The state pension increased under the Triple Lock today. Almost no one understands how the system behind it actually works, why it’s failing, or why politicians refuse to do anything about it. I’ve written a full white paper on it, published today in the hope it finds some grounding. open.substack.com/pub/wossca…

Lot of focus on Anthropic's $42B loss in their (reported) IPO numbers. Having read the highlights on the train this morning that's not the key number for me. ~$8B operating loss and ~$4B in revenue on a $2T valuation with ~$518B in forward commitments and 2 customers ~25% of sale sis the "WTF" for me. For anyone stuck on the $42B: most of that is a non cash mark on financing instruments because the valuation went up. The operating number is the one that matters. My initial thoughts: 1. That's around x430 price to revenue. Mature mega caps generally sit around x8-15, a "hyper grower" x20-40. x430 is a narrative price. "We're destined to become a utility for cognition, not a software company". Which is a bold bet, and a risk for us (look at how utility companies price) 2. That $8B operating loss on $4.6B of sales means the blended cost of training, capacity and serving still swamps revenue. Whether each API token is a loss leader isn't in the highlights specifically, but t does suggest the company as a whole is still paying up for volume and the next model. The bet is hook-in now, pricing power later, once enough products, services and enterprises rely on them. 3. 25% of revenue in two customers? Well, for all us small builders creating around the pace of change Claude supports, with the trade off being we're reliant on the model to manage the codebase, that's a helluva risk. When the prices go up, or the rate limiting comes, that's directional. 4. $518B of forward commitments over the coming years only works if the IPO pricing/follow ons show up, or revenue scales into the 10s/100s of billions. If it doesn't, that's a problem. $20B in cash doesn't fund that commitment, so it looks a bit circular: the IPO is the funding plan. 5. At this stage, with a Capex bubble the likes of which the world has never seen, if that begins to cool a x400+ valuation starts to crack pretty quickly, along with the willingness of hyperscalers (AWS/GCP etc.) to continue to keep stretching receivable shaped compute. I've been looking forward to seeing these numbers for a while. And they're as entertaining as I thought they might be. One of the most interesting thought experiments surrounds the AI productivity boom. Something lots of enterprise level orgs are now down the road with. These numbers support, as I mentioned above, a view that resource may be cheap now, but the business model is that it won't be cheap always. Food for thought. New paradigm or old risk? Pick your poison.

The 10-year finished at 5.184%, the highest close since 2007, and the indices didn't crack. The MOVE rose 19.06%. The VIX didn't. #42 is up. Enjoy and share. wosscapital.substack.com/p/4…

Technical Analysis: A Lesson in Structure. I’m a classic Dow technician, with a small, back-tested overlay of my own. This is a clean example of structure building the narrative. For the first time since the peak, Bitcoin has taken out a labelled W1 major swing high. Can it hold and confirm on the weekly close? If it does, that’s only Point 1 of a 3-point recovery. Next we want a higher-low retest that stays inside the current range. Only then do we consider a primary trend reversal. Process first. The close does the work. Positive for the bulls, not so much for the bears. What do you think?

Quick reminder, this week's report was published yesterday: wosscapital.substack.com/p/4…

Wednesday was the vote that would maintain or question Warsh’s credibility post Jackson Hole. I think safe to say it maintained it. A committee that can hike 25 basis points against the preference of the fiscal authority hasn’t lost the last word on the funds rate and monetary dominance, for now, seemingly remains. wosscapital.substack.com/p/4…

#40 out now. wosscapital.substack.com/p/4…

The hold I said was under threat last week, but was still live on Thursday, is now a residual risk. The market's calling hike. With oil heading north and Oman $17.28 above Brent, it's hard to argue. Warsh's Jackson Hole words backed him into a corner? wosscapital.substack.com/p/4…

People respond to incentives. That's the single most important sentence in economics. The video is on YouTube tonight: piped.video/watch?v=XVbnq_S2… The article is here: wosscapital.substack.com/p/t…

Supporting video for yesterday's article. piped.video/GrmfLE9Jp2I?si=3Y7P…

#39 released yesterday. Oil climbs again as jobs numbers in the US give FOMC a reason to test Warsh's Jackson Hole statement. Read the full article here: wosscapital.substack.com/p/3…

Just saying.

Last week's oil fade didn't stick. Oman is $5.49 above Brent again, and the corridor is still a conversation next to a week of fire. Friday's 162,000 took the labour excuse off the table, and the Strait still isn't open. #39 out now. wosscapital.substack.com/p/3…

#39 all copy-edited and ready for tomorrow. 11:00 as always.

#38 Summary Video Enjoy! piped.video/4h4qcVgDj2E

#38: wosscapital.substack.com/p/3… Reminder that #38 dropped yesterday. Main themes: Warsh didn't take hikes off the table, and NVidia fought back against circular financing claims no-one had made yet. Full detail in article.

Bonus Sunday Ccntent via some testing work I'm doing on code written for the Fasera platform. Hoping to launch the initial product in September. Article used for testing: wosscapital.substack.com/p/i… Enjoy.

Warsh didn't hike, and he didn't take the hike off the table. If inflation isn't moving to 2% "clearly and at sufficient speed," the Fed has "work to do." #38 is live. Enjoy and share. wosscapital.substack.com/p/3…

#38 prepped and scheduled for tomorrow. Another interesting week in the world of markets. Published at 11:00 BST, Sunday 30th. Subscribe now for instant access. wosscapital.substack.com/

Recent article with my thoughts on the "Wealth Tax" debate. wosscapital.substack.com/p/w…