EndGame Macro
Data Centers…Demand Is There, Permission Is Not
This isn’t about demand slowing for data centers. Demand is still there. What’s changing is the politics around them. DeKalb County’s extension of the moratorium is really about local pushback finally catching up with an industry that grew fast and quietly. Residents are focused on the tangible costs like noise, diesel generators, power and water usage, quality of life issues and local governments are responding by slowing things down so they can rewrite the rules.
Why the business model gets harder
For companies building and operating data centers, time suddenly matters a lot more. Delays push revenue out while costs keep running. New projects now face more permitting risk, more legal and community negotiations, and more required spending to mitigate environmental and health concerns. On top of that, the generous tax incentives that used to smooth approvals are increasingly under scrutiny, which compresses returns even further.
The irony is that constrained supply can lift pricing for existing, approved facilities, but it also makes growth less predictable. What used to be a fairly straightforward infrastructure play is turning into a capital intensive, politically exposed business. Going forward, the winners are likely the operators who already have sites permitted and plugged into the grid and the losers are the ones counting on frictionless expansion that no longer exists.
tweet
Data Centers…Demand Is There, Permission Is Not
This isn’t about demand slowing for data centers. Demand is still there. What’s changing is the politics around them. DeKalb County’s extension of the moratorium is really about local pushback finally catching up with an industry that grew fast and quietly. Residents are focused on the tangible costs like noise, diesel generators, power and water usage, quality of life issues and local governments are responding by slowing things down so they can rewrite the rules.
Why the business model gets harder
For companies building and operating data centers, time suddenly matters a lot more. Delays push revenue out while costs keep running. New projects now face more permitting risk, more legal and community negotiations, and more required spending to mitigate environmental and health concerns. On top of that, the generous tax incentives that used to smooth approvals are increasingly under scrutiny, which compresses returns even further.
The irony is that constrained supply can lift pricing for existing, approved facilities, but it also makes growth less predictable. What used to be a fairly straightforward infrastructure play is turning into a capital intensive, politically exposed business. Going forward, the winners are likely the operators who already have sites permitted and plugged into the grid and the losers are the ones counting on frictionless expansion that no longer exists.
DeKalb County extends data center moratorium amid public health concerns
#MacroEdge tweet
X (formerly Twitter)
MacroEdge (@MacroEdgeRes) on X
DeKalb County extends data center moratorium amid public health concerns
#MacroEdge
#MacroEdge
Offshore
Photo
EndGame Macro
Why Full Exposure Feels Good Until It Doesn’t
This is a positioning story. Retail investors are already heavily in stocks, and professional managers aren’t hanging back either, they’re close to fully invested. When everyone is leaning the same way, the market doesn’t get stronger, it gets more fragile. There just isn’t much new money left to push prices higher if something changes.
Why managers end up fully invested
It’s about incentives. Managers are judged against benchmarks and peers, constantly. Being cautious too early is usually worse for a career than being late. So they stay invested, ride what’s working, and assume they’ll dial back risk once conditions shift. The problem is that markets don’t usually give you a slow, polite warning. When things change, they tend to change fast.
Why this matters for retirees
For retirees, this doesn’t feel like being all in. It feels diversified…spread across funds, accounts, and strategies. But when managers across the system are near max exposure, those distinctions matter less. When volatility spikes, risk isn’t reduced ahead of time. It’s reduced after prices are already falling. That means selling into weakness and locking in losses at exactly the wrong moment when stability matters most.
We’ve seen this setup before
No major financial crisis including 1907, 1929, 2000, or 2008 started with managers defensively positioned. They all started with confidence, crowding, and high exposure. Risk only came off after liquidity broke, volatility surged, and correlations snapped higher.
This kind of setup isn’t a timing signal. It’s a risk signal. When positioning is this full, upside gets harder and downside tends to arrive faster than people expect, not because everyone is wrong, but because everyone is already in.
tweet
Why Full Exposure Feels Good Until It Doesn’t
This is a positioning story. Retail investors are already heavily in stocks, and professional managers aren’t hanging back either, they’re close to fully invested. When everyone is leaning the same way, the market doesn’t get stronger, it gets more fragile. There just isn’t much new money left to push prices higher if something changes.
Why managers end up fully invested
It’s about incentives. Managers are judged against benchmarks and peers, constantly. Being cautious too early is usually worse for a career than being late. So they stay invested, ride what’s working, and assume they’ll dial back risk once conditions shift. The problem is that markets don’t usually give you a slow, polite warning. When things change, they tend to change fast.
