## 1. The IS/LM model

## Source details

**Canonical URL:** [1. The IS/LM model](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp13186.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp13186.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp13186.pdf.json)

---

### Introduction and conceptual framing
- IS/LM conceptualizes the relationship between interest rates and real output: intersection of IS and LM curves yields simultaneous equilibrium in goods/services and money markets.
- Inward shifts in IS (e.g., contraction of C+I+G lowering output to YB) can be neutralized by shifting LM out (lowering nominal or real interest rates) to restore output to YA.
- Financial accelerator literature (Bernanke, Gertler, Gilchrist, 1996; BGG, 1998) documents that endogenous credit-market shocks (via flight to quality and higher cost for risky projects) shift IS inward and can amplify collateral-constraint shocks non-linearly and quantitatively significantly.
- Collateral underpins secured funding (mostly by nonbank investors), repurchase agreements (repo), and hedging (primarily OTC derivatives); it is essential to financial plumbing and intra-day debits/credits.

### Price of money, price of collateral, and market segmentation
- In jurisdictions with IOER offered only to depository institutions (example jurisdictions: U.S., U.K.), the price of money and short-term money-market rates are not fully market-determined, producing a wedge between banks and nonbanks.
  - Example figures:
    - Fed Funds rate on July 29, 2013 was 9 bps.
    - Deviation of 16 bps compared to likely first step in tightening rates; compare average daily absolute deviation of Fed Funds from FOMC target of 3 bps.
    - IOER cited as 25 basis points in footnote context.
- Collateral (repo) rate: rate at which cash is lent against collateral for an agreed tenor; collateral shortage lowers repo rates; collateral abundance increases repo rates.
- Regional observations:
  - Eurozone collateral/repo rates have dipped below zero for German/French/Dutch and also Danish/Swiss repo rates.
  - U.S. GC repo rates remain in positive territory; IOER has contributed to a wedge between U.S. and Eurozone repo rates.
  - Operation Twist provided extra T-bills in 2012, affecting GC rates.
- Technical drivers of cross-asset differences in "good collateral" prices include home bias, relative liquidity/depth/size of European collateral vs. U.S. T-bill market, differences in QE design (Fed vs. ECB), and cheapest-to-deliver collateral.

### Collateral, financial lubrication, and velocity (Box 1 and Box 2 findings)
- Financial collateral (including collateral re-use) was sizable and comparable with monetary aggregates like M2 or broad money.
- Before Lehman, financial lubrication (money plus collateral) was over US$30 trillion; about one-third came via pledged collateral.
- Subsequent decline in available collateral and associated re-use was an estimated US$4–5 trillion (difference between pre-crisis and post-crisis series).
- As of end-2012:
  - Overall financial lubrication rebounded to approximately US$30 trillion.
  - The "mix" shifted to money accounting for around 80 percent of financial lubrication.
  - Much of the increased money sits as excess reserves with central banks.
- Velocity of pledged collateral (volume of secured transactions divided by stock of source collateral):
  - End-2012: 2.2
  - End-2011: around 2.5
  - End-2007: 3.0
- As of end-2012: pledged collateral received by major banks that could be onward re-pledged in their own name: US$6 trillion.
  - Sources and amounts:
    - Non-hedge funds (pension, insurers, official sector, asset managers): US$1 trillion in securities on loan (without secondary market churning).
    - Hedge funds: AUM of US$2.4 trillion, provided US$1.8 trillion collateral to major banks via prime brokerage (PB) and non-PB activities.
      - PB-related collateral coming to banks: about US$700 billion.
      - Non-PB (largely repo) related collateral coming to banks: approx. $1.1 trillion.
- Example velocity calculation reported: 6 trillion/2.8 trillion = 2.15 (reported rounded/illustrative value near 2.2).
- Interpretation: decline in collateral velocity reduces financial lubrication; central bank asset purchases that withdraw good quality collateral reduce collateral reuse and velocity (silo-ed collateral has zero velocity).

