Neil Sethi NEW account, please refollow

The @neilksethi account was suspended by X for posting some BoA charts. I'm trying to sort that out but don't have high hopes. So, I am restarting with this account. If you could follow and repost, I would appreciate it, so people know where to find me now. cc: @C_Barraud @Chartfest1 @dailychartbook @lisaabramowicz1 @LiveSquawk @MikeZaccardi @profplum99 @RyanDetrick @SamanthaLaDuc @TheStalwart @TheShortBear @TheChartReport @WarrenPies

🚨 BREAKING: Keir Starmer is officially set to resign as Prime Minister Several media reports say he'll outline a resignation timetable on Monday after consulting with his wife and family today

John Authers: Outside the post-pandemic shock, the NFIB’s index of [firms raising average selling prices] is its highest since 1981. [The historic average is 13%].

DB: Positioning in traditional cyclicals (Industrial Cyclicals, Consumer Cyclicals, Financials, Energy, Materials) has fallen to the 8th %ile since 2010, the lowest since 2024.

Firms longer run inflation expectations (5-10 year) though rose to 2.77% from 2.75%. The 2025 peak was 2.83%.

@VISUALDATAT Yeah, just keeps getting more extreme.

@neilsethinew x.com/VISUALDATAT/st…

BofA FMS: In June, 80% of FMS investors think “long global semiconductors” is the most crowded trade, a record high in FMS history. In May, the most crowded trade was also "long global semiconductors” (73%).

In terms of the pipeline of new residential units, total units under construction edged down to 1.266mn (-0.9% m/m, -7.1% y/y), 15k above the least since December 2020, continuing their slow grind lower since peaking October 2022 at 1.714mn. SF under construction was essentially flat at 587k (-0.3% m/m, -5.9% y/y) while MF under construction ticked down to 662k (-1.3% m/m, -8.1% y/y), both just above the least since 2021.

From the Markets Update (neilsethi.substack.com/p/markets-upda…): I believe we set a new 52-week high for volume from the top stock on the Nasdaq Thursday at 5.2bn for penny stock $ADTX. In total we had 6.4bn in the top 3 stocks (also a one-year high) with 13 companies trading over 100mn shares (still down from 16 Tuesday, one of the highest readings over the past year).

The gap in performance between Goldman's AI basket and AI-at risk basket has widened to over 100%.

Goldman updates their World Cup model following the US' solid 2-0 victory over Australia. They note that with the win, and Türkiye's loss, the US is guaranteed to advance to the round of 32 while Turkiye has been eliminated, meaning their game against the US Thursday night "will thus be of little interest for the group standings but Paraguay-Australia should be an exciting and decisive game for positions 2 and 3." "In the round of 32, team USA will be playing against one of the best third-ranked teams on July 1st in San Francisco." Spain and France remain neck-and-neck as the most likely winner while the US has risen to 11th (1.8%).

The Q2 real GDP tracker from the Atlanta Fed was updated Wednesday, and as expected got an upgrade from the stronger than expected retail sales report which boosted the contribution from consumption to +1.87%, raising the estimate to +3.04% from 2.83% the day before, which was the least since it initiated tracking April 30th. The reading remains for now above the blue chip consensus* but that has pushed up to ~+2.2%. Here’s the breakdown of the components as of June 17th and changes from my last update June 16th: Consumption = +1.87% (+0.22%) Nonresidential fixed investment (biz spending) = +1.06% (0.00%) Inventories = +0.36% (-0.02%) Gov’t = +0.31% (0.00%) Residential investment = +0.05% (0.00%) Net exports = -0.62% (-0.01%) GDPNow Forecast: +3.04% (+0.21%) *The top (bottom) 10 average forecast is an average of the highest (lowest) 10 forecasts in the Blue Chip survey. **(as a reminder their Q1 estimate came in -0.75% too low vs the first estimate after their Q4 estimate came in +1.5% too high (both due mostly to not adjusting federal spending appropriately for the shutdown), while Q3 was eight tenths low, but was along with Goldman’s the closest of the trackers for that quarter. Q2 came in just a tenth low (and right in line in Q3 & Q4 ‘24, and a tenth off for Q2 ‘24), but was -1.2% too low for Q1 ‘25)

@Jc71006290 Saw that.

@neilsethinew x.com/i/status/20683… Life comes at you fast.

Goldman: We now assume that Persian Gulf exports normalize to pre-war levels by end of July and Persian Gulf crude production recover by October and see risks to the Mideast supply outlook as two-sided, but skewed to the downside on net. We estimate that this normalization in Gulf exports to pre-war levels might be achieved with a 13mb/d increase in Hormuz flows from current levels to around 70% of pre-war levels. #oott

Saturday Poll. The next 100 points for the S&P?

