Minseok

Whats more interesting is for a company like google, where tokens are generated using internal hardware, a $5000 per head per month cost is $60,000 per year per employee. If you can make a $300,000 programmer just 20% more efficient, youre breaking even. This is a very good internal ROIC for google.

I heard him speak on another podcast of the same opinions. its was really helpful to understand the limits and usecases of AI models. One of the conclusions I came to that this algorithm is incredibly useful for programming because much of the work is structured and documented (frameworks/source codes/documentation) and source code to train from, but not as much (yet) in other traditional disciplines. He spoke about his CFO who “uses coding to execute their job” being profoundly more productive than those who work in “analog” mode. Analog being Intuitions/ tribal knowledge/ non standardized white collar labour stuff. Majority of human expertise is inside peoples heads and not open for models to train out of. This would be bullish for software based enterprises like google/MS/Amazon whi will reap incredible benefits from worker productivity alone, but not in areas like engineering/law firms/ accounting firms where much of the work is still done by humans in manual ways. Example on point is in the field of mathematics. There is a langauge called LEAN which makes formal mathematical proofs be expressed in a precise language like coding). This enables LLM to be much more effective in generating and verifying proofs. So I see a world where when most job becomes a programming job, we will finally see significant economic gains with AI.

Korean market was highly margined, samsung sk drop was more severe than MU/SNDK. Seeing as the index drop was more severe than other countries, perhaps there is a US invasion of North Korea fear related derisking being played here added to some margin calls and profit taking? Who knows.

The conundrum for me is, if software is going to be free to make, and openclaw was so disruptive that it caused massive drops in SAAS company, why did openAI acqui-hire Peter Steinberger? He said in a podcast he built the thing over a weekend. Cant Sam Altman make a clone during his lunchbreak with some help from his engineers and codex? it seems an open admission that AI tools are not going to be that disruptive.

I didnt run the numbers myself and i forget the podcast episode but i think in aggregate the cloud providers are using a conservative ratio of cashflow for capex. Enough margin to pay for buybacks and dividends with spare. Something like 15-30% of cashflow. If anything this suggest they are saving the full throttle for later when they really feel the fomo. it could have been from one of: The real eisman playbook compound and friends

First hand experience here. My amortization jumped from 30 years to 59 years. Of course it hit the "trigger rate" where I've got contractual obligations to consider, but before the bank could act first, I increased my payment by $1200. My colleague waited to see what contractual obligations the banks might exercise, and they simply increased the payments on him without any formal notification. In both cases we were mortgagers with major banks and we passed stress test when qualifying for the loan. So banks knew the increased payment wouldnt be an issue for us to pay. Besides, 20% of all mortgages may be variable, but only about 25% of home owners have mortgages at all. It means only 5% of homes affected. Even then there is virtually no risk of default due to stress test. Like mentioned above though, if you bought at the april 2022 peak using a third party/private loan going at 10% APR 1 year term loan, you'd see the shit hitting the fan about now. But this period was also a period of historically low sales numbers. There will be distressed sellers, but the numbers are likely to be miniscule. There is a supply shortage anyways. In the GTA these days, homes are still selling fast and around asking price, ther is just not that many out on the market (literally just a couple of homes available in a neighborhood at a time) People simply pay up the increased mortgage payments and refrain from selling. Condo preconstruction projects which is a significant jolt of supply in the hundreds of units still sell out in a few weeks, so it looks like the market is soaking up the extra supply quite readily. Only if when people start loosing jobs and getting paycuts though, its a house of cards waiting to fall down. I dont see it happening though.

SU, CNQ, CVE, STLC 3% each.

That day is already here. Banks are increasing payment on variable rate mortgages without the consent of home owners when the trigger rate is hit. I voluntarily increased my payment at every rate hike for a total of $1250/month but one of my colleagues has had his bank forcebly increase payment by a total of $800/month through the hiking period. This is from a tier 1 bank.

Its the management i suppose. Brand and market: You look to the aritzia shoppers and their psyche, their position in the market is unique and lacking competiton. Other retailers like gap/h&m discount their inventories to oblivion where as Aritzia tries to grow their brand and command the premium. Theyre not quite luxury tho so you still have a large segment to target. You visit their stores and theyre always so crowded. The one local to where I live is doubling the square footage and he demand is so high that they wont make a dent in the sales per square foot. The target is also young women in 20-30’s who are making money for themselves so they dont want to settle for cheap clothes, but at the same time cant afford to fill their closets with LV/channel/hermes. Store expansions: USA has ten times the population of canada. Canada has 68 aritiza stores, USA has 33. I dont see anything fundamentally different about the culture/interests/behaviors among the target market among these two countries, so likely runway to 600+ aritizia stores in US is real. Seeing how stores are placed in canda, the strategic placement of brick and mortar store are well done. In major malls, far apart, and boutique feel. The numbers are in the filings but basically couple of millions invested in a store will net you guaranteed revenue expansions. They have a back log of potential locations they are tackling at a pace they can absorb. online sales: The real upside surprise for Atz this year has been the online sales. They explain in the calls that where ever they open up shop, they see spike in online sales in that area. Its a relatively unknown store in the US so this explains that they are doing a good job building brand equity, and people are liking what theyre seeing. Online sales are surging triple digits some quarters. Women in my family have shopped through their online channels and are liking it very much as well. Conclusion: Its really a play on a growing brand. It makes the stock especially difficult at the same time if you believe in the managemnt I think the reward will be especially great. Over the next decade or so you either get a lululemon/canada goose or failed brands like forever21/american eagle/hollister.