### Empirical/analytical effects of collateral shifts within IS/LM
- Collapse in financial collateral (estimated US$4–5 trillion) shifts IS inward, lowering output to YB and decreasing real interest rates.
- QE shifts LM right to compensate; LM shifts may be sizable and ongoing, potentially resulting in real interest rates below zero (optically masked by nominal money-market distortions).
- Fed/QE quantitative effects and rate-path implications:
  - New York Fed President suggested Fed’s QE actions may have lowered nominal rates by additional 150–200 basis points.
  - Unadjusted real rates (if Fed balance sheet had remained at US$800 billion as of end-2007) may be much lower relative to adjusted real rates that account for expanded balance sheet via QE.
  - Historical tightening cycles averaged close to 400 bps; if the policy rate starts from minus 2 percent (LM shifted out to approx minus 2 percent to reach YA), a 400 basis points tightening cycle could anchor the new policy rate at 2 percent (unless balance sheet also returns to 2007 size).

### Sources of pledged collateral, velocity table (exact figures)
- Table (In trillions of U.S. dollars; velocity in units)
  - 2007: 1.7 / 1.7 / 3.4 / 10.0 / 3.0
  - 2010: 1.3 / 1.1 / 2.4 / 5.8 / 2.4
  - 2011: 1.3 / 1.05 / 2.35 / 6.1 / 2.5
  - 2012: 1.8 / 1.0 / 2.8 / 6.0 / 2.2

### Prime Brokerage and collateral accounting (illustrative calculations)
- PB related strategies (equity long/short) share: 33 percent.
- PB related collateral to the “street” calculation:
  - US$2.4 trillion x leverage of 1.5 x 33 percent share x delta bias of long/short of 0.6 ≈ about US$700 billion.
- Non-PB related collateral calculation:
  - US$2.4 trillion x 33 percent share x “adjusted leverage margin” ≈ US$1.1 trillion.
- One-third of strategies (managed futures, distressed or cash positions) do not require leverage; collateral does not come to the street from such strategies.
- Base figures preserved:
  - US$2.4 trillion base used in illustrative calculations
  - Leverage of 1.5; delta bias of long/short of 0.6; PB share 33 percent
  - Estimated PB-related collateral to the street: about US$700 billion
  - Estimated non-PB related collateral: US$1.1 trillion

### Central bank mechanics, unwind dynamics, and institutional constraints
- Definitions used:
  - D: Bank deposits at the central bank. (D excludes banknotes, vault cash)
  - C1: Good collateral in all states of nature; can be converted to D at no haircut
  - C2: Collateral that under normal market conditions is “good”; else loses value
  - C1 held by banks: Only banks can convert C1 to D overnight. Nonbanks cannot change C1 to D
  - Ultimate Liquidity: D plus C1 held by banks
- Central bank collateral behavior and policy implications:
  - If central banks substitute D for assets that retained value as collateral (C1), little liquidity relief is achieved.
  - Effective liquidity relief requires injecting central bank money and liquid collateral against illiquid or undesirable assets (C2) to increase supply of unencumbered collateral.
  - Central banks that are slow to loosen collateral policies risk a Gresham’s Law effect: bad collateral driving out good collateral from markets.
- QE effects and collateral possession vs ownership:
  - Major central banks have taken good collateral from markets and replaced it with freshly printed money (with ECB noted as taking in bad collateral).
  - Securities at central banks do not circulate (no velocity); excess reserves (D) do not substitute for C1 (good collateral), resulting in net reduction in financial lubrication.
- Reverse repos, nonbanks, rehypothecation, and collateral velocity:
  - Reverse repos with nonbanks (e.g., MMMFs) move liabilities and can shrink bank balance sheets, but collateral received by nonbanks via reverse repos cannot be rehypothecated or onward re-pledged.
  - Only banks can rehypothecate collateral received via reverse repo and thereby increase collateral velocity; collateral released to nonbanks will not convert D to C1.
  - The Fed’s SOMA lending facility is a Treasury-for-Treasury swap and does not transfer possession without ownership in the way needed to convert D to C1.
- Unwind outcomes and constraints:
  - Unwind—voluntary release of collateral or involuntary roll-off—moves LM curve left (absorbing money) and IS curve up (release of collateral), raising both money interest rate and repo rate.
  - Repo rates should be contained below policy rate to avoid inflationary dynamics.
  - Collateral release to the market is likely to be limited; central banks that withdrew collateral via QE become large holders of good collateral.
  - Basel III and tighter regulation reduce banking appetite for securities except for capital buffers, limiting bank balance sheet space for rehypothecation.
  - Nonbank balance sheet space is key to any unwind of collateral; collateral given to nonbanks will not restore rehypothecation-driven financial lubrication.
  - Excess reserves on bank balance sheets could be converted into reverse repos if reverse repos were perfect substitutes (they are not, due to D vs C1 distinction).
  - Fed reverse repo eligibility expansion may include nonbanks such as GSEs and selected MMMF; size of RRs and broader market supply/demand will determine repo rate relative to IOER.
  - Large reverse repos could push repo rates above IOER, risking inflationary dynamics or expectations thereof, so collateral release will be “measured.”

