## wpiea2020046-print-pdf

## Source details

**Canonical URL:** [wpiea2020046-print-pdf](https://www.imf.org/-/media/files/publications/wp/2020/english/wpiea2020046-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2020/english/wpiea2020046-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2020/english/wpiea2020046-print-pdf.pdf.json)

---

### Overview and context
- Large scale asset purchase programmes (LSAPs) used since the global financial crisis to stimulate growth and fulfil inflation objectives in a zero lower bound environment.
- ECB measures (APP, negative deposit rate, TLTROs) reduced euro area long-term risk-free rates by around 80 basis points in the period June 2014 to December 2016.
- Public Sector Purchase Programme (PSPP) main APP component; accounted for approximately 80% of the entire APP until end 2016. PSPP start: March 2015.
- Introduction of PSPP followed by net capital outflows peaking in mid-2016 at nearly 5% of euro area GDP (portfolio investment net outflows).
- Net purchases of foreign securities by domestic investors since PSPP start were almost entirely in the form of long-term bonds.

### Research objective and key questions
- Objective: analyse capital outflows during the APP at security-by-security level using ESCB SHSS data for 2015Q1–2016Q4.
- Key questions:
  - Did euro area investors rebalance away from the same securities targeted by the ECB under PSPP?
  - Which securities (issuer countries and sectors) did euro area investors buy outside the euro area?
  - What determined these capital flows?
  - Were there heterogeneous investment patterns across euro area sectors and countries?

### Data and empirical framework (summary)
- Primary dataset: ESCB Securities Holding Statistics (SHSS), security-level holdings and transactions for 19 euro area countries; quarterly coverage since 2013Q4; analysis period 2015Q1 to 2016Q4.
- Sector aggregation (ESA 2010): MFI (excluding monetary authorities), ICPF, OFI (includes mutual funds), NFCs, general government, households.
- Asset characteristics merged via ISIN from CSDB: instrument type, issuer country/sector, currency, yields, original maturity, rating, outstanding amount.
- Main empirical specification: log-linear transaction equation with dependent variable ln(flow_a,i,j,t) = log of country-sector i’s net transactions (2015Q1-2016Q4) of security a issued by country-sector j.
- Controls include ln(A_a,i,j,t−1) (holdings at end-2014Q4), change in outstanding amount, euro-denomination dummy (EUR), ln(original maturity), average yield and rating in extensions.
- Fixed effects: holder country-sector (α_i), issuer country-sector (α_j), bilateral holder-issuer (γ_h,s). Standard errors clustered at holding country*sector level.
- Push-pull extensions: bilateral differences in news-based policy uncertainty, GDP forecast revisions, and yield differential.

### Main hypotheses and implementation details
- Preferred-habitat / search-for-yield:
  - PSPP eligible = 1 if security eligible for PSPP (issued by euro area governments or qualifying supranationals; maturity between 2 and 30 years; ratings above credit quality step 3; yield to maturity above deposit facility rate).
  - PSPP substitute = 1 for sovereign debt issued by specified advanced non-euro-area countries that fulfil PSPP requirements (listed: Australia, Canada, Czech Republic, Denmark, Hong Kong, Japan, New Zealand, Norway, Singapore, Sweden, Switzerland, United Kingdom, United States).
  - If mechanism holds: PSPP eligible coefficient expected negative; PSPP substitute coefficient expected positive.
- Contextual numbers:
  - Deposit facility rate at launch in January 2015 = -20bp.
  - By December 2016, stock of securities purchased under PSPP = 1.25 trillion EUR.
- Time-horizons tested:
  - Baseline: 8 quarters (2015Q1–2016Q4).
  - Additional: short term = 2 quarters; medium-term = 6 quarters; pre-PSPP window (2014Q3–2014Q4).
- Active vs passive decomposition:
  - Active: ln(flow_{a,i,j,t}) — net transactions.
  - Passive: [∆ln(stock_{a,i,j,t}) − ln(flow_{a,i,j,t})] — valuation changes.
  - Total change: ∆ln(stock_{a,i,j,t}).

### Descriptive evidence (aggregate, sectoral, geography, currency, maturity)
- Aggregate net transactions (2015Q1-2016Q4):
  - Euro area investors were net sellers of PSPP eligible securities.
  - More than EUR 250 billion PSPP eligible securities sold in net terms by euro area investors in 2015Q1-2016Q4.
  - Net sales of other euro area debt securities larger still; largest share issued by euro area banks.
  - Net purchases of foreign sovereign debt by euro area residents ≈ around EUR 350 billion; about 40% of these qualify as closest substitutes for PSPP eligible assets.
  - Net purchases of foreign private-sector debt slightly higher than foreign sovereign purchases.
- Sector drivers:
  - MFIs, households and OFIs accounted for largest net sales of PSPP eligible and other euro area debt securities.
  - ICPFs were net buyers of PSPP eligible and other euro area debt securities.
  - OFIs bought the largest amounts of PSPP substitutes and foreign debt securities overall, followed by MFIs and ICPFs.
- Geography:
  - Rebalancing towards non-euro area debt driven by Ireland and Luxembourg (financial centres hosting investment funds), and to a lesser extent Germany and France.
  - Spanish investors sold the largest amount of PSPP eligible securities, followed by the Netherlands, Germany, France, and Italy.
  - Italy was the largest net seller of other euro area debt securities, followed by Germany and France.
  - Heavy investment into US debt securities, then UK and “rest of the world”; largest net purchases for US government sector debt, then US NFCs.
- Currency exposures:
  - Sizeable net sales of euro-denominated debt — consistent with weakening euro-denomination bias.
  - Increased exposure to US dollar-denominated debt and British pounds.
  - All sectors were net buyers of US dollar-denominated debt; more than half of net purchases in US dollar were by OFIs.
  - ICPFs and, to a lesser extent, OFIs were net buyers of euro-denominated debt securities.
- Maturity composition:
  - Majority of net purchases in original maturity bucket > 10 years.
  - Vast majority of net sales were assets with maturity between 2 to 5 years after origination.
  - More than 50% of net purchases were securities with maturity exceeding 10 years.
  - Purchases of assets with minimum maturity of 10 years driven by ICPFs and OFIs; large net sales of 2–5 year securities mainly driven by MFIs and households.
- Investment funds and equities:
  - Euro area investors rebalanced towards euro area investment fund shares, debt securities issued outside the euro area, and to a lesser extent listed shares.
  - Largest net inflows by residents into ‘mixed’ investment funds, followed by bond-mandate funds.
  - Luxembourg-based OFIs accounted for 86% of all euro area OFI net purchases of extra-euro area government bonds.

### Key descriptive conclusions
- Strong support for search-for-yield and preferred-habitat hypotheses: rebalancing from euro area debt securities to foreign debt and investment fund shares, including sovereign debt of advanced non-euro-area countries.
- Evidence for duration-risk channel: > 50% of net purchases were securities with maturity > 10 years.
- Sector heterogeneity: ICPFs net buyers of PSPP eligible and other euro area debt; MFIs and households principal net sellers.
- Cross-border and currency effects: pronounced shift into US dollar-denominated debt, large net purchases of US government debt, and prominent role of OFIs (notably Luxembourg and Ireland) in channeling flows abroad.

