@AtlasShrug1 If AI stonks are overvalued but AI itself is going to have meaningful impact on our future... Does this imply memory stonks are overvalued, but physical memory prices will remain high?
@AtlasShrug1@retard3dinvst0r Thanks for explaining.
I see the fundamental case for PM/PGM - long term inflation/devaluation. But copper/uranium is proxy energy/reindustrialization... Do you see that trend lasting? Seems too correlated with AI/data center buildout?
@AtlasShrug1@retard3dinvst0r 5-8 years?! So, a super cycle then...
I'm a new-ish investor, but in the past few years I've heard about the tin, copper, lithium, o&g super cycles. They were cycles alright, cycles that lasted only 6 months or so 🤣 now I'm a super cycle cynic...
What makes this one different?
@SamanthaLaDuc Disclaimer: macro tourist here
It seems the current subdued prices indicate reserve draws are doing their job. Even though the draw is record high, reserve levels are still healthy.
There are no rules that say reserves must be filled back up asap.
So what am I missing?
@sizov_andre I'm a moron tourist trying to keep up with what's happening 🤣
Thank you for your great content!
Still confused. On the one hand, US conditions seem to deteriorate by the day. On the other Eastern Europe seems decent. And Eastern Europe quantities dwarf everything else?
@KingKong9888@JohnHow68155062 Any views on Chinese solar panel manufacturers? If not for years of involution, overcapacity, and high stockpiles, current record exports due to energy sovereignty concerns by most countries should be massively bullish.
@sizov_andre@EmuConspiracy@BorisBartlog Appreciate the response! Indeed govs will keep grains within their borders.
Another question! I read grain prices usually lead fert, not the other way round (fert being demand-pull).
Are things really so bad that fert will lead this time? There's little historic precedence...
@BubleQe@arcresources Why are you not worried about these things?
1. Continued Attachie problems
2. Fcf hit from weak domestic (aeco) prices?
Jkm/ttf spike may not be enough to overcome these, no?
@EmuConspiracy@sizov_andre@BorisBartlog If I were in the milling business (I'm not), and I saw those WASDE numbers:
"No hurry to stock up now, which would raise my inventory keeping costs. If grain prices spike in the future governments will just release their stores to suppress prices to prevent famine and riots."
@sizov_andre@BorisBartlog Why isn't the current grain stock, which is near record highs according to WASDE, sufficient to subdue price movement to the upside?
@Yampeleg At what size does MoE start to make sense? I imagine if the E's are too small, they will be a bunch of retards, and a mixture of retards is...? Although, would it be similar to boosting of weak learners in classical ml?
@Yampeleg@furkangozukara Is this LR schedule dependent? Or do you mean, given any kind of schedule, an expanding batch size is always better?
What are the "best" schedules nowdays?