### Policy implications and recommendations
- Consider pledged collateral market conditions when setting monetary policy, given cross-border funding reliance and collateral's role in financial plumbing.
- A rebound in pledged collateral markets may be more effective in easing liquidity constraints than further QE, because pledged collateral typically has higher velocity than base money and is more widely usable across banks and nonbanks.
- When designing QE or large-scale asset purchases, account for collateral composition: buying "good collateral" (e.g., U.S. Treasuries) can drain high-velocity collateral from markets, reducing financial lubrication.
- Monitor and address market segmentation created by IOER or similar facilities (wedge between banks and nonbanks) that affect money-market functioning and repo market spreads.
- During unwind, aim to keep repo rates close to policy rate; large QE efforts make controlling the wedge between repo rates and policy rate harder.
- Consider alternative interventions (for example, securities lending facilities, targeted collateral provision) that preserve or restore collateral velocity rather than only expanding D on central bank balance sheets.
- Jurisdictional considerations:
  - When good collateral is paid interest on excess reserves (examples: 25 bps at the Fed or 50 bps in the U.K.), policy rate becomes “loaded” with inflationary ammunition; collateral release must be controlled so repo rate stays below policy rate.
  - RBA-style approaches that work via collateral transformation at a penalty rate may help cope with regulatory demand for high quality liquid assets without issuing more debt securities.

*Source: 1. The IS/LM model (IMF staff paper content from supplied PDF).*

### 1. The IS/LM model .....................................................................................................

### 1. The IS/LM model

### Introduction
- IS/LM conceptualizes the relationship between interest rates and real output: intersection of IS and LM curves yields simultaneous equilibrium in goods/services and money markets.
- Inward shifts in IS (e.g., contraction of C+I+G lowering output to YB) can be neutralized by shifting LM out (lowering nominal or real interest rates) to restore output to YA.
- The financial accelerator literature (Bernanke, Gertler, Gilchrist, 1996; BGG, 1998) documents that endogenous credit-market shocks (via flight to quality and higher cost for risky projects) shift IS inward and can amplify collateral-constraint shocks non-linearly and quantitatively significantly.
- Collateral underpins secured funding (mostly by nonbank investors), repurchase agreements (repo), and hedging (primarily OTC derivatives); it is essential to financial plumbing and intra-day debits/credits.

### Price of money and price of collateral
- In countries with IOER (interest on excess reserves) offered only to depository institutions (e.g., U.S., U.K.), the price of money and short-term money-market rates are not fully market-determined, producing a wedge between banks and nonbanks.
  - Example figures:
    - Fed Funds rate on July 29, 2013 was 9 bps.
    - Deviation of 16 bps compared to likely first step in tightening rates; compare average daily absolute deviation of Fed Funds from FOMC target of 3 bps.
    - IOER cited as 25 basis points in footnote context.
- Collateral (repo) rate: rate at which cash is lent against collateral for an agreed tenor; collateral shortage lowers repo rates; collateral abundance increases repo rates.
- Regional observations:
  - Eurozone collateral/repo rates have dipped below zero for German/French/Dutch and also Danish/Swiss repo rates.
  - U.S. GC (general collateral) repo rates remain in positive territory; IOER has contributed to a wedge between U.S. and Eurozone repo rates.
  - Operation Twist provided extra T-bills in 2012, affecting GC rates.
- Technical factors driving cross-asset differences in "good collateral" prices include home bias, relative liquidity/depth/size of European collateral vs. U.S. T-bill market, differences in QE design (Fed vs. ECB), and cheapest-to-deliver collateral.

