Search podcasts by meaning, not just keywords.
Run a semantic search for a thesis or sector, like "GLP-1 exposure" or "commercial real estate," and read what experts have said about it, speaker-attributed and timestamped.
Good morning partner,
Few things to cover this morning:
Yesterday’s Session
A new spread extreme is beginning to show up in the AI trade.
What used to be long semis/memory and short software has quickly turned into the hyperscalers vs chips trade, as you can see from yesterday’s session most spenders rallied at the expense of those receiving the spending.
I warned about this relationship being the most important AI divergence of 2026.
Because it spells out a very important dynamic now starting to accelerate:
Rising cost of capital (through higher yields) creates a headwind for capex
Which helps the negative FCF hyperscalers, and hurts receivers like semis/memory.
With the treasury enacting the bonds buyback yesterday, I think markets have taken this as a knee-jerk reaction to the bond market’s efforts to put a stop to the mania before it’s too late.
Problem is…
It may already be too late, and that’s okay.
That’s why I put on this specific trade as a put option spread, the best R:R setup I’ve seen since I called the South Korean selloff.
Let’s take a look at yesterday’s leaders and laggers:
Basic Materials, Healthcare, Cyclicals.
Apart from Targa Resources and other material stocks rallying on earnings, I don’t think this has much to do with the AI buildout, especially as we just saw the spender vs receiver dynamic above.
From Healthcare, this leadership is due to the Moderna rally after announcing a new cancer-treatment drug release.
Where things get interesting is in consumer cyclicals, led by names like Celsius up over 8% as new insider buying takes place.
Lucky for you, that name is part of our consumer portfolio sleeve.
Now focus on the laggards in Industrials, Financials, Technology.
The industrial buildout of AI will take a hit every time hyperscalers go up on the belief of slower capex and better FCF down the line, another confirmation of the spread trade.
For financials, I do think this may be a reaction to the bond buybacks, as it may diminish the future amount of net interest income (NII) generated from these banks.
Technology is self-explanatory from the hyperscaler dynamic spelled out here.
The leading theme is now becoming clearer, growth shock driven by the hyperscaler financing tailwind (sponsored by a bond selloff.)
Noise vs Signal
In yesterday’s FOMC meeting, Kevin Warsh said something nobody was really expecting him to…
Stocks are near dot com valuation levels now.
When you were told that valuation multiples are far from being near the internet bubble, you weren’t entirely misled.
P/E ratios are nowhere near being a concern, as long as you take the reported earnings at face value.
As I’ve stated many times before,
Adjust for non-cash items (like unrealized gains in Anthropic & OpenAI stakes), or the manipulation of depreciation schedules ballooning earnings beyond reason.
And you quickly get to 2-3x the current level of P/E ratios.
In other words,
We are in more of an earnings bubble than a valuation bubble, and many respectable participants - like Goldman Sachs - have gone public about this, there’s just no advertising for it.
Still,
I expect all of the triggers to remain relatively quiet as we run up to a rushed Anthropic IPO.
After that, it’s anyone’s guess as to what can happen.
News
Fed Minutes Show Most officials see rate hikes as needed, as there is no real strategy to keep fighting the bond market in making this call. Honestly, we need hikes to reset the euphoria that keeps driving inflation higher.
Goldman Sachs tests the waters for a datacenter related loan, which will be assigned to CoreWeave. As you now know, these neoclouds along with memory are some of the most exposed industries in the topping capital cycle right now.
South Korean Gamblers pivoted to the US stock market, making the same leveraged bets around AI-related stocks like memory, even after millions of margin calls were issued last month.
Uber Partners With Zipline as its delivery business begins to overtake mobility, a union that could allow for dozens of thousands of deliveries to be made each year. This was part of our thesis when we pitched it on the Offside Portfolio.
Movers & ES Levels
Honeywell 📈 Gained 6% after a recent upgrade from Morgan Stanley, recall my last PMI breakdown where I explicitly told you gains were about to show up for the aerospace and defense industry.
Celsius Holdings 📈 Closed higher by 8% after a few insiders bought the stock and some leadership changes were made, boosting my original thesis behind buying the stock for the Offside Portfolio, happy to keep owning it.
Nebius 📉 Lost close to 10% as the company showed a need to raise $4.5 billion for new data center construction, amplifying the fact that these neoclouds are now heavily exposed to a topping capital cycle.
La-Z-Boy 📉 Slipped over 16% after reporting weaker results, a direct tie to the slowdown in housing and the bottom K of the consumer economy.
Now let’s get into some ES levels for today.
** Due to low volatility, yesterday’s take on the index remains the same for today
$7,650 - $7,800 remains the base to trade for now, until we see the potential September effect on volatility (which so far is right on schedule.)
We continue to see a “sell the rally” behavior currently, with the latest evidence shown at $7,740 and $7,720.
Should be see more aggressive selling activity at or below $7,700 and I suspect we may as well reach the lower end of the current base at $7,650.
What happens after that will be up to the momentum factor (AI) and how the components around memory and neoclouds continue to behave.
For now I wouldn’t have too strong of a view on the index unless these levels break on renewed aggressive offering from the sellers.
More importantly, we should continue to see passive bidding from the buyers in order to keep prices below that $7,800 turn distribution I mentioned on Monday.
Know that we are at an extremely important cutoff here at $7,700.
Below that we find a 150-point distribution running from $7,400 to $7,550 which I believe could trade and consolidate should we see that aggressive offer from sellers come in.
Portfolio
New all-time high for the portfolio at 14.3% on the quarter.
For the first time since the portfolio’s inception last quarter, our cash balance has gone below 50%.
The reason is I decided to pull the trigger on a name I’ve been watching for over a month now, and just yesterday it gave me a very clear signal of constructive action.
By luck or design, the stock went up close to 2% after I alerted Premium members of my buying activity.
This is a name designed to win when government spending comes to save the day if/when rising bond yields put a stop to the AI investment wave.
By the way, there’s over 100% upside underwritten for this company.
Most members have 15x their membership costs to Offside Premium over the past quarter.
Financial growth starts at finding these compounding deals and letting the system run.
Here’s the positioning update for today’s premium members:
Subscribe to our premium content to read the rest.
Become a paying subscriber to get access to this post and other subscriber-only content.
Upgrade