The US Two-Tier Banking System: 2020 QE as Deficit Monetization and MMT in Practice

The two-tier banking analysis framework applies equally to the United States (upper layer: the Fed; lower layer: N commercial banks; bottom layer: households). Dollar internationalization causes gaps in the balance sheet, but in practice they are not severe. By decomposing 2020 data through balance-sheet change tables (end-of-period minus beginning-of-period): direct finance does not create M2; in 2020 M2 growth surged from its usual ~5% to 25%, driven by the Fed’s unlimited-QE “four-step liquidity injection” (buying Treasuries + MBS) — in essence a deficit monetization in which the fiscal authority “bridges” through banks to the central bank, i.e., MMT (Modern Monetary Theory) in practice. How this ends if continued: no historical precedent, no answer.

The Framework As It Stands

This section is organized from compiled research notes: the original framework’s structure, terminology, and key formulations are preserved, including editorial bridges and external factual annotations; charts are drawn by the compiler following the structure of the original text.

Core agenda and hidden threads

The framework’s main-line judgment: the two-tier banking analysis framework applies equally to the United States (upper layer: the Fed / lower layer: N commercial banks / bottom layer: households). Dollar internationalization introduces off-balance-sheet gaps, but “not severely so.” Apply the framework used to analyze China directly to the US, using a balance-sheet change table (end-of-period minus beginning-of-period) to decompose 2020 and see clearly “why M2 growth surged from its usual ~5% to 25%.”

Two hidden threads:

• Hidden Thread A — Direct finance does not create M2: Direct finance is essentially a transfer of existing M2 (money from investors moves to enterprises), not an increase in M2. Therefore the 25% figure in 2020 must have an extraordinary force creating money.
• Hidden Thread B — Deficit monetization / MMT is this cycle’s substance: The fiscal authority uses “bank bridging” to ultimately sell government bonds to the central bank = the central bank directly lending money to the fiscal authority. This is called MMT in practice; how it ends is unknown and without historical precedent.

Ten propositions:

1. The two-tier banking analysis framework applies equally to the US (upper layer: central bank / lower layer: N commercial banks / bottom layer: households); dollar internationalization causes off-balance-sheet gaps, but not severe ones.
2. Direct finance does not create M2 (it is essentially a transfer of existing M2); therefore the surge to 25% in 2020 (versus the usual ~5%, crisis-period ~10%) must have an extraordinary money-creation force behind it.
3. Decompose the 2020 Fed balance sheet using the change table: asset side Treasuries +$2.36trillion,MBS+$0.62 trillion ≈ total increment = unlimited QE.
4. Liability side: funds deposited by banks at the Fed (= base money) +$1.49 trillion; as the Fed buys Treasuries/MBS, government and bank deposits at the Fed increase correspondingly.
5. Commercial banking sector total assets +$2.84trillion,corporateloansonly+$0.35 trillion; mortgages securitized into MBS bought by the Fed → “deposits at the Fed” line items surge.
6. The four-step liquidity injection: ① banks lend, creating deposits; ② banks buy Treasuries (government deposits increase); ③ fiscal spending; ④ (most critical) the Fed’s unlimited QE buys approximately $3 trillion+ from banks.
7. Deficit monetization: countries legally prohibit the fiscal authority from borrowing directly from the central bank; the workaround = fiscal → bank → central bank “bridge,” functionally equivalent to the central bank lending money directly to the fiscal authority.
8. MMT: the 2020 operation = Modern Monetary Theory in practice (continually injecting money via fiscal deficits); still contested academically; how it ends if continued is unknown, without historical precedent.
9. The framework applies to all major mainstream economies using the two-tier banking system; each country differs only in the specific mechanics.
10. Core conceptual correction: banks do not “take deposited money and lend it out” (in reality loans create deposits); practitioners must sort out this foundational concept.

Reasoning chain (full-chain US 2020 QE):

flowchart TD
  B1["Two-Tier Banking Framework Applies to the US\n(Dollar Internationalization Gaps, but Not Severe)"] --> B2["Direct Finance Does Not Create M2 (Stock Transfer)\n→2020 M2 Surged to 25% (Normally ~5%) — Extraordinary Money Creation Required"]
  B2 --> B3["Balance-Sheet Change Table (End Minus Start) — Decomposing 2020"]
  B3 --> B4["Fed Asset Side: Treasuries +2.36T · MBS +0.62T = Unlimited QE\nLiability Side: Banks' Deposits at Fed (Base Money) +1.49T"]
  B3 --> B5["Commercial Banks: Total Assets +2.84T but Corporate Loans Only +0.35T\nMortgages Securitized into MBS Bought by the Fed"]
  B4 --> B6["Four-Step Liquidity Injection: Lending → Buy Treasuries → Fiscal Spending → Unlimited QE Buys ~3T+"]
  B5 --> B6
  B6 --> B7["Deficit Monetization: Fiscal → Bank → Fed Bridge = Fed Lending Directly to Fiscal"]
  B7 --> B8["= MMT in Practice; No Answer If Continued · No Historical Precedent\nFramework Applies to All Two-Tier Banking Economies"]
  B8 --> B9["Foundation: Loans Create Deposits, Not the Other Way Around"]