### Econometric results — overall and robustness
- Baseline (2015Q1–2016Q4):
  - PSPP eligibility coefficient significant and negative: euro area investors rebalanced away from PSPP-targeted securities (Table 1 example coefficient PSPP Eligibility -0.808***).
  - PSPP substitute coefficient positive but insignificant in baseline regressions; descriptive evidence suggests indirect exposure via investment funds.
  - ln(original maturity) positive and significant (example coefficient 0.564***): investors bought relatively longer-term securities.
  - EUR denomination variable insignificant in baseline: consistent with increased rebalancing towards foreign-currency denominated debt.
  - Pre-PSPP holding variable negative (mean reversion): example ln(Holdpre) -0.652***.
  - Change in outstanding amount coefficient around 0.5 and significant (∆ Amount Out 0.505***): investors partly followed market portfolio developments.
- Robustness checks:
  - Excluding domestic securities yields similar results with slightly lower PSPP eligibility coefficient.
  - Excluding Ireland and Luxembourg increases magnitude of negative PSPP eligibility coefficient and maturity coefficient.
  - Restricting to long-term debt (original maturity > 12 months) yields similar results; long-term securities ≈ 95% of sample.
  - Risk-taking checks (high-yield or low-quality dummies) largely insignificant; sample sizes shrink substantially in these checks.
  - Replacing country-pair dummies with push-pull variables leaves main results intact; average difference in news-based policy uncertainty significant: investors buy less of debt issued by economies with higher policy uncertainty than investors’ country of residence.
  - Restricting to securities issued outside euro area: bilateral yield differential significantly positive — euro area investors bought more debt issued by non-euro-area countries with higher aggregate yields.

### Sectoral heterogeneity (summary of Table 2)
- OFIs:
  - Significant negative PSPP eligibility coefficient; invest in PSPP substitutes (consistent with preferred-habitat and descriptive evidence).
  - OFIs appear to channel ultimate investors’ preferences via investment funds.
- NFCs and households:
  - Significant net sellers of PSPP eligible assets and euro-denominated assets generally.
- MFIs:
  - PSPP eligibility coefficient insignificant; EUR coefficient significantly positive — banks’ incentives to hold PSPP eligible assets (zero risk weight, collateral use).
  - MFIs’ net sales of other euro area debt securities were three times larger than their sales of PSPP eligible debt (per descriptive Figures).
- ICPFs:
  - Strong preference for longer-term securities; large net buyers of long maturities consistent with liability matching.

### Geography and country heterogeneity
- Issuer-country groups: EA (euro area), ADV (other advanced), EME (emerging).
  - Broad-based purchases of debt securities issued in advanced economies outside the euro area; not limited to closest substitutes.
  - No evidence of broad-based euro area net purchases of EME debt securities; exception: significant net buying of EME sovereign bonds by euro area MFIs.
- Country-group heterogeneity (formerly stressed vs non-stressed):
  - Both groups net sellers of PSPP eligible securities; coefficient nearly three times larger for formerly stressed countries (Cyprus, Greece, Italy, Portugal and Spain), indicating stronger rebalancing away from PSPP-eligible assets.
  - Non-stressed country investors were significant net buyers of longer-term bonds.

### Time dynamics
- Pre-announcement (2014Q3–2014Q4):
  - Euro area investors were significant net buyers of PSPP eligible assets (anticipation/loading-up).
  - Example coefficient in Table 5 for pre-announcement PSPP eligibility 1.226***.
- Short-term (2015Q1–2015Q2):
  - Rebalancing starts immediately after launch: PSPP eligible coefficient negative (example -0.425*), EUR denomination negative and significant.
- Medium-term (6 quarters):
  - Strongest rebalancing forces; PSPP eligibility and euro-denomination negative coefficients larger in magnitude than short-term.
  - Original maturity coefficient grows over time (gradual switch to longer-term maturities).
- Sector time dynamics:
  - Loading up before programme driven by investment funds.
  - Short term: MFIs and households significant sellers.
  - Medium term: strongest rebalancing by households, OFIs and NFCs.

### Active versus passive rebalancing (summary of Section 4.6 and Table 7)
- Dependent variables:
  - Column (1): ln(flow_a,h,s) — cumulative net purchases (2015Q1–2016Q4).
  - Column (2): passive rebalancing = [∆ln(stock_a,h,s) − ln(flow_a,h,s)].
  - Column (3): ∆ln(stock_a,h,s) — change in holdings 2014Q4–2016Q4.
- Subsample: securities held by a sector at both 2014Q4 and 2016Q4 with reported holdings and transactions; observations example: 331356.
- Key findings:
  - Transactions: PSPP eligibility negative and significant for transactions (example Net transactions coefficient -1.541*** overall): euro area investors net sellers of PSPP-eligible securities.
  - Valuation (passive): PSPP eligibility positive and significant in passive rebalancing (example Passive rebalancing coefficient 1.492*** overall): substantial positive valuation gains on PSPP-eligible assets.
  - Holdings change: no significant PSPP-eligible coefficient in overall ∆ holdings (example ∆ Holdings -0.0485 not significant), highlighting that changes in holdings can mask transaction patterns.
  - Overall: active net sales only partly offset by positive valuation gains.
- Sector patterns:
  - OFIs: significant net sales and positive valuation gains; valuation gains offset part of net sales.
  - MFIs: significant net sales in subsample, small valuation gains.
  - Households: significant net sales and positive valuation gains; negative coefficient in overall holdings change.
  - ICPFs: net buyers of PSPP-eligible assets and achieved sizeable valuation gains; net purchases of securities with maturity > 10 years made up 50% of net debt purchases.

### Principal empirical magnitudes and exact illustrative coefficients (selected, preserved)
- PSPP Eligibility (baseline, Table 1 example): -0.808***.
- ln(Holdpre) (baseline example): -0.652***.
- ∆ Amount Out (baseline example): 0.505***.
- ln(original maturity) (baseline example): 0.564***.
- Baseline observations: 683007 (columns vary as reported).
- PSPP substitutes descriptive net purchases: net purchases of foreign sovereign debt ≈ around EUR 350 billion; about 40% qualify as closest substitutes.
- Net sales of PSPP eligible securities in 2015Q1–2016Q4: more than EUR 250 billion.
- PSPP stock by December 2016: 1.25 trillion EUR.
- Deposit facility rate at launch January 2015: -20bp.
- Active vs Passive (Table 7 overall part I examples): Net transactions -1.541***; Passive rebalancing 1.492***; ∆ Holdings -0.0485.

### Policy-relevant implications (derived from findings)
- Central-bank large-scale asset purchases in the euro area led to significant portfolio rebalancing by domestic investors away from targeted securities, generating cross-border capital outflows and global spillovers concentrated toward advanced-economy debt and investment fund channels.
- Investment-fund sector (OFIs) played a pivotal intermediary role in channeling investor demand abroad; supervisory and market-structure considerations relevant for assessing transmission and cross-border spillovers.
- Analysis of transactions (active component) is essential: valuation gains (passive component) can mask substantial net sales when only changes in holdings are observed.
- Sectoral and country heterogeneity matters for assessing distributional effects and financial stability implications (formerly stressed country investors and households exhibited stronger propensity to sell PSPP-eligible assets).
- Duration-risk channel and shift into long maturities (majority of net purchases in maturity > 10 years) underline the need to consider duration and liability structures when evaluating long-term implications of QE for institutional investors.

