Records up, confidence down
Week ending
-
01
Reddit rose 11% for doing nothing
S&P Dow Jones said on Thursday that Reddit would join the S&P 500. The shares jumped about 11%. Nothing whatsoever changed about the company between Wednesday and Thursday.
The slot opened almost by accident. AvalonBay Communities, a residential property company already in the index, is being bought by Equity Residential, which is also in it. When one member absorbs another, a gap appears, and the committee fills it.
The reason the price moves is purely mechanical. Every fund that tracks the S&P 500 has to own what the index owns, so the moment a company is added, an enormous number of buyers are obliged to buy it whatever they happen to think of the business. That is buying pressure carrying no new information at all.
It is worth seeing what this proves. We wrote in What the S&P 500 actually is that the index is a curated list rather than a law of nature. This is that same point from the other side: being put on the list is itself worth money, separately from the company being any better than it was the day before.
One detail to keep straight. The announcement came on Thursday and the shares moved on Thursday. Reddit did not actually join the index until the following Tuesday, which belongs to next week.
-
02
A record on Thursday, and the Dow fell anyway
The S&P 500 closed at a record 7,798.99 on Thursday and crossed 7,800 during the session, finishing the week at 7,785.76. That is a third straight weekly gain. Underneath it, the market pulled in two directions.
The Russell 2000, which tracks smaller American companies, rose 3.15% on the week to fresh highs. The Dow Jones Industrial Average fell 0.6%, ending a two week run, held back by healthcare and older industrial names.
This is the most useful habit to build when you read a headline about the market. An index rising does not mean the things inside it rose. It means the weighted average rose, and money can move violently between the parts while the total barely shifts.
Smaller companies tend to do better when cheaper borrowing looks likely, because they carry more debt and less cash than the giants. So a week that made a rate rise look less likely was worth more to them than to anybody else.
-
03
The market set records and people felt worse
The University of Michigan's preliminary August reading of consumer sentiment fell to 51, from 55.2 in July, when economists had expected about 55. In the same week, retail sales dropped 0.6%, the largest fall since May 2025.
The survey noted the largest reductions among older consumers, lower income consumers, and those without a college degree. That detail matters more than the headline number, because it says the mood is not falling evenly.
Hold this next to item two. In the same five days, the S&P 500 set an all time high and the people the economy is made of reported feeling materially worse. Both are true, and they are not in conflict, because they measure different things.
A share price is a claim on a company's future profits. Consumer sentiment is how people feel about their own money right now. A company can be expected to earn more while its customers feel poorer, at least for a while. The gap between those two is worth watching, because historically it does not stay open forever.
-
04
Inflation came in tame, which was the whole point
July consumer prices rose 0.1% on the month, putting the annual rate at 3.4%. Core inflation, which strips out food and energy, ran at 2.5%. Both landed close to what economists expected, and that was enough.
Notice the gap between the two figures. Headline inflation at 3.4% and core at 2.5% means a good deal of the pressure is coming from food and energy, which is consistent with oil sitting at a three week high on continued uncertainty around the Strait of Hormuz.
The reason a dull number moved anything is that it keeps a rate rise off the table. That is now the third week running in which weak or unremarkable economic data has been treated as good news by the market, for the reason we set out in Why bad news about jobs can send the market up.
It is worth saying plainly that 3.4% is not low. It is simply lower than feared, and markets trade the gap between what happened and what was expected rather than the number itself.
-
05
The AI boom started borrowing
A wave of AI related corporate borrowing is crowding out buyers of long dated bonds, and the ten year Treasury yield rose to 4.67%. The buildout everybody has been watching in share prices is now turning up in the debt market.
Until recently the enormous spending on data centres and chips came mostly out of the cash these companies already had. Cash is finite, and the spending is not slowing, so more of it is now being funded by issuing bonds.
When a great deal of new corporate debt arrives at once, it competes for the same pool of buyers as government bonds. To attract them, borrowers have to offer more, and yields across the market drift up. A higher ten year yield then feeds back into mortgages, company borrowing costs and the value of every future profit, which is the chain in Why interest rates move everything.
This is a slow story rather than a dramatic one, and it is the kind that matters most. Borrowing to build changes what happens if the returns arrive late, which is exactly the lesson in How to lose more than you put in, scaled up from one investor to an entire industry.
-
06
Better than expected, and down 5%
Applied Materials forecast revenue above what analysts had estimated, and the shares fell about 5%. The company did better than the published expectations and worse than the real ones.
This happens constantly and it confuses almost everyone the first time they see it. The estimates printed on a screen are not what the share price is holding. The price already contains whatever investors quietly assumed, and after a strong run for chip equipment makers, those quiet assumptions had climbed well past the written ones.
Two companies in the same week showed both halves of this. Reddit rose 11% on news that changed nothing about its business. Applied Materials fell 5% on news that was objectively good. In both cases the price moved on the distance between reality and what was already priced in.
We wrote the whole idea up here: Why good news can crash a stock.
Figures cover the week ending 16 August 2026 and are sourced from public reporting. Educational content, not financial advice. A new recap goes up every week, and on @wealthstratum.