Alex

[Alex @marlantes]

@antoniogm Mt Livermore Angel Island?

[Alex @marlantes]

@pitdesi $2802.T

[Alex @marlantes]

@EverettRandle Airlines + loyalty credit cards a funny example

[Alex @marlantes]

@jake_zeller 0.5x ntm sales and 8.5x ebitda though!

[Alex @marlantes]

@BillyDraper Especially if you can redeem the 1/1 for buffet style charity lunch

[Alex @marlantes]

@pitdesi @signulll Honestly pretty clever counter positioning for the first to do it. I wonder what other industries this could work in (rhymes a bit with Costco and Amazon prime)

[Alex @marlantes]

@atelicinvest I’m in the camp that this will be a tailwind for a section of the market, but I think a credible argument is that software tends to be deflationary and an increase in lower priced substitute goods accelerates that. Reduced long term pricing compresses TV.

[Alex @marlantes]

@Deenobrown123 @TheValueist I hear you on cyclality, but quality of earnings and capital intensity pretty different vs say COST

[Alex @marlantes]

@teddy_okuyama Hi Teddy, I just met with an endowment who is potentially interested in finding a manager who focuses on small gap growth software in Japan. Would you be able to connect us?

[Alex @marlantes]

@pitdesi @chamath @theallinpod @MECEMike It depends how you execute the trade, but in this scenario I think you'd get a rebate. E.g. you short $QQQ with a cost of borrow of 25bps, but are going to get a cash rebate of ~3.8% so you'd earn 3.5%+ on the short (delta neutral). Could also use proceeds to fund the long.

[Alex @marlantes]

@RealAssetsValue So in your example the carrying cost of the trade would be ~5.6% a year (+4, -9, -0.6). However when people short they are usually funding a different long and not holding low yielding cash (e.g. they are funding WPC paying 5.75%)

[Alex @marlantes]

@RealAssetsValue Exactly - the margin rate is less relevant here. That's the cost charged to you to borrow USD to go long other assets. This is typically fed fund rate + how the broker prices risk / profit. E.g. fed funds + 50bps

[Alex @marlantes]

Very satisfying explanation

[Tivadar Danka @TivadarDanka]

In machine learning, we take gradient descent for granted.

We rarely question why it works.

What's usually told is the mountain-climbing analogue: to find the valley, step towards the steepest descent.

But why does this work so well? Read on.

[Alex @marlantes]

@robbezdjian Humans don’t do well mentally handicapping exponential outcomes. HHH pays out 3.375 and more than covers the worst 3 tail risk outcomes

[Alex @marlantes]

@DhruvBhalla_ Agentic wallet

[Alex @marlantes]

@nealkhosla @ArfurRock E.g. Sutter Hill incubated $SNOW. I think at IPO their stake was ~$12bn, but their revenue is still below OSCR

[Alex @marlantes]

@nealkhosla @ArfurRock To be fair F500 is curated based on revenue, whereas you might care more about market cap

[Alex @marlantes]

@endowment_eddie Have you seen https://t.co/8M6XBzEp1d?

[Alex @marlantes]

@Billy_Draper @cutsclothing @rhone @buckmasonusa @tenthousandcc Savvy marketers will know cents per running hour > cents per mile