*Source: wpiea2020046-print-pdf — selected sections and empirical results from 1. INTRODUCTION, 2. EMPIRICAL FRAMEWORK, 2.3 Empirical implementation, 3 Descriptive evidence, 4.1 Overall results, 4.6 Active vs. passive rebalancing, and associated Tables and Figures.*

### 1. INTRODUCTION________________________________________________________5

### 1. INTRODUCTION

### Overview and context
- Large scale asset purchase programmes (LSAPs) are an unconventional monetary policy tool used since the global financial crisis to stimulate economic growth and fulfil inflation objectives in a zero lower bound environment.
- The European Central Bank’s (ECB) unconventional measures, covering the large scale Asset Purchase Programme (APP), a negative deposit rate, and targeted longer-term refinancing operations (TLTROs), reduced euro area long-term risk-free rates by around 80 basis points in the period June 2014 to December 2016.
- The Public Sector Purchase Programme (PSPP) is the main component of the APP and accounted for approximately 80% of the entire APP in the period until end 2016.
- Introduction of the PSPP (start: March 2015) was followed by significant net capital outflows; at their peak in mid-2016, net outflows of portfolio investment reached an all-time high of nearly 5% of euro area GDP.
- Net purchases of foreign securities by domestic investors since the start of the PSPP took place almost entirely in the form of long-term bonds.

### Research objective and questions
- Objective: analyse the capital outflows observed during the APP at a very granular, security-by-security level.
- Key questions:
  - Did euro area investors rebalance away from the same securities that were targeted by the ECB under the APP?
  - Which securities did euro area investors buy outside the euro area (issuer countries and sectors), thereby creating global spillovers?
  - What were the determinants of these capital flows?
  - Were there heterogeneous investment patterns across euro area sectors and countries?

### Contribution and empirical advantages
- Uses security-by-security data from the ESCB Securities Holding Statistics by sector (SHSS) to provide a detailed account of euro area portfolio rebalancing over the first eight quarters of the PSPP period: 2015Q1 to 2016Q4.
- First, to the authors’ knowledge, to analyse actual net transactions (net purchases/net sales) at the security level in a bilateral cross-country-sector setting.
- Ability to decompose portfolio rebalancing into:
  - "Active" component: capital flows (transactions).
  - "Passive" component: valuation changes due to fluctuations in security prices and exchange rates, following Tille and van Wincoop (2010).
- Allows examination of heterogeneity across investor sectors and countries due to:
  - Different initial positions (notably in exposure to public sector bonds).
  - Different degrees of investor sophistication, informational frictions, asset and liability management strategies, and regulatory constraints.

### Channels and mechanisms discussed
- Portfolio rebalancing channel: absorption of safe long-term bonds by the central bank raises prices and lowers yields of those assets, incentivising investors to seek higher-yielding assets abroad.
- Duration risk channel: reduced supply induces investors to accept a smaller term premium for long-term bonds, further lowering yields and encouraging investment in higher-yielding foreign assets.
- Preferred-habitat behaviour: rebalancing toward bonds with similar characteristics.
- Risk-taking channel: rebalancing toward riskier assets.
- Signalling channel: commitment to keep short-term rates low can create persistent interest rate differentials and incentivise cross-border carry trades.
- Inflation channel: expectations of higher future inflation and purchasing power parity considerations could trigger appetite for foreign assets.
- Confidence channel: market interpretation of QE about the central bank’s economic outlook can affect investor sentiment and risk aversion.

### Relation to existing literature
- Builds on literature documenting cross-border spillovers of QE (e.g., Neely (2010), Moore et al. (2013), Altavilla et al. (2015), Andrade et al. (2016)).
- Complements microdata studies of portfolio rebalancing focused on single sectors or countries (e.g., Calvet et al. (2009), Hau et al. (2017)) by using cross-country-sector granular transactions.
- Related SHSS-based studies:
  - Koijen et al. (2018): use changes in security holdings to assess rebalancing 2015Q2–2016Q4.
  - Albertazzi et al. (2018): find portfolio rebalancing with country differences.
  - Boermans and Vermeulen (2018): preferred-habitat preferences remained stable during the APP.
  - Bua and Dunne (2019): evidence of rebalancing away from euro area government bonds by Irish investment funds after the June 2016 PSPP scale-up.
- Incorporates push and pull factors literature on international capital flows (Calvo et al. (1993, 1996); Fernandez-Arias (1996); Chuhan et al. (1998); Griffin et al. (2004); Fratzscher (2012); Forbes and Warnock (2012); Cerutti et al. (2017); McQuade and Schmitz (2019)).

---

### 2. EMPIRICAL FRAMEWORK

### Data (Section 2.1)
- Primary dataset: ESCB Securities Holding Statistics (SHSS) covering security-level portfolio holdings and transactions of all 19 euro area countries.
- Data collection and scope:
  - Collected by national central banks from (i) financial investors and (ii) custodians.
  - Covers all short-term and long-term debt securities, listed shares, and investment fund shares identified by ISIN.
  - Includes directly and indirectly reported securities; only assets held in custody for non-financial investors are included to avoid double reporting.
- Time coverage and frequency:
  - Collected on a quarterly basis since 2013Q4.
  - Releases used in analysis: up to 2016Q4.
  - Analysis focuses on the PSPP period 2015Q1 to 2016Q4 (first eight quarters).
- Sector aggregation (following ESA 2010) to six sectors:
  - monetary and financial institutions (MFI) excluding monetary authorities,
  - insurance companies and pension funds (ICPF),
  - other financial institutions (OFI) (includes mutual funds),
  - non-financial corporations (NFCs),
  - general government,
  - households.
- Asset characteristics merged via ISIN from ESCB Centralised Securities Database (CSDB): instrument type, issuer country and institutional sector, currency of denomination, yields, original maturity, rating, outstanding amount.

### Econometric approach (Section 2.2)
- Empirical specification (log-linear transaction equation):
  - Dependent variable: ln(flow_a,i,j,t) — the log of country-sector i’s net transactions (net sales or purchases) accumulated over t (the PSPP period 2015Q1-2016Q4) of security a issued by country-sector j.
  - Pre-PSPP holdings control: ln(A_a,i,j,t−1) — holdings at the end of 2014Q4, included to control for the pre-existing level of a sector’s investment in a specific security.
    - If net transactions over 2015Q1-2016Q4 are negative, the logarithm of the absolute value is taken and multiplied by -1 (following Levy Yeyati et al. (2007)).
- Main variables of interest:
  - PSPP Eligible: dummy = 1 if the security is eligible to be bought under PSPP.
  - PSPP Substitute: dummy = 1 for sovereign debt securities issued by certain advanced countries outside the euro area (definition per 2.3.1 in source).
- Security-level controls:
  - Change in the outstanding amount of a security (market prices) to capture shifts in the market portfolio.
  - Currency denomination: binary for euro-denominated securities.
  - Original maturity (log days).
  - In additional estimations: average yield and rating of a security over the PSPP period.
- Identification and fixed effects:
  - Holder country-sector fixed effects (α_i) and issuer country-sector fixed effects (α_j).
  - Bilateral fixed effects (γ_h,s) between holder country h and issuer country s to capture bilateral factors affecting capital flows (absorb gravity variables, transaction costs, incomplete markets).
  - Standard errors clustered at the holding country*sector level.
- Push and pull factors (supplementary estimations replacing bilateral fixed effects γ_h,s):
  - Bilateral differences in:
    1. News-based policy uncertainty index (Baker et al. (2016)).
    2. GDP forecast revisions as a proxy for relative changes in macroeconomic expectations.
    3. Change in the yield differential at the country-level.
- Heterogeneity analysis:
  - Estimate varying coefficients (β^d_1,...,β^d_k) across sectors or country groups to assess heterogeneous responses simultaneously and allow direct statistical inference on differences.