Key data anchors (2020 values cited in the course, not current data):

• US M2 growth rate comparison: normally ~5%, financial crisis ~10%, 2020 pandemic surge to 25%.
• 2020 Fed change table: asset side Treasuries +$2.36trillion\ast \ast ,MBS\ast \ast +$0.62 trillion (combined ≈ $3trillion\approx totalassetincrement);liabilitysidegovernmentdeposits+$1.26 trillion, bank deposits (base money) **+$1.49trillion\ast \ast ,currencyincirculation+$0.8 trillion.
• 2020 commercial bank change table: total assets +$2.84trillion\ast \ast ,corporateloansonly\ast \ast +$0.35 trillion, “deposits at the Fed” +$1.48 trillion (caused by mortgages securitized into MBS bought by the Fed).
• Four-step injection volumes: the fourth step — Fed unlimited QE bought Treasuries $2.36trillion,MBS$0.62 trillion, other securities $0.35trillionfrombanks,totalapproximately$3 trillion+.

⚠️ The figures above are empirical case values cited in the course (course estimate); for current-period or other-country judgments, one must obtain and label data from official sources such as the Fed’s H.4.1 balance-sheet release.

Application method (decomposing a country’s monetary injection using the two-tier banking framework, US 2020 QE template):

1. First confirm structural applicability: Does the target country have an upper-layer central bank + lower-layer N commercial banks + bottom-layer households structure? The off-balance-sheet gaps from dollar internationalization are not severe; the same framework applies.
2. Pull the change table: Subtract the beginning-of-period balance sheet from the end-of-period (central bank table + banking sector aggregate table), observe each line item’s “increment” over the year; assets and liabilities must still balance.
3. Read the central bank side: What did the asset side buy (Treasuries/MBS → QE scale)? Where did the liability side grow (banks’ deposits at the central bank = base money)?
4. Read the banking side: Total assets vs. corporate loan increment; did mortgages/Treasuries pass through securitization or get sold “pass-through” to the central bank (causing “deposits at the central bank” to surge)?
5. Chain the four steps: lending creates deposits → banks buy Treasuries → fiscal spending → the Fed’s unlimited QE buys away ~$3 trillion+.
6. Characterize the nature: Is this deficit monetization (fiscal → bank → central bank bridge = central bank lending directly to the fiscal authority)? Is this MMT in practice? Carry an honest caveat of “no historical precedent, no answer on how it ends.”
7. Correct the common misconception: Always ground the analysis in the “loans create deposits” foundation from The Two-Tier Banking System: T-Accounts and Loans Create Deposits; never say “banks take deposits and lend them out.”

Compiler’s Perspective

Coordinates: Category = Monetary System & Circulation / axis_h = Shu / axis_v = Its Place in the Whole

Bridging layer

A person relying solely on the aggregate narrative “the Fed printed money” will skip the most critical step: direct finance does not create M2. Equity issuance and corporate bonds are positional shifts of existing M2, not net additions. To explain why M2 growth, normally ~5%, could jump to 25% in 2020, one must decompose the balance-sheet change table to see the source of the “extraordinary money-creation force” — the answer lies in the fourth step of the four-step injection: the Fed’s unlimited QE bought $2.36trillioninTreasuries+$0.62 trillion in MBS from banks, approximately $3trillion+;thissumwasconvertedintobank{s}^{\prime }depositsattheFed(basemoney+$1.49 trillion), then amplified into broad money through the loan-creates-deposits mechanism.

The substance of deficit monetization is a “bridge” that circumvents legal prohibitions: national laws prohibit the fiscal authority from borrowing directly from the central bank, but the fiscal → bank → central bank path is functionally equivalent. Describing this path as “the printing press spinning faster” is a descriptive simplification; the actual mechanical differences determine the risk-release pathway: the bank layer is a real intermediate node, and mortgage securitization, the MBS market, and bank balance sheets are all components of the transmission chain — any abnormality in any link alters the landing point of monetary policy.

The framework’s exclusive claim: “no historical precedent, no answer” for MMT in practice is not a term of modesty — it is the most important piece of information. The 2020 figures (M2 25%, Treasuries +$2.36 trillion) tell us the scale; “no historical precedent” tells us there is no successful exit experience to draw on — these two pieces of information together constitute an honest judgment of the true risk of deficit monetization. Quoting either one alone is extracting half a conclusion.

Total Social Financing and Broad Money M2: Accounting Definitions and Divergence provides the accounting foundation for the M2 definition; The Zero-Rate QE Era: The Global Spread of the Monetary Experiment is the broader macro backdrop for this round of global diffusion; this article is the operational decomposition layer using the balance-sheet change table as empirical evidence.

See Also

• The Two-Tier Banking System: T-Accounts and Loans Create Deposits
• Total Social Financing and Broad Money M2: Accounting Definitions and Divergence
• The Zero-Rate QE Era: The Global Spread of the Monetary Experiment
• Central Bank Policy Tools and Monetary Policy Transmission

Sources

• Compiled research notes z-0193 · collected 2026-07
• External course (identity-stripped for collection), lectures 12-2 (primary) and 11-16 (mechanism foundation, see z-0192)