### Box 1 — Pledged Collateral and Money Aggregates (key statistics and findings)
- Financial collateral (including collateral re-use) was sizable and comparable with monetary aggregates like M2 or broad money.
- Before Lehman, financial lubrication (money plus collateral) was over US$30 trillion; about one-third came via pledged collateral.
- Subsequent decline in available collateral and associated re-use was an estimated US$4–5 trillion (difference between pre-crisis and post-crisis series).
- As of end-2012:
  - Overall financial lubrication rebounded to approximately US$30 trillion.
  - The "mix" shifted to money accounting for around 80 percent of financial lubrication.
  - Much of the increased money sits as excess reserves with central banks.
- Fed’s balance sheet included over US$ 2 trillion in excess reserves of large banks (as of discussion), which does not contribute fully towards financial lubrication for nonbanks.
- Policy implication highlighted: a rebound in the pledged collateral market may be more effective in easing liquidity constraints than further QE; costs and trade-offs of QE become more apparent regarding central bank balance-sheet composition and collateral availability.

### Collateral and monetary policy via the IS/LM framework
- Declines in collateral use slow financial intermediation, producing effects similar to drying up of interbank markets.
- Pledged financial collateral need not be AAA/AA if securities are liquid, mark-to-market, and part of legal cross-border master agreements; such collateral functions as "cash equivalent".
- Demand for collateral is likely to increase over time due to:
  - Short-term counterparty risk concerns making secured funding more attractive.
  - Greater arm’s-length transactions in a globally integrated financial system seeking collateral security.
  - New regulations increasing demand for collateral-based operations.
- Empirical/analytical shifts:
  - Collapse in financial collateral (estimated US$4–5 trillion) shifts IS inward, lowering output to YB and decreasing real interest rates.
  - QE shifts LM right to compensate; LM shifts may be sizable and ongoing, potentially resulting in real interest rates below zero (optically masked by nominal money-market distortions).
- Fed/QE quantitative effects and rate-path implications:
  - New York Fed President suggested Fed’s QE actions may have lowered nominal rates by additional 150–200 basis points.
  - Unadjusted real rates (if Fed balance sheet had remained at US$800 billion as of end-2007) may be much lower relative to adjusted real rates that account for expanded balance sheet via QE.
  - Historical tightening cycles averaged close to 400 bps; if the policy rate starts from minus 2 percent (LM shifted out to approx minus 2 percent to reach YA), a 400 basis points tightening cycle could anchor the new policy rate at 2 percent (unless balance sheet also returns to 2007 size).

### Box 2 — Velocity of pledged collateral (2012) (key statistics and calculation)
- Velocity of pledged collateral (volume of secured transactions divided by stock of source collateral) was adversely impacted by central bank actions that mop up good collateral.
- Velocity estimates:
  - End-2012: 2.2
  - End-2011: around 2.5
  - End-2007: 3.0
- As of end-2012:
  - Pledged collateral received by major banks that could be onward re-pledged in their own name: US$6 trillion.
  - Sources and amounts:
    - Non-hedge funds (pension, insurers, official sector, asset managers): US$1 trillion in securities on loan (without secondary market churning).
    - Hedge funds: AUM of US$2.4 trillion, provided US$1.8 trillion collateral to major banks via prime brokerage (PB) and non-PB activities.
      - PB-related collateral coming to banks: about US$700 billion.
      - Non-PB (largely repo) related collateral coming to banks: approx. $1.1 trillion.
  - Velocity calculation example: 6 trillion/2.8 trillion = 2.15 (reported rounded/illustrative value near 2.2).
- Interpretation: decline in collateral velocity reduces financial lubrication; central bank asset purchases that withdraw good quality collateral reduce collateral reuse and velocity (silo-ed collateral has zero velocity).

