the KOSPI broke below 5,000. The only sign of a turn is that last Friday safe-haven flows began rotating from the dollar to the yen. $310 billion of dip buying, 3% betting on a hard landing Hartnett's second key word this week is "clearing event"he is visibly impatient with the word's overuse. His positioning data explains why: without anyone completing a clearing, there will be no clearing event. BofA's November fund manager survey shows only 3% of investors expect a "hard landing," 50% expect a "soft landing," and 42% expect "no landing"; cash positioning is at an extremely low 3.6%; asset allocation is long stocks and short bonds; and inflows into equity funds and ETFs over the past four weeks reached a record $310 billion. Hartnett's conclusion is: prices have capitulated, positioning has not, and before the latter happens it is not appropriate to give an all-clear buy signal. The disagreement is real. Goldman's Prime book shows hedge fund net leverage at the lowest in more than five years, below the low seen on "Liberation Day," the fastest pace of tech stock selling in a decade, and single-stock shorting at the largest scale since June 2022. But hedge funds are not the main holdersthe ones absorbing losses are ETFs, and ETF investors were still buying the dip in AI stocks last week. BofA's Bull & Bear Indicator fell from 7.4 to 7.1, due to large outflows from emerging market debt and junk bonds, wider high-yield and bank subordinated bond spreads, deteriorating market breadth, and futures positioning turning short. The indicator is still closer to "sell" than "panic": global market breadth is in the 18th percentile, equity fund flows are in the 99th percentile, and fund manager positioning is pinned at the 100th percentile. No pressure valve in the bond market: long bonds may post a sixth consecutive year of negative returns For most of this selloff, bonds failed to provide a buffer. Hartnett notes that 30-year Treasuries are heading for a sixth consecutive year of negative price returns, the longest streak since the Vietnam War period of 1963 to 1969. As of the time of his writing, the 30-year yield was only 25 basis points from its recent high, while the Nasdaq had fallen 15%, bank stocks 16%, and crude oil 22%. His core concern is that in the next recession the risk is for yields to rise rather than fall, because debt levels are too high. Last Friday, the 30-year Treasury yield closed at 5.31%, down 18 basis points for the week, and the pressure valve appeared to open a crack. Citadel Securities sees a record 3.5-standard-deviation gap between yields and growth. But there is tension within Hartnett's note: page one says the risk of rising yields in a recession is the greatest, while page three's hard-landing script opens with allocators buying bonds; and last week he also said no one wants to touch Treasuries until yields reach 5.5% to 6.5%the three cannot all be true at once. Soft-landing script: below 7,000 is a buying point Hartnett's soft-landing path is: if before spring inflation falls, employment stabilizes, and productivity gains are confirmed, with earnings growth maintained above 10%, then below 7,000 on the S&P is a buying pointlast Friday's close was already 75 points below that level. At that point a "breadth bull market" restarts, and sectors that did not participate in the AI rally are no longer expensive: the equal-weighted S&P trades at 16x earnings, mid-caps at 15x (30-year average 16x), and non-U.S. stocks at 12x (average 13x). Sequence also matters. His advice is to preserve capital first, wait for panic to become explicit, wait for the narrative to shift from rate hikes to rate cuts, and then buy cyclical "Cs": crypto, chips, copper, consumer, and apartment REITs. The valuation matrix shows that 7,338 on the S&P corresponds to 18x forward EPS of $408, implying 16% earnings growth, close to the market consensus. Last Friday's 6,925 corresponds to 17x, or 17.9x based on $387 under 10% growth. If earnings grow only 5% ($370) and are given a 16x valuation, the matrix gives 5,913; Goldman's Brian Garrett believes that a full replay of Q4 2018 would end at 6,294, exactly 17x the same earnings. Two institutions, two methods, and the bad-case landing zones are highly close. Hard-landing script: from "boom loop" to "doom loop" Hartnett is betting on the bad case. He believes the "hard landing" probability is underestimated and that the market's consensus for a "boom"5% to 6% nominal GDP growth and more than 15% EPS growth in 2027is too high, and both forecasts rely on two weakening supports. The first is household equity wealth. It has shrunk by about $5 trillion since last quarter; although it is still up about $8 trillion for the year, the S&P lost about $1.3 trillion in market value on last Thursday alone. The second is AI capital expenditure: hyperscaler capex is expected to reach $1.3 trillion in 2027, up from $822 billion this year and $412 billion in 2025, while their combined free cash flow will swing from positive $191 billion in 2025 to negative $61 billion this year and further to negative $283 billion next year. Someone must finance that gap, which is why the market has been tense over the past two weeks: the largest IPO in history approached its issue price the day after listing, SpaceX is trying to pay $47 billion for Nvidia chips without "borrowing," Blackstone's data center company failed to sell an $8 billion investment-grade bond, and the ecosystem's "lender of last resort" says it will sharply raise financing thresholds. Hartnett says voters, the bond market, and asset allocators are sending the same signal: