Offshore
Photo
memenodes
#NewYear2026 plans
tweet
#NewYear2026 plans
i'm gonna be so productive today
also me: grok remove the dress https://t.co/JWBYGBgCoj tweet
Offshore
Video
memenodes
Why are you still single? You don't have a girlfriend?
me in my prime: https://t.co/CSUmyWxJIL
tweet
Why are you still single? You don't have a girlfriend?
me in my prime: https://t.co/CSUmyWxJIL
tweet
Offshore
Photo
memenodes
Everyone's out enjoying New Year's Eve, and you're gooning in your room with Grok https://t.co/5WGP7Ef15F
tweet
Everyone's out enjoying New Year's Eve, and you're gooning in your room with Grok https://t.co/5WGP7Ef15F
tweet
Offshore
Photo
EndGame Macro
H.4.1 Signals: Late Cycle Liquidity, Calm on the Surface, Tightening Underneath
Total Fed assets now sit at $6.58 trillion, up $24.4 billion week over week, but still down about $305 billion from a year ago. QT hasn’t reversed outright, but it has clearly changed character.
The most important shift is where the Fed is holding assets…
• Treasuries: $4.23T, up $23.4B WoW, but down $81B YoY
• Mortgage backed securities: $2.05T, down $2.9B WoW and nearly $195B YoY
QT is no longer about shrinking the balance sheet aggressively. It’s about reshaping it. The Fed is quietly moving toward shorter duration Treasuries while letting MBS continue to roll off.
That matters more than the headline balance sheet number.
A Quiet but Important Detail: Bills Are Doing the Work
Almost the entire weekly increase in Treasuries came from T-bills, which rose $23.1B in a single week. Notes and bonds didn’t change at all.
This isn’t random. Bills…
• Carry less duration risk
• Reprice faster at higher rates
• Generate income more quickly
In plain English the Fed is managing interest rate risk and cash flow, not easing financial conditions.
Liquidity Is Still There But the Shock Absorber Is Gone
Reserve balances ended the week at $2.96T, down $3.5B WoW and down nearly $292B from a year ago.
On its own, that weekly move is small. The issue is what’s no longer there to cushion future drains.
Reverse repos the old pressure release valve have collapsed…
• Total RRP: $327B, down $184B YoY
• Of that, $322B belongs to foreign official accounts
• Only $4.7B remains in domestic “other” usage
That means the domestic ON RRP facility is effectively empty.
So going forward…
• Treasury cash builds
• Currency demand spikes
• Deposits shift
…and those pressures hit reserves directly, with much less buffer than in 2022–2023.
This doesn’t cause stress by itself. It raises sensitivity.
Discount Window Use: Not a Crisis, But Not Nothing
Total Fed loans are still small $9.2B but context matters…
• Loans are up $301M WoW
• Up $6.9B year over year
• Nearly all of it is primary credit (discount window)
That’s not panic level borrowing. But it is a sign that some institutions are choosing the Fed over private markets at the margin.
Historically, the signal isn’t the level, it’s persistence. If this fades after year end, it’s noise. If it keeps climbing, it becomes information.
Liquidity Swaps: A Flicker, Not a Fire
Central bank dollar swaps rose to $481M, up $392M WoW.
In absolute terms, this is tiny. In directional terms, it’s worth logging.
When swaps move off zero, it usually reflects localized offshore dollar tightness, not global stress. This is a watch item, not a warning siren.
The Accounting Reality Everyone Ignores
Buried in the liabilities is a line most people wave away…
Earnings remittances due to the U.S. Treasury: –$242.1B
That negative number is the Fed’s deferred asset, the accumulated shortfall from paying high interest on reserves while holding a low coupon bond portfolio.
This doesn’t impair the Fed’s ability to operate. But it does tell you…
• High rates are expensive
• Time matters
• The Fed is incentivized to manage income carefully
That helps explain the shift toward bills and balance sheet stabilization.
So What’s the Actual Risk Signal?
Several things narrow the margin for error…
• Reserves are lower YoY and more exposed
• The ON RRP buffer is gone
• Discount window usage is rising modestly
• Liquidity management is becoming more deliberate, not looser
At the same time…
• The yield curve is positive
• Emergency facilities are quiet
• Asset composition remains conservative
This is late cycle balance sheet behavior, not crisis behavior.
Bottom Line
Recession risk implied by the balance sheet alone will be very dependent on how labor, Treasury issuance, and funding markets evolve from here.
Interactive guide to our weekly #Balanc[...]
H.4.1 Signals: Late Cycle Liquidity, Calm on the Surface, Tightening Underneath
Total Fed assets now sit at $6.58 trillion, up $24.4 billion week over week, but still down about $305 billion from a year ago. QT hasn’t reversed outright, but it has clearly changed character.
The most important shift is where the Fed is holding assets…
• Treasuries: $4.23T, up $23.4B WoW, but down $81B YoY
• Mortgage backed securities: $2.05T, down $2.9B WoW and nearly $195B YoY
QT is no longer about shrinking the balance sheet aggressively. It’s about reshaping it. The Fed is quietly moving toward shorter duration Treasuries while letting MBS continue to roll off.
