Liquidity, rates and prices tell different stories

Start with dates and units

The cached Fed balance sheet is $6,747.6B on Oct 7, 2026, +12.0 billion dollars compared with Jul 8, 2026. The 10-year minus 2-year Treasury spread is +0.42 percentage points on Oct 13, 2026. They describe different things: the size of the central bank’s balance sheet and the shape of the Treasury yield curve.

A small balance-sheet change can coexist with substantial changes in Treasury cash or short-term funding. Read the components before calling the whole environment easier or tighter.

The balance sheet is one component

The change above compares the latest cached observation with an observation at least 91 calendar days earlier. WALCL is weekly and reported in millions of dollars at its source; these charts convert it to billions. This is a descriptive comparison, not a stock-market forecast.

Treasury cash and reverse repo have their own dates

These series help unpack the movement of cash around the financial system. A convenient “net liquidity” shortcut leaves out other flows and uses components released at different times. Compare their observation dates before combining them.

A spread needs its underlying yields

The latest aligned yields are 5.21% for 10 years and 4.79% for 2 years. This chart subtracts same-date observations. A positive spread describes a higher long-term yield; it does not by itself establish the reason for the move or the next direction of equities.

Prices and employment provide another check

The CPI comparison uses the seasonally adjusted CPIAUCSL index in Sep 2025 and Sep 2026. It may differ from a headline rate based on the unadjusted series. A national price basket also differs from a particular household’s expenses.

Employment, prices and financial conditions belong in the same conversation. A chart becoming more favorable for one narrative is a reason to examine the other readings.

Source and interpretation

Source: the Federal Reserve balance sheet ↗, Treasury yields ↗, BLS consumer prices ↗ and BLS unemployment ↗, retrieved through FRED. Charts use our dated cache, not a real-time terminal. Each observation date is visible; a later source release may already exist. Historical observations can be revised.

The useful question for the next review: which component changed enough to alter the interpretation, and which other measure would contradict it?