Bad ‘hard’ data creeps in

Digest opened free editor

Rola Khaleda, FT editor, chooses her favorite stories in this weekly newsletter.

Good morning. president Donald Trump threatened customs duties on any country importing Russian oil, hoping to push Russian President Vladimir Putin to end the war he started in Ukraine. Like Trump’s tariff for Venezuelan oil importers, this is mainly an additional penalty for Russian oil, which will press up to the global price. Does oil break $ 90 a barrel this year? Or will Trump will decline if oil prices rise enough to notice his voter? Email us: Robert.armstrong@ft.com and iden.reiter@ft.com.

Personal consumption expenses

In a month of large market movements, it is difficult to know the meaningful moves that are noise. It seems that the decrease S&P 500 percent on Friday was the previous.

The main catalyst was to issue personal consumption expenses that were bad at multiple levels. It was widely expected that the PCE price index was a little hot In the inflation report at the last consumer price index, the components feeding on PCE were higher than expected. PCE exceeded Friday until those expectations. Core PCE increased by 2.8 percent, or 4 percent a month, which is the largest monthly rise since January 2024.

There were other bad signals in the numbers. The personal savings rate by American families for the second month increased to 4.6 percent. The consumption had risen, but with less than expected-although there was a great leap in the consumption of strong goods from the previous month, indicating some front tariffs. A more hot, less spent, and more savings: the smell of this group emanates from the recession, which is a smell that witnesses the market. The Federal Reserve, which is preferred by PCE over the consumer price index as an inflation measure, implicitly at its last meeting that the customs tariff can be stagnant. This seems more likely.

Last week, we asked whether the soft data was bad or when it turns into bad solid data. On Friday, PCE can be an early example. The soft soft data continues in: Friday also appeared in reading the consumer index in Michigan. On the same day, the tracker of GDP in Atlanta in the first quarter – amended to exclude the effect of gold imports – declined to below zero. As we have noticed in the past, the level of GDP is distorted through the treatment of exports. However, the direction of change is related.

(Reich)

Readers’ response to Kuruv

On Friday, I said that Coreweave should be estimated on a cash basis such as a real estate project or Rit, not with an EBITDA double, like the technology company. I also wrote that it was not quite clear to me why the major technology companies will use external sources to calculate artificial intelligence to a relatively small start with high capital costs when they can create data centers with their cheapest capital (which they do, to tens of billions). These arguments sparked some interesting responses.

Iuri Struta Books asking: “If you appreciate Coreweave like Reit, can you Amazon Aws, MSFT’s Azure and Google Cloud’s Alphabet the same way?” A great question, and the answer is yes, to the point that these units do not confuse the value -added services in their service computer offers. If they mix with more services, they are mixed for programs. What I understand is that Coreweave is a relatively light display of services, but, like every other company in space, they want to sell more than electricity and servers, to protect its margins.

Why then invest Big Tech in Coreweave, when they have significant investments in their data centers? The commentator, “independent”, argued that the use of external sources to Coreave reduces the intensity of major technology companies, and that the high use rates in Coreave enhances efficiency that helps compensate for their capital costs. The Independent also argued that Kuruv “provides diversification in a slower use [and] Ease [the] Blow (the lowest models) if these assets suffer from technological statute of limitations. In other words, it is hedging the absorption of artificial intelligence and infrastructure. Full point From Coreave is that its capacity can be trivial without damaging the Big Tech for its customers in technology. If a generation of NVIDIA graphics units spread by Coreaeve become old before their time, Poem From Coreove is that it gives large technology a way to stay away from the problem. So, what focuses on Coreweave on customer contracts in the long run, which seems to undermine the advantages that indicate independent? If Coreave sells flexibility, then why does anyone sign a multi -year agreement or pay with them?

A regular correspondent, Ken Favaro, wrote to argue that I was the best wrong tree by focusing on the cost of higher capital:

The cost of the capital with the project, not the project owner. . . The risky project must have a higher cost of capital than an unprotected project, even if it is sponsored by the actor himself. The implicit idea in your question is that hyper excessive should finance their own devices because they have a lower cost of capital is the same thinking that Gee is in a problem: I think Welch and IMMELT can use the AAA classification to finance the GE CAPITAL growth. But of course, the company’s credit classification decreased because Gepital was investing in the high risk things from the place where the AAA classification came: industrial companies with high advantages.

Favaro then indicated that what is or not is used by external sources should be driven by companies ’capabilities – what is good – and not its cost of capital.

I love this argument a lot. But I point out that large technologies have entered into the ownership of data centers on a large scale, so they clearly believe that they have abilities there, or at least they think they have no choice but to build their own ability, given the amount of what they need. The question is why, given this, they are a little restored to Coreweave on the margin? In addition, I think it is also possible for some adult technology to make a GE mistake by becoming a huge operator of the data center. The very low capital costs may be drowned in entering an area in which the returns may be lower, and the risks are higher, than they believe. If this is true, it would be better if there was a larger Corweef soon, so that the large technology could have grabbed what is good.

One good reading

Libyan oil.

Bodcast is not available

Not enough without destroyer? Listen to our new podcast, to diving for 15 minutes in the latest news on markets and financial addresses, twice a week. Annexing previous versions of the newsletter here.