That matters more than the headline balance sheet number.
A Quiet but Important Detail: Bills Are Doing the Work
Almost the entire weekly increase in Treasuries came from T-bills, which rose $23.1B in a single week. Notes and bonds didn’t change at all.
This isn’t random. Bills…
• Carry less duration risk
• Reprice faster at higher rates
• Generate income more quickly
In plain English the Fed is managing interest rate risk and cash flow, not easing financial conditions.
Liquidity Is Still There But the Shock Absorber Is Gone
Reserve balances ended the week at $2.96T, down $3.5B WoW and down nearly $292B from a year ago.
On its own, that weekly move is small. The issue is what’s no longer there to cushion future drains.
Reverse repos the old pressure release valve have collapsed…
• Total RRP: $327B, down $184B YoY
• Of that, $322B belongs to foreign official accounts
• Only $4.7B remains in domestic “other” usage
That means the domestic ON RRP facility is effectively empty.
So going forward…
• Treasury cash builds
• Currency demand spikes
• Deposits shift
…and those pressures hit reserves directly, with much less buffer than in 2022–2023.
This doesn’t cause stress by itself. It raises sensitivity.
Discount Window Use: Not a Crisis, But Not Nothing
Total Fed loans are still small $9.2B but context matters…
• Loans are up $301M WoW
• Up $6.9B year over year
• Nearly all of it is primary credit (discount window)
That’s not panic level borrowing. But it is a sign that some institutions are choosing the Fed over private markets at the margin.
Historically, the signal isn’t the level, it’s persistence. If this fades after year end, it’s noise. If it keeps climbing, it becomes information.
Liquidity Swaps: A Flicker, Not a Fire
Central bank dollar swaps rose to $481M, up $392M WoW.
In absolute terms, this is tiny. In directional terms, it’s worth logging.
When swaps move off zero, it usually reflects localized offshore dollar tightness, not global stress. This is a watch item, not a warning siren.
The Accounting Reality Everyone Ignores
Buried in the liabilities is a line most people wave away…
Earnings remittances due to the U.S. Treasury: –$242.1B
That negative number is the Fed’s deferred asset, the accumulated shortfall from paying high interest on reserves while holding a low coupon bond portfolio.
This doesn’t impair the Fed’s ability to operate. But it does tell you…
• High rates are expensive
• Time matters
• The Fed is incentivized to manage income carefully
That helps explain the shift toward bills and balance sheet stabilization.
So What’s the Actual Risk Signal?
Several things narrow the margin for error…
• Reserves are lower YoY and more exposed
• The ON RRP buffer is gone
• Discount window usage is rising modestly
• Liquidity management is becoming more deliberate, not looser
At the same time…
• The yield curve is positive
• Emergency facilities are quiet
• Asset composition remains conservative
This is late cycle balance sheet behavior, not crisis behavior.
Bottom Line
Recession risk implied by the balance sheet alone will be very dependent on how labor, Treasury issuance, and funding markets evolve from here.
Interactive guide to our weekly #Balanc[...]
Offshore
EndGame Macro H.4.1 Signals: Late Cycle Liquidity, Calm on the Surface, Tightening Underneath Total Fed assets now sit at $6.58 trillion, up $24.4 billion week over week, but still down about $305 billion from a year ago. QT hasn’t reversed outright, but…
eSheet report: https://t.co/75xiVY33QW #FedData - Federal Reserve tweet
Offshore
Photo
Offshore
Photo
The Few Bets That Matter
RT @0xDevShah: Meta acquired @ManusAI. Not a model company, they acquired an environment company, and the distinction is important.
I have a solid argument favoring that intelligence cannot exist in isolation. It cannot be dissociated from the context and environment in which it operationalizes itself. Manus has internalized this completely.
Manus runs on Claude with its custom tools built for orchestration and grounding. Their agentic environment enables the agents to browse, write code, manipulate files, and execute multi-step workflows without human in the loop.
They also beat OpenAI on GAIA. An interesting thing here is that they didn't build a foundation model. They built the most compatible environment for models to reason and act within.
I'm coining a new term here: Situated Agency. Situated Agency is an idea that agentic capabilities are not intrinsic to the model alone, but they emerge from the coupling of a model with tools, memory, and execution environment. Manus is perhaps the first company to productize Situated Agency at scale. And now Meta owns it.
Actually, this changes everything.
Meta spent a lot of time struggling to build SOTA models. Llama 4 was a disappointment. Behemoth was delayed because it couldn't compete with other frontier models. They built the Superintelligence team. Acquired Scale AI. All attempts were made to close the gaps.
And now the execution layer.
Manus has achieved SOTA agentic performance without training a single model. They engineered the environments and let Claude handle the inference-time compute. Meta might be positioning to become an agentic infrastructure company, not a foundation model company.