slow down data center construction. Voters spoke on November 3, allocators gave a record share citing "overinvestment" in the survey, and the bond market has been applying pressure since September. The two indicators that would confirm a hard landing have never changed: initial jobless claims above 250,000, and investment-grade spreads above 100 basis points. If triggered, the "boom loop" (stock prices rise, wealth rises, spending rises, capex rises, stock prices rise again) will turn into a "doom loop," and allocators will be forced to cut equities and add bonds, Chinese assets (ample fiscal space), the dollar, and assets negatively correlated with AI, such as REITs and real estate. His summary is "long apartments, short chips." In other words, the hard-landing trade is the reversal of the four big trades that defined the 2020s: buy everything except bonds (30-year Treasuries' peak-to-trough price return -62%), buy everything except China (Chinese stocks have fallen 5% a year on average this decade, the world's worst), buy everything except the dollar (bitcoin and gold have compounded at 44% and 23% a year, respectively, until peaking in October 2025 and January 2026), and go all-in on AI (Nasdaq market value has increased by $26 trillion since March 2023). One of these has clearly reversed, the second began on October 6, and the third began to loosen after the bond rebound. Concentration risk remains: the top ten AI stocks still account for 38% of the U.S. market, in the same tier as the Nifty Fifty's 40%, Japan's 44%, and TMT's 41% during the internet bubble, and far below the railway era's 63%. The 2000 script: twelve rallies on the way down While most of Wall Street references 2018, Hartnett goes back to an earlier cycle. He writes that in May the market was still trading "bubble," and now it is trading "post-bubble." The 2000 path was: tech stocks peaked in March, the Fed still raised rates by 100 basis points that year, and afterward the market broadened into "boring sectors"consumer staples, utilities, financials and health care rose 40% to 60%, while tech stocks nearly halved. This is highly similar to the moves since the summer of 2026. What has not yet appeared is 2001: rate hikes caught up with the economy, safe havens failed, and the heroes of 2000, utilities, became the worst-performing sector of 2001. Two details are worth noting: the dollar was strong in both 2000 and 2001; and the Nasdaq produced twelve rallies of more than 10% on the way down, including one of 35% and two of more than 40%. This is a necessary reminder for investors waiting to buy rebounds after breadth rules trigger. A third detail directly addresses Hartnett's two questions: the last rate hike of that cycle was in May 2000, and the Fed cut rates by 475 basis points over the course of 2001. The answer to "when will rate hikes end" is "soon"; but the answer to "will they end before a recession" is no. In the 2018 version, Powell hiked in December, the bottom came before Christmas, and the pivot came in the first week of January; in the 2000 version, the Fed stopped early, and the bear market still dragged on for two years because of earnings problems. The difference between the two is the difference between a positioning accident and a capex collapse. Fund flows: $66.6 billion into stocks, bonds see first outflow since April 2025 In the week through last Wednesday (covering Nvidia's earnings and pre-IPO cash raising, but not the Anthropic listing or the two worst trading days), investors put $66.6 billion into stocks ($73.5 billion into ETFs, $6.8 billion out of mutual funds), $400 million into cash, and $200 million into gold; crypto saw $1.1 billion of outflows and bonds $7.4 billion. Within fixed income, bank loans remained favored, consistent with the allocation logic of investors who still believe in rate hikes; emerging market debt lost more than 1% of assets in a week. For the full year, several 2026 inflows set records: global equities $1.5 trillion (annualized), global bonds $909 billion, U.S. equities $751 billion, U.S. large caps $669 billion, investment-grade bonds $443 billion, Treasuries $243 billion, tech funds $229 billion, and Korea $79 billion; consumer funds saw a record $20 billion outflow. How investors with 3.6% cash positioning see 2027 The survey shows fund managers expect a soft landing, falling inflation, and slower AI capex growth, and entering 2027 they are overweight stocks, commodities, emerging markets, health care and industrials, and underweight cash, bonds, REITs, consumer and energy. The best asset class choices were international stocks (36%) and U.S. stocks (24%), the best index choices were the Nasdaq and MSCI Emerging Markets (21% each), and the best currency was the yen (42%). Their end-2027 targets are: 10-year Treasury yield 4.9%, EUR/USD 1.12, S&P 7,926, gold $4,341. The first was reached last Friday, thirteen months early. The S&P target is 14% above the current price, and the gold target 24% above. For contrarian investors, Hartnett lists the pair trades that worked in 2026: long crude/short crypto, long energy/short consumer discretionary, long semiconductors/short telecom, long small-cap value/short large-cap growth, long cybersecurity/short defense. Annual themes include: war over peace, rate hikes over rate cuts, greed over fear, AI debt issuance over AI job losses, concentration over breadth, and IPOs over buybacks in a week with $107 billion of issuance. If these trades reverse in 2027, fear, breadth and buybacks will make a comeback. This article is reprinted from "Wall Street See" (). GMTEight editor: Jiang Yuanhua.