Concept Guide
Macro vs Micro Liquidity: Two Layers, One Dollar System
When practitioners say "dollar liquidity" they usually mean one of two related but distinct things: the macro pool created by central-bank and fiscal balance sheets, or the micro plumbing that decides whether that pool can actually flow to the borrowers and dealers who need it. Both layers matter; they fail in different ways; and an indicator built for one cannot answer the questions of the other.
Macro (broad) liquidity — the pool
Macro liquidity is the slow, policy-driven cash buffer in the system. It is set by the Federal Reserve's balance sheet (asset purchases and runoff), the Treasury General Account (how much cash the Treasury holds at the Fed rather than spending into the economy), the level of bank reserves, and the residual buffer at the ON RRP facility. It moves on weeks-to-quarters time scales and answers the question "how much usable dollar liquidity is available, in aggregate, right now?"
This is the layer the DLI scores. Its headline spine is the net-liquidity flow — the Fed balance sheet minus the Treasury General Account minus ON RRP — read as a smoothed 6-month-equivalent change, with an acute funding-stress override layered on top; credit and risk/price channels sit beside it as context panels. When the DLI reads tight, the macro pool is shallow or draining. When it reads loose, the pool is deep or refilling. The validation rubric (lag-0 vs NFCI/STLFSI/ANFCI/KCFSI) treats this as a coincident stance gauge, not a leading indicator.
Micro (market/funding) liquidity — the plumbing
Micro liquidity is the fast, structural ability of the financial system to move dollars across counterparties when someone wants them. It shows up in bid-ask spreads, dealer balance-sheet capacity, repo functioning, the cross-currency basis (FX swaps), and short-term commercial-paper funding premia. It can break even when macro liquidity is abundant — September 2019 happened against the backdrop of a large Fed balance sheet, and the cross-currency basis stays persistently negative in normal times because post-Basel III leverage rules make dealer intermediation expensive.
On this site the micro layer lives in the Offshore Reference section: Fed central-bank-liquidity-swap usage, the CP-Tbill spread (90-day AA financial CP minus 3-month T-bill), foreign holdings of U.S. Treasuries, and the published ECB/BoJ balance sheets. These are reference indicators — they are not scored in the DLI because their cadence (monthly/quarterly for some) and their signal character (zero most days, then spikes) do not fit the smooth net-liquidity headline. They answer "is the plumbing working?", not "how much is in the pool?".
The BIS bridge: macro provides the pool, micro decides whether it flows
The Bank for International Settlements consolidates the two layers into one operational definition: financial conditions are "the ease with which financing can be obtained in global financial markets." Macro is the necessary supply; micro is the sufficient distribution. Either alone is incomplete, and either alone can mislead.
A useful mental model: macro liquidity sets the ceiling on what is possible — you cannot have a credit boom if reserves are scarce — while micro liquidity sets the floor on what is reaching the real economy this week. The two layers are usually correlated (they share Fed policy as a common driver) but they diverge at exactly the moments that matter most: regime turns and crisis transmission.
Three case studies where the two layers tell different stories
Lehman 2008: Macro and micro tightened in lockstep. The Fed balance sheet had not yet expanded materially, and the cross-currency basis went sharply negative as foreign banks scrambled for dollars they could not borrow from each other. Both layers said "tight" simultaneously, which is why the policy response had to address both at once (QE for macro, central-bank swap lines for micro).
COVID March 2020: Macro tightened briefly (a flight to T-bills drained reserves into TGA) while micro broke catastrophically — the basis widened in days, commercial paper froze, dealer balance sheets refused to intermediate. The Fed's rescue was disproportionately a micro intervention: opening swap lines, the SRF, the CPFF, the PMCCF — buying time for the macro QE to take effect.
SVB March 2023: Macro was already easing (BTFP injected reserves quickly) but micro went tight at the same moment — CP-Tbill widened sharply as money funds dumped bank paper. A dashboard reading only macro flows would have missed the bank-run signal entirely; a dashboard reading only micro would have missed the policy response. Both matter, and they have to be read side by side.
How to use the two layers together on this site
Start with the DLI as the macro regime filter. Loose-and-improving means the policy pool is supportive; tight-and-deteriorating means the pool is shallow and policy is not actively refilling it. This is your baseline for whether risk assets have a tailwind or a headwind at the multi-week scale.
Then check the Offshore Reference section for micro stress. A non-zero Fed swap line reading is a binary "offshore funding is broken" flag. A widening CP-Tbill spread is a real-time bank-funding-pressure read. The BIS quarterly basis tells you the structural cost of synthesizing dollars abroad.
When the two layers agree, the regime is unambiguous. When they disagree — DLI loose but CP-Tbill widening, or DLI tight but offshore quiet — that is the more interesting state, and it usually marks either a transition or a market dislocation that has not yet been priced.