### Policy implications and recommendations (drawn from analysis)
- Consider pledged collateral market conditions when setting monetary policy, given cross-border funding reliance and collateral's role in financial plumbing.
- A rebound in pledged collateral markets may be more effective in easing liquidity constraints than further QE, because pledged collateral typically has higher velocity than base money and is more widely usable across banks and nonbanks.
- When designing QE or large-scale asset purchases, account for collateral composition: buying "good collateral" (e.g., U.S. Treasuries) can drain high-velocity collateral from markets, reducing financial lubrication.
- Monitor and address market segmentation created by IOER or similar facilities (wedge between banks and nonbanks) that affect money-market functioning and repo market spreads.

*Source: 1. The IS/LM model (IMF staff paper content from supplied PDF).*

### 2.5 for end-2011 and largely neutralizes the bounce in “source” collateral (since the total sources of

### 2.5 for end-2011 and largely neutralizes the bounce in “source” collateral (since the total sources of

### Sources of Pledged Collateral, Velocity, and Collateral (2007 and 2010–12)
- Table (In trillions of U.S. dollars; velocity in units)
  - Year / Sources / Volume of Secured Operations / Velocity / Hedge funds / Others / Total
  - 2007: 1.7 / 1.7 / 3.4 / 10.0 / 3.0
  - 2010: 1.3 / 1.1 / 2.4 / 5.8 / 2.4
  - 2011: 1.3 / 1.05 / 2.35 / 6.1 / 2.5
  - 2012: 1.8 / 1.0 / 2.8 / 6.0 / 2.2
- Sources: Risk Management Association; and IMF staff estimates. See also Singh (2011 and 2012).

### Prime Brokerage (PB) related strategies and collateral accounting
- PB related strategies (equity long/short) share: 33 percent.
- PB related collateral to the “street” calculation (as presented):
  - US$2.4 trillion x leverage of 1.5 x 33 percent share x delta bias of long/short of 0.6 ≈ about US$700 billion.
- Non-PB related collateral calculation (as presented):
  - US$2.4 trillion x 33 percent share x “adjusted leverage margin” ≈ US$1.1 trillion.
- One-third of strategies (managed futures, distressed or cash positions) do not require leverage; collateral does not come to the street from such strategies.

### Definitions (Table 1: Terms Used)
- D: Bank deposits at the central bank. (D excludes banknotes, vault cash)
- C1: Good collateral in all states of nature; can be converted to D at no haircut
- C2: Collateral that under normal market conditions is “good”; else loses value
- C1 held by banks: Only banks can convert C1 to D overnight. Nonbanks cannot change C1 to D
- Ultimate Liquidity: D plus C1 held by banks
- Source: Singh and Stella (2012).

### Central banks, collateral, and market mechanics
- Collateral classification and crisis behavior:
  - C1 (e.g., U.S. Treasuries or Bunds) and D may remain acceptable at par during crises; everything else can cease to be liquid.
  - If central banks substitute D for assets that retained value as collateral (C1), little liquidity relief is achieved.
  - Effective liquidity relief requires injecting central bank money and liquid collateral against illiquid or undesirable assets (C2) to increase supply of unencumbered collateral.
  - Central banks that are slow to loosen collateral policies risk a Gresham’s Law effect: bad collateral driving out good collateral from markets.
- QE effects and collateral possession vs ownership:
  - Major central banks have taken good collateral from markets and replaced it with freshly printed money (with ECB noted as taking in bad collateral).
  - Unwind of central bank balance sheets will increase money interest rates and collateral (repo) rates.
  - Central banks will seek to keep policy rate and repo rate close to avoid a wedge; repo rates are an important gauge during unwind.
  - Larger QE efforts make controlling the wedge between repo rates and policy rate harder.
  - Securities at central banks do not circulate (no velocity); excess reserves (D) do not substitute for C1 (good collateral), resulting in net reduction in financial lubrication.
- Reverse repos, nonbanks, rehypothecation, and collateral velocity:
  - Reverse repos with nonbanks (e.g., MMMFs) move liabilities and can shrink bank balance sheets, but collateral received by nonbanks via reverse repos cannot be rehypothecated or onward re-pledged.
  - Only banks can rehypothecate collateral received via reverse repo and thereby increase collateral velocity; collateral released to nonbanks will not convert D to C1.
  - The Fed’s SOMA lending facility is a Treasury-for-Treasury swap and does not transfer possession without ownership in the way needed to convert D to C1.
  - TBAC has considered a securities lending facility (as noted in TBAC presentation).