Why this matters for retirees
For retirees, this doesn’t feel like being all in. It feels diversified…spread across funds, accounts, and strategies. But when managers across the system are near max exposure, those distinctions matter less. When volatility spikes, risk isn’t reduced ahead of time. It’s reduced after prices are already falling. That means selling into weakness and locking in losses at exactly the wrong moment when stability matters most.
We’ve seen this setup before
No major financial crisis including 1907, 1929, 2000, or 2008 started with managers defensively positioned. They all started with confidence, crowding, and high exposure. Risk only came off after liquidity broke, volatility surged, and correlations snapped higher.
This kind of setup isn’t a timing signal. It’s a risk signal. When positioning is this full, upside gets harder and downside tends to arrive faster than people expect, not because everyone is wrong, but because everyone is already in.
Investment Managers now have close to 100% exposure to the stock market 🚨🚨 https://t.co/tFuKfiE14s tweet
Offshore
Video
The Few Bets That Matter
Imagine the size of demand if retail image generation forced a major player to shift R&D compute R&D to production.
A bubble? Hm...
We have so many features, so many products we wanna launch, that get held back because we don't have enough compute
https://t.co/WZKzLyFVNz
tweet
Imagine the size of demand if retail image generation forced a major player to shift R&D compute R&D to production.
A bubble? Hm...
We have so many features, so many products we wanna launch, that get held back because we don't have enough compute
https://t.co/WZKzLyFVNz
tweet
Offshore
Video
EndGame Macro
https://t.co/VMNQAOybUq
tweet
https://t.co/VMNQAOybUq
Fed's Waller: We are not seeing the job market go off a cliff tweet
Offshore
Photo
Fiscal.ai
Shares of Roblox are now down 23% since they added 40 million DAUs in a single quarter.
$RBLX https://t.co/vUcex9YYQM
tweet
Shares of Roblox are now down 23% since they added 40 million DAUs in a single quarter.
$RBLX https://t.co/vUcex9YYQM
tweet
Offshore
Photo
Dimitry Nakhla | Babylon Capital®
Oracle is now down 40%+ from its highs
After peaking at ~46x earnings, $ORCL multiple has been nearly halved
That’s the risk of chasing stocks at euphoric valuations — when sentiment cools, multiple compression can do far more damage than fundamentals
$ORCL balance sheet 🚨 https://t.co/D6F5yFHBcD
tweet
Oracle is now down 40%+ from its highs
After peaking at ~46x earnings, $ORCL multiple has been nearly halved
That’s the risk of chasing stocks at euphoric valuations — when sentiment cools, multiple compression can do far more damage than fundamentals
$ORCL balance sheet 🚨 https://t.co/D6F5yFHBcD
tweet
EndGame Macro
Why JPMorgan Is Locking In Yield Now
JPM is moving money that earns whatever the Fed pays today into Treasuries that lock in a yield for longer. It’s less about making a call on the next meeting and more about getting set for the next phase of the cycle.
What they expect from the Fed and the economy
In my opinion JPM sees rates drifting lower and growth slowing, not reaccelerating. In that world, cash stops being attractive pretty quickly. Every cut makes reserves less valuable. Treasuries, on the other hand, do two things at once..they hold their yield and they tend to rise in price when the Fed leans easier. That’s a good trade if you think policy support is about stability, not reflation.
There’s also a plumbing angle. When the Fed steps away from QT, it’s usually because the system needs breathing room. Banks that remember 2019 and 2008 know how fast liquidity can matter again.
Why history backs this up
JPM has been here before. In 2019, small reserve changes turned into a repo shock. In 2008, the institutions with clean collateral and flexibility survived while others scrambled. And in 2022, loading up on duration too early was a painful lesson many banks learned the hard way.
This is JPM preparing for a softer economy, easier policy, and a market that starts caring less about squeezing returns and more about staying liquid. When a bank this size makes a move like this, it’s usually about being early to what matters next.
tweet
Why JPMorgan Is Locking In Yield Now
JPM is moving money that earns whatever the Fed pays today into Treasuries that lock in a yield for longer. It’s less about making a call on the next meeting and more about getting set for the next phase of the cycle.
What they expect from the Fed and the economy
In my opinion JPM sees rates drifting lower and growth slowing, not reaccelerating. In that world, cash stops being attractive pretty quickly. Every cut makes reserves less valuable. Treasuries, on the other hand, do two things at once..they hold their yield and they tend to rise in price when the Fed leans easier. That’s a good trade if you think policy support is about stability, not reflation.
There’s also a plumbing angle. When the Fed steps away from QT, it’s usually because the system needs breathing room. Banks that remember 2019 and 2008 know how fast liquidity can matter again.
