Forty-eight hours that decide the quarter

This week has the heaviest calendar we’ve seen in over a month, and almost all of it is packed into Wednesday and Thursday. The Fed announces its decision on Wednesday, and Kevin Warsh holds his press conference right after. Then 2 hours later, Microsoft and Alphabet report their earnings. On Thursday we get the ECB decision, the first estimate of Q3 GDP, September PCE inflation, and Amazon after the market close. The Bank of Japan then wraps up the week on Friday. Futures are only giving about 10% odds of a hike this week, so a hold is pretty much a done deal, and what we really need to hear is what Warsh says about December, which is priced at around 84% right now. Moreover, all of this is happening less than a week before the US midterm elections on 3 November.

Where we’re starting from

Last week was mostly about how much AI is costing Big Tech. Meta beat on revenue at $64.6 billion, but the quarterly capital spending of $34.9 billion pushed its free cash flow into negative, and on top of that it raised the floor of its 2026 spending plan to $135 billion. The stock fell around 7%, and its credit default swaps hit a record. Then on Thursday, the Nasdaq dropped 2.6%, its worst day since early June. Tesla went the other way and cut its spending plan. For the whole week, the S&P 500 fell 0.5% to 7,623.68, and Nasdaq fell 0.7%, while the Dow gained 0.4% and small caps went up around 2%. The equal-weight S&P has also beaten the cap-weighted index for three weeks in a row now, so money is still in stocks, it’s just moving away from the biggest AI names.

Oil was the other big mover. Houthi strikes on Saudi oil infrastructure pushed Brent to its highest since April at above $110, and WTI went above $100. The US data also stayed firm. Flash manufacturing PMI hit a 53-month high at 56.0, with the selling-price gauge rising to 59.8, jobless claims fell to 194k, and Michigan one-year inflation expectations climbed to 4.7% (highest since May) while sentiment dropped to a five-month low. The 10-year yield eased to around 5.0%.

Theme 1: What does Warsh say about December?

We’re expecting a hold at 3.75-4.00%, so the decision itself shouldn’t move much. The statement won’t help us much either, since Kevin Warsh took forward guidance out of it, so everything will come down to his press conference. The data since the September hike has been pointing in both directions. On the side against another hike, September payrolls came in at just around 29k, and unemployment rose to 4.2%. On the side for another hike, core CPI has been up around 0.3% for two months straight, producer prices jumped 1.3% in September, household inflation expectations are at 4.7%, manufacturing PMI is at a 53-month high, and oil is sitting above $100.

The market has priced this as “pause now, hike in December”, and we need to see if Kevin Warsh goes along with that, pushes back on it, or hints that October was a closer call than the market thinks. I’ll be listening for two things. 1. How he talks about the oil shock. If he calls it a temporary price jump, the Fed can look through it, but if he calls it a risk to inflation expectations, the Fed has to act on it, and with Michigan expectations at 4.7%, the second one is quite possible. 2. Whether he ties any more hikes to the labour market. If he does, the next jobs report pretty much decides December.

Other central banks are also tightening. Bank of Canada and RBNZ decide on Wednesday, the ECB on Thursday, and the BoJ on Friday (expected to hold at 1.25%). The market is leaning toward more hikes from both the ECB and BoJ, so what Lagarde and Ueda say will move the dollar too, along with Warsh.

What would change the story: Kevin Warsh calling the oil shock a threat to inflation expectations, or refusing to rule out more hikes in total. In that case, December becomes close to certain, the market starts pricing hikes into 2027, and the front end of the curve goes up.

Theme 2: Is the economy running hot, or just prices?

Thursday’s data comes out the morning after the Fed meeting, so it’s more of a December story. Q2 GDP was 2.2%, and the Atlanta Fed’s GDPNow is pointing to around 3.5% for the Q3 first estimate. Most of that comes from a near-18% jump in business equipment spending, a lot of which is AI investment. If growth comes in that strong, the Fed has more room to hike again.

Still, last week’s retail data showed us why the headline isn’t enough. Retail sales beat in September only cause petrol got more expensive, and the control group rose just 0.1%. Real consumer spending is under pressure. So if Q3 growth turns out to be mostly AI capex while households are spending less, the economy is more split in two than overheating, and that’s a tougher argument for a hike.

Then there’s September PCE. Core PCE is expected to rise around 0.2% MoM, after 3.0% YoY in August. PCE is the measure the Fed actually targets, so after two hot CPI reports, a soft reading here would carry a lot of weight. Tuesday’s durable goods & consumer confidence and the employment cost index later in the week fill in the rest. I’d keep an eye on the ECI especially, since it tells us if wages are pushing inflation up or cooling down.

What would shift the whole situation: a strong GDP number led by consumer spending instead of business investment, together with core PCE at 0.3% or higher. That would make the economy look properly overheated, and December would be hard to argue against.

Theme 3: Can Big Tech’s AI spending clear the market’s new bar?

After last week, the market has made it clear that revenue growth alone won’t save a stock if spending drags free cash flow below zero, and that’s what happened to Meta. This week, Microsoft, Alphabet and Amazon (around $11 trillion combined) go through the same test. All three have cloud businesses that turn AI investment into revenue, so they have a better chance than Meta to show a return. Alphabet already dropped around 6% last week along with Meta, so expectations are a bit lower.

The numbers I care about most are cloud revenue growth (Azure, Google Cloud and AWS) compared with 2027 capex guidance. If cloud growth speeds up enough to justify the extra spending, the market can start to tell apart the AI builders that make money on it from the ones that just spend. If capex grows faster than cloud revenue, I’d expect the Meta reaction to spread across the whole group. Intel, Samsung and SanDisk also report on Thursday, which gives us the chip-supply side after a rough month for semis.

What would change this entire story: all three raising 2027 capex while cloud growth slows down. Then the Meta selloff stops being a one-company problem and becomes a sector repricing, hitting the biggest weights in the S&P 500.

×

What it means for gold

Gold has the same push and pull as last week, with the Fed meeting now in the middle of it. Oil above $100 and inflation expectations at 4.7% support gold as an inflation hedge. A Fed that locks in a December hike pulls the other way by keeping real yields high and the dollar supported. Since Warsh’s words matter more than the decision, I’d expect gold’s biggest move of the week around the 2:30 p.m. New York time press conference on Wednesday, rather than at the 2:00 announcement.

Two more things to keep in mind. First, the midterms on 3 November bring event risk over the following weekend, so traders may keep their positions light going into Friday. Second, if the market doesn’t like what Microsoft, Alphabet or Amazon report, the AI-credit stress from last week could come back. In a messy selloff, gold can get haven buying, but it can also get sold in the first wave when people need cash.

The one thing to watch

What Warsh says when someone asks him about December. Most of the other releases this week are already priced in or come out after the decision anyway. If he treats the oil shock as something to look through, or something to lean against, that will set the direction for rates, the dollar, and gold heading into the midterms and the next jobs report.

Key economic events this week

Weekly economic calendar

26 – 30 Oct

Date	Time (UTC)	Country	Event
Tue 27	12:30	US	Durable Goods Orders MoM (Sep)
Wed 28	13:45	CA	BoC Interest Rate Decision
Wed 28	18:30	US	FOMC Press Conference
Thu 29	13:15	EU	ECB Interest Rate Decision (Oct)
Fri 30	09:00	DE	German GDP QoQ (Q3)

Dates and times correct at time of publication. See the full economic calendar