### Unwind dynamics and constraints
- Unwind outcomes:
  - Unwind—voluntary release of collateral or involuntary roll-off—moves LM curve left (absorbing money) and IS curve up (release of collateral), raising both money interest rate and repo rate.
  - Repo rates should be contained below policy rate to avoid inflationary dynamics.
  - Collateral release to the market is likely to be limited; central banks that withdrew collateral via QE become large holders of good collateral.
- Institutional and regulatory constraints:
  - Basel III and tighter regulation reduce banking appetite for securities except for capital buffers, limiting bank balance sheet space for rehypothecation.
  - Nonbank balance sheet space is key to any unwind of collateral; even so, collateral given to nonbanks will not restore rehypothecation-driven financial lubrication.
  - Excess reserves on bank balance sheets could be converted into reverse repos if reverse repos were perfect substitutes (they are not, due to D vs C1 distinction).
  - Fed reverse repo eligibility expansion may include nonbanks such as GSEs and selected MMMF; size of RRs and broader market supply/demand will determine repo rate relative to IOER.
  - Large reverse repos could push repo rates above IOER, risking inflationary dynamics or expectations thereof, so collateral release will be “measured.”

### RBA proposals and policy implications (Conclusion)
- RBA approach:
  - The RBA’s proposals cope with upcoming regulatory changes requiring additional high quality liquid assets without issuing more debt securities.
  - The route resembles collateral transformation at a penalty rate, intended to keep collateral re-use from declining.
  - Demand collateral = Supply collateral * re-use factor (as presented).
  - RBA does not face the same QE dilemmas as Fed or BoE where release of good collateral is constrained by interaction between repo rate and policy rate.
- Jurisdictional differences:
  - When good collateral is paid interest on excess reserves (e.g., 25 bps at the Fed or 50 bps in the U.K.), policy rate becomes “loaded” with inflationary ammunition; collateral release must be controlled so repo rate stays below policy rate.
  - Only some jurisdictions may emulate RBA’s approach (examples listed in source: Singapore, Hong Kong, New Zealand, South Africa, etc.).
- Ownership vs possession distinction:
  - Central banks that took good collateral (ownership) out of the market removed collateral services that transfer with possession; markets need possession (and velocity) rather than ownership to restore collateral services.

### Key statistics and exact figures (preserved)
- Table figures:
  - 2007: 1.7; 1.7; 3.4; 10.0; 3.0
  - 2010: 1.3; 1.1; 2.4; 5.8; 2.4
  - 2011: 1.3; 1.05; 2.35; 6.1; 2.5
  - 2012: 1.8; 1.0; 2.8; 6.0; 2.2
- PB collateral computations and totals:
  - US$2.4 trillion base used in illustrative calculations
  - Leverage of 1.5; delta bias of long/short of 0.6; PB share 33 percent
  - Estimated PB-related collateral to the street: about US$700 billion
  - Estimated non-PB related collateral: US$1.1 trillion
- Policy-relevant rates and values mentioned:
  - Interest on excess reserves examples: 25 bps (Fed), 50 bps (U.K.)

*Source: Excerpt from IMF Working Paper content provided in the supplied PDF text.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp13186.pdf_