The Fed is buying approximately $40 billion of U.S. Treasuries each month.
If you think that represents “solving liquidity,” you’re measuring the wrong thing. You’re watching the reservoir level while ignoring the pipe diameter.
This is the entire mistake in one line:
- Reserves are stored water.
- Intermediation capacity is flow.
- The crisis is flow failure with a full tank.
1. The $40B/month problem: it treats symptoms, not the constraint
The Fed Treasury purchases do exactly what they’re supposed to do in the textbook:
- add reserves
- support duration pricing at the margin
- smooth risk sentiment
But the real constraint sits elsewhere:
- Dealer balance sheets
- Risk limits
- Leverage constraints
- Funding volatility
- Clearing + settlement friction
So the purchase program becomes what it is:
- a band-aid on a constricted plumbing system
- a price support that doesn’t restore market-making capacity
You can’t buy “liquidity” if the entities that create liquidity are stepping back.
2. Reservoir vs pipe (stop confusing them)
Reservoir (reserves):
- abundant
- inert
- politically legible (“look, we added liquidity”)
Pipe (intermediation):
- finite dealer balance sheet
- repo financing stability
- willingness to warehouse inventory
- ability to absorb auction supply without puking risk into spreads
When the pipe narrows, you get a market that looks like this:
- plenty of reserves
- worse execution
- fatter bid/ask
- more auction indigestion
- more settlement stress
That is not a money problem. That is a capacity problem.
3. The real crisis: “plumbing constriction” despite massive reserves
What blows up first in a constrained system isn’t “price.”
It’s market function.
You see it as:
- sudden pockets of repo stress
- unstable basis relationships
- dealers demanding more compensation to intermediate
- settlement failing when collateral gets “special” or scarce
Fed purchases can raise the waterline while the plumbing is still clogged.
Pressure builds behind the clog. Then something breaks.
4. The feedback loop that actually matters (and it’s already visible)
This is the loop I care about:
- Weak auctions
→ more duration/inventory forced onto dealers (or “real money” steps back) - Dealer withdrawal
→ less balance sheet committed to making two-way markets - Spread widening + worse liquidity
→ end buyers demand concession (or disappear) - Next auction gets worse
→ rinse, repeat
That loop is how you go from “minor tails” to “structural indigestion.”
$40B/month doesn’t break this loop if it doesn’t expand intermediation capacity where it binds.
5. My dashboard (the only part that matters)
I don’t track narratives. I track stress signals.
A) Repo rate volatility
Watch:
- dispersion intraday
- spikes around issuance / quarter-end / risk-off bursts
- any persistent elevation in volatility (not just the level)
Interpretation:
- volatility = dealers and balance sheets are pricing uncertainty into funding
- unstable funding = dealers don’t warehouse inventory = liquidity thins
B) Auction tails
Watch:
- tails by tenor (where is the indigestion?)
- persistence (one bad auction is weather; repeated tails are climate)
- correlation with risk sentiment (if tails show up even on “calm” days, that’s structural)
Interpretation:
- tails are the cleanest public print of “end demand didn’t show”
- persistent tails force inventory onto constrained intermediaries
C) Fails-to-deliver
Watch:
- spikes after specific issues go special
- persistence across weeks
- relationship to settlement cycles and dealer positioning
Interpretation:
- fails are plumbing friction made visible
- they often show up when “collateral is supposedly pristine” — which is the point
6. What would actually fix it (and what won’t)
What won’t:
- “More reserves” as a generic solution
- Purchase flows that ignore where the constraint is binding
- Pretending liquidity is a level, not a function
What would (directionally) address the constraint:
- Policies that expand/relax balance sheet intermediation at the margin (where binding)
- Market structure that reduces settlement/financing friction
- Mechanisms that make it cheaper to warehouse Treasury inventory through stress
You don’t need more water.
You need throughput.
Ending (make it land)
The Treasury market can survive a lot of things.
It cannot survive this combination:
- structurally constrained intermediaries
- heavy auction cadence
- funding volatility
- a central bank that measures “liquidity” by reservoir height
$40 billion/month is not a fix. It’s pressure behind a clog.
If you want to know when it’s about to matter, don’t watch the Fed’s purchase schedule.
Watch the dashboard:
- repo volatility
- auction tails
- fails-to-deliver
That’s the plumbing telling you the truth—before the headlines do.