---

### 3. PAPER ORGANISATION AND NEXT STEPS
- Section 2: dataset and empirical framework (introduced above).
- Section 3: descriptive evidence on portfolio rebalancing since the launch of the APP.
- Section 4: econometric results, including:
  - Overall results,
  - Sector heterogeneity,
  - Geography and country heterogeneity,
  - Time dynamics,
  - Active vs. passive rebalancing.
- Section 5: conclusion.

*Source: wpiea2020046-print-pdf - 1. INTRODUCTION________________________________________________________5*

### 2.3  Empirical implementation

### 2.3  Empirical implementation

### 2.3.1  Main hypotheses
- Preferred-habitat / search-for-yield prediction:
  - Euro area investors are expected to be net sellers of assets targeted by the Eurosystem under the PSPP and to rebalance into the closest substitutes.
  - Security-level exogenous variables:
    - PSPP eligible = 1 for securities eligible to be purchased by the Eurosystem under the PSPP (i.e. securities issued by euro area governments or securities of international or supranational institutions that fulfil requirements such as a maturity between 2 and 30 years, ratings above credit quality step 3 in the Eurosystem’s harmonised rating scale, and yield to maturity above the deposit facility rate).
    - PSPP substitute = 1 for securities issued by the public sector in non-euro area advanced economies which otherwise fulfil all the requirements of the PSPP.
  - If mechanism holds: significantly negative coefficient for PSPP eligible and positive coefficient for PSPP substitute.
  - Empirical context notes:
    - Deposit facility rate at launch in January 2015 = -20bp.
    - By December 2016, the stock of securities purchased under PSPP = 1.25 trillion EUR.
    - PSPP substitute issuers include: Australia, Canada, Czech Republic, Denmark, Hong Kong, Japan, New Zealand, Norway, Singapore, Sweden, Switzerland, United Kingdom, and the United States.
- Duration-risk channel:
  - Investors expected to rebalance into securities with longer maturities to gain duration risk as euro area yield curve shifted downwards and flattened but did not invert.
- Currency denomination / euro-bias channel:
  - Expect weakening of the euro-denomination bias in debt securities; PSPP impacts extended to euro-denominated private sector debt via signalling channel; corporate sector (CSPP) also affected NFC yields in final six months of observation.
- Risk-taking / foreign rebalancing channel:
  - Expect portfolio shifts towards other foreign debt securities, possibly including emerging economies, but more likely towards other advanced-economy debt if investors maintain certain risk profiles.
- Sector heterogeneity:
  - Retail investors (households) expected to be significant sellers of PSPP-eligible bonds; substitution often happens via investment funds (OFI).
  - Banks may be limited sellers due to low (zero) risk weight and collateral use.
  - Rebalancing into longer-term maturities expected to be driven by ICPFs (insurance companies and pension funds) due to long-term liabilities and risk regulation; banks retain preference for liquid assets.
- Country heterogeneity:
  - Formerly stressed countries (Cyprus, Greece, Italy, Portugal and Spain) may sell more eligible assets and take greater advantage of rebalancing opportunities.
  - Non-stressed countries expected to be more able to rebalance towards longer-term securities due to looser financial constraints.

### 2.3.2  Extensions of baseline approach
- Time-horizon variations:
  - Baseline: 8 quarters since start of PSPP.
  - Additional horizons: short term = 2 quarters; medium-term = 6 quarters.
  - Also test pre-PSPP period (2014Q3 and 2014Q4) to capture anticipation effects including ECB credit easing of June 2014 and Draghi (2014) speech at Jackson Hole.
- Active versus passive channels:
  - Replace dependent variable ln(flow_{a,i,j,t}) with:
    - ∆ln(stock_{a,i,j,t}) — change in stock of holding (active + passive).
    - [∆ln(stock_{a,i,j,t}) − ln(flow_{a,i,j,t})] — passive changes due to valuation (price) effects.
  - This distinguishes net transactions from valuation-driven changes.

### 3  Descriptive evidence on euro area rebalancing during the PSPP period

#### 3.1  Debt securities — aggregate and sectoral patterns
- Net sales and purchases (period 2015Q1-2016Q4):
  - Euro area investors were net sellers of PSPP eligible securities in 2015Q1-2016Q4.
  - More than EUR 250 billion PSPP eligible securities were sold in net terms by euro area investors in 2015Q1-2016Q4.
  - Even larger net sales recorded for other euro area debt securities, largest share issued by euro area banks (attributed to negative net issuance and PSPP signalling channel).
  - Net purchases of foreign sovereign debt by euro area residents ≈ around EUR 350 billion; around 40% of these qualify as closest substitutes for PSPP eligible assets.
  - Net purchases of foreign private-sector debt were slightly higher than foreign sovereign purchases.
- Sector drivers (summary):
  - MFIs, households and OFIs (mainly investment funds) accounted for the largest net sales of PSPP eligible and other euro area debt securities.
  - Insurance companies and pension funds (ICPFs) were net buyers of both PSPP eligible and other euro area debt securities.
  - OFIs bought the largest amounts of PSPP substitutes and foreign debt securities in general, followed by MFIs and ICPFs.
- Country residence patterns:
  - Rebalancing towards non-euro area debt driven by financial centres Ireland and Luxembourg (hosting many euro area investment funds), and to a lesser extent Germany and France.
  - Spanish investors sold the largest amount of PSPP eligible securities, followed by the Netherlands, Germany, France, and Italy.
  - Italy was the largest net seller of other euro area debt securities, followed by Germany and France.

#### 3.2  Beyond debt securities — investment funds and equities
- Aggregate portfolio shifts:
  - Euro area investors overall net sellers of euro area debt securities and rebalanced towards:
    - Euro area investment fund shares,
    - Debt securities issued outside the euro area,
    - To a lesser extent euro area and foreign listed shares.
- Sectoral flows into investment funds:
  - Net sales of euro area debt driven by MFIs and households.
  - ICPFs, households and OFIs bought the largest amounts of euro area investment fund shares.
  - OFIs were by far the largest net buyers of foreign debt securities, followed by MFIs and ICPFs.
  - Households channelled investments into overseas debt largely via investment funds.
- Investment fund mandates:
  - Largest net inflows by euro area residents went into ‘mixed’ investment funds, followed by bond-mandate funds.
  - Households and ICPFs were large net buyers of ‘mixed’ funds; OFIs focused more on bond funds.
  - ECB aggregate statistics indicate euro area investment funds mainly bought shares of other investment funds, debt securities and listed shares in the analysis period.
- Geography of investment fund purchases:
  - Largest net purchases of euro area investment fund shares originated from Germany, Italy, Spain and France.
  - Luxembourg-based OFIs accounted for 86% of all euro area OFI net purchases of extra-euro area government bonds.

