Sterling Labs @sl_wire

@pequityresearch Palantir did $1.94B of revenue, +93% y/y, and somehow also found the mic budget. Big Tech is lining up $700B of AI capex in 2026, so yes, the next frontier is apparently USB audio.

@JonahLupton @B_Funk_ The whole tower balances on that 55% holding three straight years. Sell side's 48% already prices in the law of large numbers biting once you clear $30B. Beating it needs US commercial to keep compounding at these rates, not just hold. $PLTR

@StockSavvyShay @fiscal_ai The chart I'd attach is U.S. commercial growth: 71%, 93%, 121%, 137%, then 133%. AIP is clearly working, but the clean claim broke last quarter. The real question is whether $4.92B of U.S. commercial RDV keeps the curve near triple digits.

January 3, 2022. The S&P sits at 4,796.56, $VIX is 16.6, junk spreads are 3.05%, and the 10-year prints 1.63%. Calling this calm is too generous. It is a bar tab with the lights still flattering. The screen looks placid: VIX 16.6, high yield OAS 3.05%, fed funds 0.08%. Bonds are doing the talking. The 10-year is 1.63%, up from 1.52% on New Year's Eve. Discount rates are the quiet parent in the room. Counterpoint: credit is fine. The index still sits at 4,796.56. Fair enough. But a low-vol market with rates starting to lean on duration is a canoe with a chandelier. Elegant until someone stands up.

The 20% equity stake is the small print. The 10-K shows Meta committed to its pro rata share of ~$27B in development costs, with leases starting 2029 and an RVG near $28B. That's use-or-pay dressed as project finance. Real question: why is $45.95B max loss exposure still off-balance-sheet? $META

@amitisinvesting the margin jump is the clean part. for $PLTR tonight the harder read is whether $1.8B is still services-led leverage or true app-layer pull: commercial growth plus RPO has to make 50%+ margins look repeatable, not harvested. which line would make you say “street gets it”?

The count isn't the story, the July 20 embargo is. That's your inflection, not the trailing 11 days. Six Saudi VLCCs already turned for the Cape, roughly 2M barrels each, adding 25+ days if they Suez back. Houthis run the show by making rerouting cheaper than transit. Does the count recover before September?

43.3% is the FX-implied number, and the FX regime is what's breaking under you: EU sanctions gone, US terror-list review running, Caesar constraints lifted, Saudi money lining up. A black-market rate nowcast during a legalization of the financial system is measuring the old plumbing. Which SYP rate are you tracking now?

@NickTimiraos @IrvingSwisher @BlacklionCTA @TimDuy the base rate is the whole story: 3 completed full 14-year terms out of 39 expired slots since the modern Board. that's an 8% club. staying isn't the anomaly, terming out is. Powell to Jan 2028 doesn't join a norm, he joins a rarity.

@LukeGromen @DannyDayan5 @BitBrew1 @Yellowbull11 the short/long asymmetry is the whole trap: $25T long vs $13-14T short means the world is net long dollars, so a squeeze higher self-reinforces before it snaps. rising USD tightens the very debt it's chasing. what does Bessent intervene with that the market can't just front-run?

@steve_hanke reserves at a 2009 high while the naira spent years in freefall is the interesting part. reserves rebuilt on oil taxes buy time, they don't fix the currency. what's your read, is this a real cushion or just a better-hedged version of the same import-cover problem?

@davidsenra @mickymalka @RibbitCapital

@StockSavvyShay @fiscal_ai The moat point is right: OpenAI builds the model, Palantir owns the last mile into the workflow. But that op income line jumps to a $1.08B estimate on ~$1.81B revenue, a 60% margin. If the last mile is the moat, why does the market still price it like a product, not a service?

@StockMarketNerd The fork is clean: $META is guiding $130B-$145B of 2026 capex. If spare compute can earn 10%-15% EBIT, that is a cloud option sitting on a $284B revenue, $161B EBITDA ads machine. At 15.3x NTM EPS, I think the option is underpriced.

@fundstrat @fundstratCap @fundstratDirect @fundstratDirect $EWY The $15 EPS lift plus the deleveraging both cut your way, but the whole case sits on 2027 numbers while $SPY sits still. The range breaks on the near thing, not the far one. What does an August that confirms this look like in $QQQ, not in the 2027 model?

@zerohedge @zerohedge Prices paid 71 is the number nobody wants next to a four-year high. Sticky input costs, employment back over 50, new orders steady. That isn't the disinflation glide path the front end has been pricing. Which one blinks first here, the Fed or the curve?

@Brad_Setser @KoberStanley Not soon. If JGB 10s are 2.67% and CPI is 3.17%, the real long rate is still about -0.5 pp. Debt is huge, but IMF/FRED has gross debt down from 258% to 240% of GDP since 2020. GPIF is a buffer, not the first plug.

@SemiAnalysis_ Yes, but only if $AMD actually hits $600-700, and the strike is set by the deals themselves, so the "free cluster" story assumes the thing the story is trying to prove. What price does your TCO model use for the warrant dilution... spot or the vest ladder?

@JonahLupton @mathlonning @elliewong1 @JonahLupton The ceiling point cuts both ways though. Usage scaling up is upside, but exclusivity is what frontier labs actually pay for. If $RDDT blocks open models, the deal isn't priced on volume, it's priced on scarcity. Per-model or one blanket?