Aritzia ATZ - its a retailer growing 50% year over year, 80% in the US alone where they are still a small player. Very well managed company targetting a niche “everyday luxury” market. Yahoo finance posted PE is 32. Not terribly cheap but market does not think its too expensive, seeing as how recent downturn has not affected it too much (-14% from peak).

On the topic of labour, Just an anecdote here. I started my career in oil and gas industry in canada around 2011 and switched to nuclear 2017. I was only one of a whole cohort of engineers who switched when the oil industry downturn started. When the industry capex has such a close correlation to the job market of your profession, you can feel the ebb and flow of the industry capex without looking at any cashflow statements. When you see your former colleagues, colleagues of colleagues, and former engineering contractors your worked for and their competitors bidding for nuclear jobs, and then you meet new colleagues in Nuclear who used to do fly-in fly-out fort mcmurry gigs in alberta ‘back in the good ol’ days’, you get a sense of how dry the oil capex is. With bleak o&g capex forecast, and decades of new nuclear capex (public money) in plan, many engineers have switched already and there is no sign of going back. Outside of niche and specialized incumbent knowledge, the vast majority of engineering work and skill involving steel/concrete/piping/rotating equpment/ project execution is nearly identical and transferrable, and transfer they have. If oil companies want them back, you now have to pry their fingers off with very competitive salaries. How high? As a baseline, aside from qualitative factors such as nuclear jobs are close to the cities with limited travel requirements. A competent and experienced nuclear engineer can easily push $200,000 simply by willing to work 3 hours north of gta at Bruce Power. Even at OPG east of GTA, mid level engineering managers and project managers can push $200,000. So What would it take to move them to Alberta and fort mcmurry? Away from civilization (and spouse who cant quit work and kids who cant move school) and into the tundras? You have to bring back the $300,000 $400,000 and above salaries just to get started on a negotiation. But nuclear jobs are cushy (unionized at client level, little competition among vendors), slower paced (35 hour work weeks) and most of all project execution in o&g runs circles around anything nuclear does relieving the individual of concentrated accountability for errors in execution in a typical nuclear project ( and private investment demands more accountability than public investment), so the job is relatively stress free. All the while staying close to and spending time with your family. The gap is even wider for blue collar. A qualified Nuclear operator already pushes $300,000, $400,000 salaries. Normally we used to say if your company didn't win the job, you can easily jump ship to the competitor who did. If oil industry is down, you can stay afloat by going over to nuclear for a while. Now with the renaissance of nuclear investments allowed by public sentiment change to ESG-ish mentality, government is allowing for significant capex in new builds and developments, so you have to stay in nuclear to experience the boom times. The cycle seems to have made an unusual turn this time indeed.

https://financialpost.com/commodities/energy/oil-gas/eric-nuttall-fear-is-gripping-energy-markets-but-the-facts-suggest-oil-fundamentals-are-still-strengthening Canadian energy sector averaging 31% FCF yield and on track to be debt free by Q1 2023. The whole industry can privatize itself in 3.1 years! This is on a backdrop of oil demand growth which dont turn negative even in hard times, just slower. Hypothetical inventory surplus of 1M BPD is only enough to replenish historical stock piles over a whole year. Canadian TSX capped energy index ETF seems like a good hold in these times.

https://apple.news/AZfDc8w_9TLyTaa9W-fAKIw gasoline demand falling but diesel demand is robust. refineries in US are still shutting. and heavy canadian crude with high diesel distilates should bode well. Dont know how material it is to the thesis but might be part of the overreaction in the markets lately.

This is bearish for POS terminal vendors( e.g square).

But the points! Why would I use apple pay if i can get 2% cashback on my mastercard? Most of the merchant fee (2% to banks 0.15% to V/MC) is charged by the banks backing the credit. And most of that is going to consumers in the form of kickbacks. In order for a new payment processing provider to tear down the moat, it needs to first win over the customers. Who will bribe 2% of transaction fee to customers just so they can earn measily 0.15% in transaction fees? So you say they can just copy the business model. Just look at the history of V/MC. Each bank used to have their own cards, compatible with select registered merchants. Industry went through consolation to pool compatible merchants to arrive at two “standards” who are largely compatible. Now we are seeing once a gain fragmentation, but will customers prefer V/MC which are compatible with 99.9% of merchants or X-pay who may only have 20-% of merchants registerd. to whose benefit is it to overtake the duopoly? Certainly not in the customer’s interest. visa and mastercard themselves cant erode their own moats. They are still using a dozen or so character limit for the transaction description because this was the limit of database software of the 60’s. There is just no incentive to overtake V/MC. They can reduce their fee by half to kill all competition and they can operate business as usual.