#### 3.3  Geography, currency, and maturity composition
- Geography:
  - Euro area residents net sellers of debt securities issued in their home countries.
  - Heavy investment into US debt securities, followed by UK and “rest of the world”.
  - Largest net purchases recorded for US government sector debt, then US NFCs.
  - MFIs and households sold domestic securities; OFIs’ net purchases focused almost entirely on extra-euro area debt, largely US and UK.
  - Majority of net equity purchases were domestic and other euro area, reflecting investment fund shares.
- Currency exposures:
  - Sizeable net sales of euro-denominated debt securities — consistent with weakened euro-denomination bias.
  - Increased exposure to US dollar-denominated debt and British pounds.
  - All sectors were net buyers of US dollar-denominated debt; more than half of net purchases in US dollar were by OFIs.
  - ICPFs and, to a much smaller extent, OFIs were net buyers of euro-denominated debt securities.
- Maturity structure:
  - Majority of net purchases of debt securities were in original maturity bucket > 10 years.
  - Vast majority of net sales consisted of assets with maturity between 2 to 5 years after origination.
  - More than 50% of net purchases consisted of securities with maturity exceeding 10 years.
  - Purchases of assets with minimum maturity of 10 years driven by ICPFs and OFIs; ICPFs’ behavior consistent with matching long-term liabilities.
  - Large net sales of 2-5 year securities were mainly driven by MFIs and households.

### Key descriptive conclusions
- Strong support for hypotheses: euro area investors rebalanced from euro area debt securities to foreign debt and investment fund shares, consistent with “searching for yield” and moving into closest substitutes to PSPP-targeted securities (sovereign debt of advanced non-euro-area countries).
- Evidence for duration-risk channel: more than 50% of net purchases were securities with maturity > 10 years.
- Sector heterogeneity confirmed: ICPFs were net buyers of PSPP eligible and other euro area debt since PSPP launch; MFIs and households were principal net sellers.
- Cross-border and currency effects: strong shift into US dollar-denominated debt, large net purchases of US government debt, and important role of OFIs (notably Luxembourg and Ireland) in channeling flows abroad.

*Source: wpiea2020046-print-pdf — 2.3  Empirical implementation*

### 4.1  Overall results

### 4.1  Overall results

### Overall estimation results (PSPP period 2015Q1 to 2016Q4)
- PSPP eligibility variable is significant with a negative sign, confirming hypothesis number 1: euro area investors significantly rebalanced portfolios away from individual securities targeted under the PSPP, controlling for security-specific, country and sector factors.
- PSPP substitute variable is positive but insignificant in column (2); descriptive evidence suggests indirect exposure to PSPP substitutes via net purchases of investment fund shares.
- Original maturity coefficient is significantly positive, confirming hypothesis number 2: euro area investors were net buyers of relatively more long-term securities.
- Euro currency denomination variable is insignificant, in line with hypothesis 3, suggesting increased rebalancing towards foreign-currency denominated debt securities.
- Pre-PSPP holding variable has a negative coefficient, confirming “mean reversion” (investors sold assets they held relatively large amounts of in 2014Q4 and bought those held relatively small amounts).
- Change in outstanding amount over the PSPP period has a significantly positive coefficient of around 0.5, consistent with CAPM predictions that investors partly followed developments of the overall market portfolio.
- Robustness checks (columns 3–8):
  - Excluding domestic securities (column 3) yields similar results with a slightly lower PSPP eligibility coefficient, indicating sales particularly of securities issued by governments of investors’ country of residence.
  - Excluding Ireland and Luxembourg (column 4) produces a more negative PSPP eligibility coefficient and a larger maturity coefficient.
  - Restricting to long-term debt securities (original maturity > 12 months) (column 5) delivers very similar results; long-term securities comprise around 95% in the sample.
  - Risk-taking checks:
    - Dummy for securities with average yield within the highest decile over the PSPP period (column 6) is insignificant; remains insignificant if replaced with a dummy for the upper 25% of yields or with the yield itself (unreported).
    - Dummy for securities rated within the worst rating category (out of four standardised categories) (column 7) is insignificant; worst rating corresponds to BB+ and below for Standard & Poor’s and Fitch.
    - Sample size shrinks by more than 50% in columns (6) and (7) due to limited data availability.
  - Replacing country-pair dummies with push and pull variables (column 8) leaves results intact; the only significant variable is the average difference in news-based policy uncertainty (Baker et al., 2016): investors show significantly lower net purchases of debt securities issued by economies with higher economic policy uncertainty than investors’ country of residence.
  - When restricting the sample to securities issued by countries outside the euro area (unreported), the bilateral yield differential turns significantly positive: euro area investors bought larger amounts of debt securities issued by non-euro area countries with higher aggregate yields (push-pull framework).

### Sector heterogeneity (Table 2)
- Heterogeneous coefficients across sectors reflect differing regulatory restrictions, asset management strategies, and information sophistication.
- OFIs:
  - Significant negative coefficient on PSPP eligibility.
  - Invest in PSPP substitutes, consistent with preferred habitat motives and with descriptive evidence (Figures 7–9).
  - Evidence suggests channeling of ultimate investors’ preferences via investment funds; insignificant closest-substitute in Table 1 likely driven by investments into closest substitutes via mutual funds rather than direct foreign sovereign debt purchases.
- Non-financial corporations (NFCs) and households:
  - Significant net sellers of PSPP eligible assets and euro-denominated assets more generally.
- MFIs (banks):
  - PSPP eligible coefficient is insignificant; euro-denomination coefficient is significantly positive.
  - Reflects banks’ incentives to hold PSPP eligible assets due to zero risk weight and collateral role in monetary operations.
  - Despite Figure 5 showing MFIs as the sector with the largest net sales of PSPP securities, regression controls (e.g., pre-PSPP levels) explain the insignificant PSPP eligibility coefficient; MFIs’ net sales of other euro area debt securities were three times larger than those of PSPP eligible debt securities (Figure 5).
- ICPFs (insurance corporations and pension funds):
  - Strong preference for net purchases of longer-term securities (duration risk and preferred habitat channels); particularly large for ICPFs, followed by NFCs and households.
- Pre-PSPP holdings negative coefficient and positive sign on changes in outstanding amounts remain significant across all sectors.

### Geography (Table 3)
- Geography split: issuing sectors from three country groups — euro area (EA), other advanced (ADV), emerging (EME) economies.
- PSPP eligibility remains significant overall (column 1); net investment by euro area OFIs into closest substitutes remains significant.
- Securities issued by euro area residents:
  - Significant net purchases of debt securities issued by OFIs (driven by household and NFC sectors); these largely are bonds issued by financial intermediaries linked to euro area NFCs (or MFIs) and securitisation entities.
- Euro area sectors invested significantly in bonds issued by banks and OFIs resident in advanced countries outside the euro area.
- Euro area ICPFs and households also invested significantly in bonds issued by advanced economies’ NFCs.
- Evidence indicates broad-based purchases of debt securities issued in advanced economies, not only closest substitutes — consistent with risk-taking channel evidence (Ammer et al., 2018) for non-US investors rebalancing into riskier US corporate bonds.
- Emerging market exposures:
  - No evidence of broad-based euro area net purchases of EME debt securities.
  - Significant net buying of EME sovereign bonds by euro area MFIs.
- Overall: shift towards foreign debt was broad-based across sectors but mainly targeted at other advanced economies.

### Country heterogeneity (Table 4)
- Heterogeneous coefficients across formerly stressed and non-stressed euro area country groups (formerly stressed: Cyprus, Greece, Italy, Portugal and Spain; Ireland and Luxembourg excluded).
- Both country groups were significant net sellers of PSPP eligible securities.
  - Coefficient almost three times as large for formerly stressed countries, indicating larger propensity to rebalance away from PSPP-eligible assets.
  - Suggests investors in formerly stressed countries realized capital gains on PSPP eligible securities; capital gains were particularly sizable for banks in formerly stressed countries due to prior purchases of domestic public debt during the sovereign debt crisis (moral suasion and carry trade motives).
- Only investors of non-stressed countries were significant net buyers of longer-term maturity bonds, likely due to lower yields in these countries and looser financial constraints enabling longer investment horizons.