Why history backs this up
JPM has been here before. In 2019, small reserve changes turned into a repo shock. In 2008, the institutions with clean collateral and flexibility survived while others scrambled. And in 2022, loading up on duration too early was a painful lesson many banks learned the hard way.
This is JPM preparing for a softer economy, easier policy, and a market that starts caring less about squeezing returns and more about staying liquid. When a bank this size makes a move like this, it’s usually about being early to what matters next.
JPMORGAN PULLS $350 BLN FROM FEDERAL RESERVE TO BUY UP TREASURIES - FT tweet
X (formerly Twitter)
First Squawk (@FirstSquawk) on X
JPMORGAN PULLS $350 BLN FROM FEDERAL RESERVE TO BUY UP TREASURIES - FT
Clark Square Capital
RT @BerkelKip: Any small US fund managers recommend their auditor? Mine got acquired by a larger firm last year and jacked prices...
tweet
RT @BerkelKip: Any small US fund managers recommend their auditor? Mine got acquired by a larger firm last year and jacked prices...
tweet
Offshore
Photo
EndGame Macro
When Oracle’s AI Story Starts Showing Up in CDS
The market is starting to look at Oracle less like a pure software company and more like a company taking on a massive, long dated infrastructure build. AI data centers sound exciting, but they’re capital heavy, slow to monetize, and extremely sensitive to financing terms. When news breaks that a major funding partner steps away or renegotiates, it forces investors to ask tougher questions about how these projects actually pencil out.
Why the CDS move matters more than the stock
The real signal is Oracle’s credit default swaps pushing above 150 basis points. CDS is the price of insuring Oracle’s debt. When that cost jumps, the credit market is saying it sees higher risk ahead. Not default tomorrow, but more leverage, more execution risk, and less margin for error.
Credit tends to get there before equity because it has to. Equity can live on optimism and long term stories. Credit lives on cash flow, balance sheets, and timing. Once spreads widen, funding gets more expensive, deal terms get tougher, and the math on big projects gets harder. That’s when confidence starts to wobble.
My View
This is a reminder that the AI build out isn’t free. When financing friction shows up, the market stops pricing upside first and starts pricing risk first. CDS widening tells you the market wants to be paid more to underwrite that risk and that usually happens when expectations shift from this will work to show me it can.
tweet
When Oracle’s AI Story Starts Showing Up in CDS
The market is starting to look at Oracle less like a pure software company and more like a company taking on a massive, long dated infrastructure build. AI data centers sound exciting, but they’re capital heavy, slow to monetize, and extremely sensitive to financing terms. When news breaks that a major funding partner steps away or renegotiates, it forces investors to ask tougher questions about how these projects actually pencil out.
Why the CDS move matters more than the stock
The real signal is Oracle’s credit default swaps pushing above 150 basis points. CDS is the price of insuring Oracle’s debt. When that cost jumps, the credit market is saying it sees higher risk ahead. Not default tomorrow, but more leverage, more execution risk, and less margin for error.
Credit tends to get there before equity because it has to. Equity can live on optimism and long term stories. Credit lives on cash flow, balance sheets, and timing. Once spreads widen, funding gets more expensive, deal terms get tougher, and the math on big projects gets harder. That’s when confidence starts to wobble.
My View
This is a reminder that the AI build out isn’t free. When financing friction shows up, the market stops pricing upside first and starts pricing risk first. CDS widening tells you the market wants to be paid more to underwrite that risk and that usually happens when expectations shift from this will work to show me it can.
BREAKING: Credit default swap spreads on Oracle's, $ORCL, debt have surged above 150 bps.
Investors are selling the stock over concerns around the company's ability to deliver on recently announced data center buildouts. https://t.co/40SIicXLJd tweet
AkhenOsiris
$NET $JD
Cloudflare and JD Cloud announced an expansion of their partnership to create a global platform for AI inference workloads. The companies project the collaboration will reduce latency for AI inference by up to 80 percent.
The partnership aims to connect global developers to the China market and Chinese developers to international markets through a unified networking experience. The integration allows customers to activate their presence in China without changing code, with traffic automatically routed to local JD Cloud data centers while other traffic uses Cloudflare locations worldwide.
tweet
$NET $JD
Cloudflare and JD Cloud announced an expansion of their partnership to create a global platform for AI inference workloads. The companies project the collaboration will reduce latency for AI inference by up to 80 percent.
The partnership aims to connect global developers to the China market and Chinese developers to international markets through a unified networking experience. The integration allows customers to activate their presence in China without changing code, with traffic automatically routed to local JD Cloud data centers while other traffic uses Cloudflare locations worldwide.
tweet