Yardeni: We were not surprised by the blockbuster increase in retail sales, given that we have been closely following and reporting on the Redbook same-store retail sales index, which continued to accelerate in May (chart).

Goldman: We are moderating our forecast for gold price appreciation for two reasons. First, we reduce our forecast for rate-sensitive gold ETF demand, as our economists pushed back the final two Fed rate cuts earlier this month to June and December of 2027 (vs. December 2026 and March 2027 previously). Second, as the surprisingly hawkish first Fed meeting under Chair Warsh is likely to limit market concerns about DM central bank independence in coming quarters, we now assume stable macro-policy hedges demand for gold (vs. a gradual recovery to early January 2026 levels previously). We now expect the gold price to rise to $4,900/toz by December 2026 (vs. $5,400 previously). Our gold price views remain structurally constructive but tactically cautious with near-term downside risk and medium-term upside risk.

In another piece of evidence that the US might be up against the limit of shale oil production, the EIA notes that the gas-oil ratio (GOR) "has steadily increased over the past five years. In 2025, the GOR averaged nearly 4,200 cubic feet of natural gas per barrel of oil (cf/b), a 16% increase from 2021. "As the region matures, we expect the GOR to continue to increase. As the GOR rises, we estimate that the production growth rate of natural gas will continue to exceed the growth rate of crude oil in the Permian region."

Mark Hulbert: "There’s a 68% probability that the U.S. stock market will be higher at the end of this year. ... "The 68% odds simply reflect the percentage of calendar years since the Dow Jones Industrial Average was created in 1896 in which the Dow rose from July through December. "I have been unable to find any objectively defined subset of years since the late 1800s in which the stock market’s second-half probability was significantly higher or lower than this all-year average. ... "These results don’t mean that news doesn’t matter to the stock market. But it’s only over longer horizons where indicators boast significant forecasting records. "The stock market’s returns over shorter-term time horizons are largely statistical noise. "Consider a statistic known as the 'R-squared,' which measures the degree to which a given stock-market indicator is able to predict or explain the market’s future return. At the one-year horizon, the R-squared is zero for the Buffett Ratio — despite it being “probably the best single measure of where valuations stand at any given moment,” as Warren Buffett claimed a number of years ago. The Buffett Ratio’s R-squared grows to 6% when forecasting five-year returns and a highly significant 29% at the 10-year horizon."

John Kemp in a couple of Friday blog posts (you can subscribe at the link at the bottom) notes that "oil traders and hedge funds increasingly anticipated the imminent re-opening of the Strait of Hormuz for more than a month before the United States and Iran concluded a memorandum of understanding ending their war on June 15." "Hedge funds and other money managers sold Brent futures and options in nine of the ten weeks leading up to the deal, cutting their combined position by the equivalent of 221 million barrels [to below year ago levels but above the same week in 2024]. And they have continued selling post-deal, paring the equivalent of 95 million barrels in ICE Brent futures and options over the seven days ending on June 16, the most in over a year (and fourth most since the start of 2020). "Funds hold a bearish position of 114 million barrels (19th percentile for all weeks since 2011) down from 209 million (54th percentile) the week before the memorandum and an extremely bullish 429 million (92nd percentile) on March 31." That is below this week a year ago, but roughly the same as in 2024.

BoA FMS Hartnett: asked what stage of investment cycle AI stocks currently in, 56% say “boom” (i.e. FOMO = more buyers) vs. 21% say “euphoria” (price/value in danger zone.

As always, nice collection of posts from @C_Barraud. A theme was competition from Chinese AI models. As Christophe notes, "China may not have won the frontier AI race yet but it is clearly winning the usage, cost, and adoption battles... If Chinese models continue to gain usage, they could put pressure on pricing, margins, and the premium attached to closed US models. For me, the narrative that American models will capture everything is becoming too simplistic." Click through for the posts, including an FT article and a couple from ZeroHedge. I thought I'd add another from ZeroHedge from this week: "Anthropic's flagship Claude Fable 5 model scores 60 on Artificial Analysis' Intelligence Index, compared with 44 for DeepSeek V4-Pro. The gap is meaningful on the most demanding reasoning and agentic tasks, but it comes at a steep price: roughly $3.25 per task versus just $0.05 for DeepSeek. "For many everyday use cases, the performance difference may not justify the cost. Think of it as a high-end supercar versus a reliable family wagon. The supercar is undeniably more capable, but for most trips the cheaper option gets you to the same destination at a fraction of the cost."

DB: Our measure of aggregate equity positioning increased from modestly underweight to slightly above neutral (0.05sd, 42nd percentile). Lots more from DB in the Week Ahead. Sub here and you'll get the full report Sunday: neilsethi.substack.com