### Time dynamics (Table 5 and Table 6)
- Pre-announcement period (2014Q3-2014Q4) (column 1):
  - Euro area investors were significant net buyers of PSPP eligible assets in the two quarters before the programme announcement, suggesting anticipation effects and speculative loading-up in advance of PSPP.
- Short-term (first two quarters of PSPP, 2015Q1-2015Q2) (column 2):
  - Significantly negative coefficient on PSPP eligible assets and euro-denominated debt securities; rebalancing starts immediately after launch.
  - Strong positive coefficient on changes in outstanding amount — investors generally follow the market portfolio — and a less pronounced mean reversion.
- Medium-term (6 quarters) (column 3):
  - Negative coefficients on PSPP eligibility and euro-denomination larger than in short term; rebalancing forces strongest in this period.
  - Net sales of euro-denominated assets significant only in short and medium term, but not over the entire period (column 4); may reflect strong announcement effects and sectoral needs to hold euro-denominated securities.
  - Original maturity coefficient exhibits a larger positive sign over time, suggesting gradual switch to longer-term maturities amid extended low yield environment.
- Sectoral time dynamics (Table 6):
  - “Loading up” on eligible assets before programme start is driven by investment funds, indicating speculation on valuation gains.
  - Short term: MFIs and households exhibit significant negative PSPP eligibility coefficients.
  - Medium term (6 quarters): strongest rebalancing driven by households, OFIs and NFCs; consistent with baseline period results.

*Source: https://www.imf.org/-/media/files/publications/wp/2020/english/wpiea2020046-print-pdf.pdf*

### 4.6  Active vs. passive rebalancing

### 4.6  Active vs. passive rebalancing

### Methodology and variables
- Analysis motivated by the theoretical model of Tille and van Wincoop (2010).
- Dependent variables in Table 7 (column references as in source):
  - Column (1): cumulative net purchases of debt securities over the period 2015Q1-2016Q4 ln(flow_a,h,s).
  - Column (2): valuation changes defined as [∆ln(stock_a,h,s) − ln(flow_a,h,s)].
  - Column (3): change in holdings between 2014q4 and 2016q4 ∆ln(stock_a,h,s).
- Subsample restriction: only securities held by a given sector both at the end of 2014Q4 and at the end of 2016Q4 and for which holdings and transactions were reported (necessary to compute passive rebalancing).

### Key empirical findings
- Transactions (active rebalancing):
  - Negative coefficient on PSPP eligibility for transactions (net purchases), consistent with baseline estimation — i.e., euro area investors were net sellers of PSPP-eligible securities.
  - Significant net sales of PSPP-eligible securities recorded for MFIs in the subsample (significant at the 10% level).
- Valuation (passive rebalancing):
  - Positive coefficient in the passive rebalancing estimation implying euro area investors recorded significant positive valuation gains in PSPP-eligible assets relative to other debt securities during the period.
  - Valuation gains partly offset active net sales but did not fully compensate them.
- Holdings:
  - No significant coefficient for PSPP-eligible assets in the overall change in holdings estimation (column 3) for the full period of analysis, underscoring the importance of analysing actual transactions rather than proxying with changes in holdings.

### Sectoral patterns
- Aggregate patterns driven by OFIs, MFIs and households, with varying magnitudes:
  - OFIs: significant net sales and positive valuation gains in PSPP-eligible securities; generated sizeable valuation gains while selling PSPP-eligible assets, thereby offsetting part of net sales.
  - MFIs: significant net sales of PSPP-eligible securities (in subsample) and recorded only small valuation gains during the PSPP period; MFIs were significant net buyers of euro denominated securities overall (reflecting regulatory and balance sheet management reasons).
  - Households: significant net sales and positive valuation gains in PSPP-eligible securities; significant (and negative) coefficient in the overall change in holdings estimation is obtained for households.
  - ICPFs: continued to be net buyers of PSPP-eligible assets and achieved sizeable valuation gains; net purchases of securities with a maturity exceeding 10 years made up 50% of net debt purchases.

### Descriptive volumes and dynamics
- Descriptive evidence (Figure 16) highlights:
  - ICPFs: sizeable valuation gains in PSPP-eligible assets and continued net buying of these assets.
  - MFIs and households: large net sales of PSPP-eligible assets and only small valuation gains during the PSPP period.
  - OFIs: more sizeable valuation gains while selling PSPP-eligible assets, offsetting part of net sales.
- The active (net purchases) and passive (valuation changes) channels worked in opposite directions during the PSPP period; overall net sales by euro area investors were only partly offset by positive valuation gains.

### Implications emphasized in concluding section
- Importance of analysing actual transactions rather than proxying with changes in holdings due to the distinct active vs. passive components.
- Active net sales of PSPP-eligible securities by euro area investors were only partly offset by positive capital gains, implying incomplete compensation from valuation effects for transactional outflows.