Meta has -
> Billions of users generating real-world task data and feedback loops daily
> Rayban glasses and Quest headsets as interfaces for agents
> WhatsApp, Messenger, Instagram as mediums for task delegation
> Zuckerberg also mentioned that he is pushing for personal superintelligence on all wearables
None of this requires Meta to have the SOTA model on MMLU. It requires Meta to have the best execution environment for models to act on behalf of users.
The Avocado rumours become interesting here.
Avocado is Meta's tbd closed model, reportedly being developed under @alexandr_wang. If Manus's agentic systems are genuinely model-agnostic, which their architecture suggests, then nothing blocks Meta from swapping Claude for Avocado.
Manus already runs Claude and fine-tuned Qwen interchangeably, routing different subtasks to different models based on capabilities. The architecture abstracts the model layer behind a smartly engineered tool-calling interface.
This gives Meta a production-tested agentic environment with $125M ARR that they can gradually integrate. They inherit the execution layer, the context engineering IP, the sandboxed compute infrastructure, the customer feedback loops, then port it to Avocado when the model is ready.
Things could get hot if Meta fully commits to this thesis.
OpenAI is building vertically. Foundation models, custom chips, agent frameworks, consumer applications. Google is building vertically. TPUs, Gemini, search, workspace integration. Both are betting that owning the foundation model layer is essential to capturing value.
Meta could be betting the opposite. If Situated Agency is correct, then the best strategy would be to build the best orchestration infrastructure. Let others race to improve the SOTA models, and swap in whatever model scores highest on your agent benchmarks at any given moment.
This is how Android beat iOS in market share. Google didn't build the best hardware. They built the best platform layer for hardware makers to build on, then captured the market. Meta making the same bet on agentic AI fits with Zuckerberg's playbook.
Manus may be the first sign that suggests Meta is thinking this way about AI agents.
Congrats to Meta and the complete teams at Manus AI! tweet
RT @0xDevShah: Meta acquired @ManusAI. Not a model company, they acquired an environment company, and the distinction is important.
I have a solid argument favoring that intelligence cannot exist in isolation. It cannot be dissociated from the context and environment in which it operationalizes itself. Manus has internalized this completely.
Manus runs on Claude with its custom tools built for orchestration and grounding. Their agentic environment enables the agents to browse, write code, manipulate files, and execute multi-step workflows without human in the loop.
They also beat OpenAI on GAIA. An interesting thing here is that they didn't build a foundation model. They built the most compatible environment for models to reason and act within.
I'm coining a new term here: Situated Agency. Situated Agency is an idea that agentic capabilities are not intrinsic to the model alone, but they emerge from the coupling of a model with tools, memory, and execution environment. Manus is perhaps the first company to productize Situated Agency at scale. And now Meta owns it.
Actually, this changes everything.
Meta spent a lot of time struggling to build SOTA models. Llama 4 was a disappointment. Behemoth was delayed because it couldn't compete with other frontier models. They built the Superintelligence team. Acquired Scale AI. All attempts were made to close the gaps.
And now the execution layer.
Manus has achieved SOTA agentic performance without training a single model. They engineered the environments and let Claude handle the inference-time compute. Meta might be positioning to become an agentic infrastructure company, not a foundation model company.
Meta has -
> Billions of users generating real-world task data and feedback loops daily
> Rayban glasses and Quest headsets as interfaces for agents
> WhatsApp, Messenger, Instagram as mediums for task delegation
> Zuckerberg also mentioned that he is pushing for personal superintelligence on all wearables
None of this requires Meta to have the SOTA model on MMLU. It requires Meta to have the best execution environment for models to act on behalf of users.
The Avocado rumours become interesting here.
Avocado is Meta's tbd closed model, reportedly being developed under @alexandr_wang. If Manus's agentic systems are genuinely model-agnostic, which their architecture suggests, then nothing blocks Meta from swapping Claude for Avocado.
Manus already runs Claude and fine-tuned Qwen interchangeably, routing different subtasks to different models based on capabilities. The architecture abstracts the model layer behind a smartly engineered tool-calling interface.
This gives Meta a production-tested agentic environment with $125M ARR that they can gradually integrate. They inherit the execution layer, the context engineering IP, the sandboxed compute infrastructure, the customer feedback loops, then port it to Avocado when the model is ready.
Things could get hot if Meta fully commits to this thesis.
OpenAI is building vertically. Foundation models, custom chips, agent frameworks, consumer applications. Google is building vertically. TPUs, Gemini, search, workspace integration. Both are betting that owning the foundation model layer is essential to capturing value.
Meta could be betting the opposite. If Situated Agency is correct, then the best strategy would be to build the best orchestration infrastructure. Let others race to improve the SOTA models, and swap in whatever model scores highest on your agent benchmarks at any given moment.
This is how Android beat iOS in market share. Google didn't build the best hardware. They built the best platform layer for hardware makers to build on, then captured the market. Meta making the same bet on agentic AI fits with Zuckerberg's playbook.
Manus may be the first sign that suggests Meta is thinking this way about AI agents.
Congrats to Meta and the complete teams at Manus AI! tweet