*Source: wpiea2020046-print-pdf - 4.6  Active vs. passive rebalancing*

### References

### wpiea2020046-print-pdf - References

### References list
- Full bibliography of cited works (authors, years, titles, working papers, journal articles, speeches, and technical reports) including but not limited to:
  - Ahmed, S., S. E. Curcuru, F. E. Warnock, and A. Zlate(2016): “Decomposing International Portfolio Flows,” Tech. rep.
  - Albertazzi, U., B. Becker, and M. Boucinha(2018): “Portfolio Rebalancing and the Transmission of Large-Scale Asset Programs: Evidence from the Euro Area,” ECB Working Paper 2125.
  - Altavilla, C., F. Canova, and M. Ciccarelli(2016): “Mending the broken link: heterogeneous bank lending and monetary policy pass-through,” ECB Working Papers 1978.
  - Altavilla, C., G. Carboni, and R. Motto(2015): “Asset purchase programmes and financial markets: lessons from the euro area,” ECB Working Paper Series 1864.
  - Altavilla, C., S. Simonelli, and M. Pagano(2017): “Bank Exposures and Sovereign Stress Transmission,” Review of Finance, 21, 2103–2139.
  - Ammer, J., S. Claessens, A. M. Tabova, and C. Wroblewski(2018): “Searching for Yield Abroad : Risk-Taking Through Foreign Investment in U.S. Bonds,” International Finance Discussion Papers 1224, Board of Governors of the Federal Reserve System.
  - Andrade, P., J. Breckenfelder, F. D. Fiore, P. Karadi, and O. Tristani(2016): “The ECB’s asset purchase programme: an early assessment,” ECB Working Paper 1956.
  - Baker, S. R., N. Bloom, and S. J. Davis(2016): “Measuring Economic Policy Uncertainty,” The Quarterly Journal of Economics, 131, 1593–1636.
  - Bergant, K. and M. Schmitz(2019): “International financial flows and the Eurosystem’s asset purchase programme: evidence from b.o.p and security by security data,” in Are post-crisis statistical initiatives completed?, ed. by B. for International Settlements, Bank for International Settlements, vol. 49 of IFC Bulletins chapters.
  - Boermans, M. and R. Vermeulen(2016): “International investment positions revisited: Investor heterogeneity and individual security characteristics,” DNB Working Papers 531, Netherlands Central Bank, Research Department.
  - Boermans, M. and R. Vermeulen(2018): “Quantitative easing and preferred habitat investors in the euro area bond market,” DNB Working Papers 586, Netherlands Central Bank, Research Department.
  - Brunnermeier, M. K. and Y. Sannikov(2016): “The I Theory of Money,” Working Paper 22533, National Bureau of Economic Research.
  - Bua, G. and P. G. Dunne(2019): “The Portfolio Rebalancing Effects of the ECB’s Asset Purchase Programme,” International Journal of Central Banking, 15, 1–46.
  - Bubeck, J., M. M. Habib, and S. Manganelli(2017): “The Portfolio of Euro Area Fund Investors and ECB Monetary Policy Announcements,” ECB Working Paper 2116.
  - Calvet, L. E., J. Y. Campbell, and P. Sodini(2009): “Fight or Flight? Portfolio Rebalancing by Individual Investors,” Tech. Rep. 1.
  - Calvo, G. A., L. Leiderman, and C. M. Reinhart(1993): “Capital Inflows and Real Exchange Rate Appreciation in Latin America: The Role of External Factors,” IMF Staff Papers, 40, 108–151.
  - Calvo, G. A., L. Leiderman, and C. M. Reinhart(1996): “Inflows of Capital to Developing Countries in the 1990s,” Journal of Economic Perspectives, 10, 123–139.
  - Camanho, N., H. Hau, and H. Rey(2018): “Global Portfolio Rebalancing and Exchange Rates,” NBER Working Papers 24320.
  - Cerutti, E. M., S. Claessens, and A. K. Rose(2017): “How Important is the Global Financial Cycle? Evidence from Capital Flows,” IMF Working Papers 17/193, International Monetary Fund.
  - Chari, A., K. D. Stedman, and C. Lundblad(2017): “Taper Tantrums: QE, its Aftermath and Emerging Market Capital Flows,” NBER Working Papers 23474.
  - Chuhan, P., S. Claessens, and N. Mamingi(1998): “Equity and bond flows to Latin America and Asia: the role of global and country factors,” Journal of Development Economics, 55, 439–463.
  - Coeure, B.(2017): “The international dimension of the ECB’s asset purchase programme,” Speech by Benoit Coeure at the Foreign Exchange Contact Group meeting, 11 July 2017.
  - De Santis, R. A.(2016): “Impact of the asset purchase programme on euro area government bond yields using market news,” ECB Working Paper 1939.
  - Domanski, D., H. S. Shin, and V. Sushko(2017): “The Hunt for Duration: Not Waving but Drowning?” IMF Economic Review, 65, 113–153.
  - Draghi, M.(2014): “Unemployment in the euro area,” Speech by Mario Draghi, President of the ECB, Annual central bank symposium in Jackson Hole, 22 August 2014.
  - ECB publications and working papers (2017a, 2017b, 2017c) on net portfolio investment outflows, impact of non-standard measures, and securities holdings.
  - Eisenschmidt, J., D. Kedan, M. Schmitz, R. Adalid, and P. Papsdorf(2017): “The Eurosystems asset purchase programme and TARGET balances,” ECB Occasional Paper 196.
  - Falagiarda, M., P. McQuade, and M. Tirpak(2015): “Spillovers from the ECB’s non-standard monetary policies on non-euro area EU countries: evidence from an event-study analysis,” Working Paper 1869, European Central Bank.
  - Fernandez-Arias, E.(1996): “The new wave of private capital inflows: Push or pull?” Journal of Development Economics, 48, 389–418.
  - Forbes, K. J. and F. E. Warnock(2012): “Capital flow waves: Surges, stops, flight, and retrenchment,” Journal of International Economics, 88, 235–251.
  - Fratzscher, M.(2012): “Capital flows, push versus pull factors and the global financial crisis,” Journal of International Economics, 88, 341–356.
  - Galstyan, V. and P. R. Lane(2013): “Bilateral portfolio dynamics during the global financial crisis,” European Economic Review, 57, 63–74.
  - Georgiadis, G. and J. Graeb(2016): “Global financial market impact of the announcement of the ECB’s asset purchase programme,” Journal of Financial Stability, 26, 257–265.
  - Griffin, J. M., F. Nardari, and R. M. Stulz(2004): “Are Daily Cross-Border Equity Flows Pushed or Pulled?” The Review of Economics and Statistics, 86, 641–657.
  - Hau, H., N. Camanho, and H. Rey(2017): “Global Portfolio Rebalancing under the Microscope,” Sfi working paper.
  - Heipertz, J., R. Ranciere, and N. Valla(2016): “Domestic and international sectoral portfolios: Network structure and balance sheet effects,” mimeo.
  - Koijen, R. S. J., F. Koulischer, B. Nguyen, and M. Yogo(2018): “Inspecting the Mechanism of Quantitative Easing in the Euro Area,” Banque de France Working Paper.
  - Krishnamurthy, A. and A. Vissing-Jorgensen(2011): “The Effects of Quantitative Easing on Interest Rates: Channels and Implications for Policy,” Brookings Papers on Economic Activity.
  - Lane, P. R.(2015): “Risk Exposures in International and Sectoral Balance Sheet Data,” World Economics, 16, 55–76.
  - Lemke, W. and T. Werner(2017): “Dissecting long-term Bund yields in the run-up to the ECB’s Public Sector Purchase Programme,” ECB Working Papers 2106.
  - Levy Yeyati, E., U. Panizza, and E. Stein(2007): “The cyclical nature of North-South FDI flows,” Journal of International Money and Finance, 26, 104–130.
  - Martin, P. and H. Rey(2004): “Financial super-markets: size matters for asset trade,” Journal of International Economics, 64, 335–361.
  - McQuade, P. and M. Schmitz(2019): “America First? A US-centric view of global capital flows,” Working Paper Series 2238, European Central Bank.
  - Moore, J., S. Nam, M. Suh, and A. Tepper(2013): “Estimating the impacts of U.S. LSAPs on emerging market economies local currency bond markets,” Tech. rep.
  - Neely, C. J.(2010): “The large scale asset purchases had large international effects,” Tech. rep.
  - Okawa, Y. and E. van Wincoop(2012): “Gravity in International Finance,” Journal of International Economics, 87, 205–215.
  - Tille, C. and E. van Wincoop(2010): “International capital flows,” Journal of International Economics, 80, 157–175.
  - Timmer, Y.(2018): “Cyclical Investment Behavior across Financial Institutions,” Journal of Financial Economics.
  - Vayanos, D. and J.-L. Vila(2009): “A Preferred-Habitat Model of the Term Structure of Interest Rates,” NBER Working Papers 15487.

### Figures included (Figures 1–16) — captions and key notes preserved
- Figure 1: Breakdown of euro area net portfolio investment flows
  - Last observation is December 2017. 12-month cumulative sums in bn EUR.
  - Notes: A positive (negative) number indicates net outflows (inflows) from (into) the euro area. Equity includes investment fund shares.
- Figure 2: Breakdown of euro area portfolio investment outflows
  - Last observation is December 2017. 12-month cumulative sums in bn EUR.
  - Notes: A positive (negative) number indicates net purchases (sales) of non-euro area securities by euro area investors. Equity includes investment fund shares.
- Figure 3: Average Yield to Maturity
  - Unweighted average yield to maturity of securities eligible to be bought under PSPP (PSPP eligible) vs. comparable securities outside the euro area (PSPP substitutes). PSPP eligible and PSPP substitute as defined in 2.3.1.
- Figure 4: Euro area net debt transactions
  - Cumulated net purchases by euro area residents from 2015Q1-2016Q4 in EUR bn. PSPP eligible and PSPP substitute as defined in 2.3.1.
- Figure 5: Euro area net debt transactions by sector
  - Cumulated net purchases by euro area residents from 2015Q1-2016Q4 in EUR bn. PSPP eligible and PSPP substitute as defined in 2.3.1.
- Figure 6: Euro area net debt transactions by country
  - Cumulated net purchases by euro area residents from 2015Q1-2016Q4 in EUR bn. PSPP eligible and PSPP substitute as defined in 2.3.1.
- Figure 7: Euro area net transactions: intra-(LHS) and extra-euro area (RHS)
  - Cumulated net purchases by euro area residents from 2015Q1-2016Q4 in EUR bn. Securities on the left (right) issued in (outside) the euro area.
- Figure 8: Euro area net transactions by sector: intra-(LHS) and extra-euro area (RHS)
  - Cumulated net purchases by euro area residents from 2015Q1-2016Q4 in EUR bn. Securities on the left (right) issued in (outside) the euro area.
- Figure 9: Euro area net transactions by country: intra-(LHS) and extra-euro area (RHS)
  - Cumulated net purchases by euro area residents from 2015Q1-2016Q4 in EUR bn. Securities on the left (right) issued in (outside) the euro area.
- Figure 10: Euro area net transactions of investment fund shares split by their main mandate
  - Cumulated net purchases of investment fund shares from 2015Q1-2016Q4 split by their respective main mandate of investment in bn EUR. Data on the mandates are from an extract of the CSDB on 31/01/2018.
- Figure 11: Euro area net transactions by geography: debt (LHS) and equity (RHS)
  - Cumulated net purchases by euro area residents from 2015Q1-2016Q4 in EUR bn. Debt securities on the left and equity securities on the right hand side.
- Figure 12: Euro area net transactions by currency: debt (LHS) and equity (RHS)
  - Cumulated net purchases by euro area residents from 2015Q1-2016Q4 in EUR bn. Debt securities on the left and equity securities on the right hand side.
- Figure 13: Euro area net transactions by currency: debt (LHS) and equity (RHS)
  - Cumulated net purchases by euro area residents from 2015Q1-2016Q4 in EUR bn. Debt securities on the left and equity securities on the right hand side.
- Figure 14: Euro area net debt transactions by maturity
  - Cumulated net purchases by euro area residents from 2015Q1-2016Q4 in EUR bn.
- Figure 15: Euro area net debt transactions by maturity (sector split)
  - Cumulated net purchases by euro area residents from 2015Q1-2016Q4 in EUR bn.
- Figure 16: Euro area net transactions, changes in holdings and valuation effects in PSPP eligible securities, by sector
  - Cumulative values from 2015Q1-2016Q4 of PSPP eligible assets by sector in EUR bn.

### Tables included (Tables 1–7) — key specification details and exact numeric entries preserved
- Table 1: Baseline estimation: debt securities
  - Dependent variable: (adjusted) logarithm of cumulative net purchases of debt securities during the PSPP period (2015q1-2016q4).
  - Independent variables include: ln(Holdpre), ∆ Amount Out, EUR, ln(original maturity), PSPP Eligibility, PSPP Substitute, High Yield, Low Quality, Diff. Policy Uncertainty s−h, Diff. ∆ GDP Forecast s−h, Diff. ∆ Yield s−h.
  - Observations: 683007 (columns vary: 683007, 460687, 566399, 651402, 306654, 290771, 613093 as reported per column).
  - Holder country-sector FE: yes. Issuer country-sector FE: yes. Country-pair FE: yes (except last column where Country-pair FE = no).
  - Example coefficients and significance (exact values preserved): ln(Holdpre) -0.652***; ∆ Amount Out 0.505***; ln(original maturity) 0.564***; PSPP Eligibility -0.808***.
  - Notes include precise definitions and the statement: All differences refer to 2014Q4 to 2016Q4. T-statistics in brackets. * significant at 10% level; ** significant at 5% level, *** significant at 1% level.
- Table 2: Sectoral estimation: debt securities
  - Columns: MFI, ICPF, OFI, NFC, HH.
  - Observations: 683007.
  - Example coefficients: ln(Holdpre) ranges (-0.709***, -0.556***, -0.526***, -0.658***, -0.830***). ∆ Amount Out ranges (0.640***, 0.513***, 0.487***, 0.476***, 0.445***).
- Table 3: Geography estimations
  - Columns: ALL, MFI, ICPF, OFI, NFC, HH.
  - Observations: 683,007 (listed twice across the table).
  - Notes define ADV (advanced countries listed) and EM (emerging market economies) and preserve T-statistics in brackets and significance notation.
  - Example coefficients preserved: PSPP Eligibility -0.827*** (in column ALL).
- Table 4: Country-group estimation
  - Columns: Debt Stressed, Debt Non-stressed.
  - Observations: 566399 (one column shown; holder/issuer FE and country-pair FE = yes).
  - Example coefficients preserved: ln(Holdpre) -0.904*** (Stressed), -0.712*** (Non-stressed); ∆ Amount Out 0.547*** (Stressed), 0.495*** (Non-stressed); PSPP eligibility -1.750** (Stressed), -0.685* (Non-stressed).
  - Note: The group of formerly stressed countries consists of Cyprus, Greece, Italy, Portugal and Spain. Ireland and Luxembourg are not included in these estimations.
- Table 5: Extension of the baseline estimation: Time Dimension
  - Columns correspond to periods: From 14Q3 until 14Q4; From 15Q1 until 15Q2; From 15Q1 until 16Q2; From 15Q1 until 16Q4.
  - Observations: 347851; 689561; 684168; 683007 respectively.
  - Example coefficients preserved: ln(Holdpre) -0.431***; -0.298***; -0.589***; -0.652***. ∆ Amount Out 0.578***; 0.661***; 0.542***; 0.505***. PSPP eligibility 1.226***; -0.425*; -1.047***; -0.808***.
- Table 6: Time Dimension and PSPP eligibility: overall and by sector
  - Same observation counts as Table 5: 347851; 689561; 684168; 683007.
  - Part I (Overall specification): PSPP coefficient excerpts preserve values: 1.226***; -0.425*; -1.047***; -0.808***.
  - Part II (Sectoral specification) reports sectoral PSPP coefficients for MFI, ICPF, OFI, NFC, HH with exact reported values and significance markers.
- Table 7: Extension: Active vs. Passive Rebalancing and PSPP eligibility
  - Columns: Net transactions; Passive rebalancing; ∆ Holdings.
  - Observations: 331356 across columns.
  - Part I (Overall specification) coefficients preserved: Net transactions -1.541***; Passive rebalancing 1.492***; ∆ Holdings -0.0485.
  - Part II (Sectoral specification) reports sectoral coefficients for MFI, ICPF, OFI, NFC, HH with exact reported values and significance markers.
  - Notes specify dependent variables for columns and reiterate FE structure and significance notation.

*Content unit includes the full references list, Figures 1–16 with captions and notes, and Tables 1–7 with reported coefficients, observations, fixed-effect indicators, T-statistics notation, sample periods (e.g., 2015Q1-2016Q4, 2014Q4 to 2016Q4), and exact numerical entries as presented in the source.*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2020/english/wpiea2020046-print-pdf.pdf_