## _wp1582

## Source details

**Canonical URL:** [_wp1582](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp1582.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp1582.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp1582.pdf.json)

---

### I. Research question and conceptual framework
- Research question: Are Hong Kong SAR’s business cycles more synchronized with mainland China or the United States?
- Context and motivation:
  - Linked exchange rate system (LERS) with the Hong Kong dollar pegged to the U.S. dollar.
  - Hong Kong SAR acts as a gateway between mainland China and the rest of the world; structural shift from manufacturing to a service-based economy with manufacturing migrating to Guangdong province.
  - Key hypothesis: mainland China is more important for Hong Kong SAR’s trend growth; the United States remains dominant for cyclical fluctuations.
- Distinctions:
  - Stochastic trend shocks (permanent): linked to productivity and have permanent impact on income.
  - Transitory shocks (cyclical): include transitory productivity shocks and demand shocks, causing temporary fluctuations.
- Identification strategy:
  - Estimate stochastic income trends using real GDP and consumption for Hong Kong SAR, mainland China, and the United States via a theory-guided state-space model and hierarchical SVAR.
  - Test the permanent income hypothesis for real consumption in Hong Kong SAR and mainland China.
  - Trace transmission of permanent and transitory shocks and their evolution over time.

### II. Stylized facts on integration: trade, services, and finance
- Headline vs value-added export shares (2012):
  - Headline: share of Hong Kong SAR’s merchandise and service exports to mainland China = 51 percent; U.S. share = 21 percent.
  - Value-added adjustment (merchandise exports): mainland = 22 percent; U.S. = 25 percent.
- External demand and services:
  - U.S. final demand shocks could dominate external demand fluctuations for goods in Hong Kong SAR.
  - U.S. final demand in nontourism services exports remained larger than mainland’s (Figures referenced in source).
- Financial services exports (2012):
  - U.S. demand accounted for 33 percent of Hong Kong SAR’s financial services exports; mainland share = 4 percent.
- FDI, stock market, renminbi bonds:
  - Mainland importance increased as source of inward FDI and destination of outward direct investment; U.S. shares stable or declining.
  - Share of mainland companies in Hong Kong stock market capitalization larger since 2005.
  - Offshore renminbi bond issuance increased substantially since 2010.
- Tourism and structural transformation:
  - Tourism services exports to the mainland (value-added) rose from 28 percent in 2000 to over 66 percent in 2012; U.S. share dropped from 11 percent to 4 percent over same period.
  - Rise in labor productivity and strong TFP growth associated with expansion of services export sector and financial linkages with mainland.

### III. Data and econometric approach
- Data:
  - Seasonally adjusted quarterly real GDP and real consumption for Hong Kong SAR, mainland China, and the United States; all data end on 2013:Q2.
  - Hong Kong GDP: starts 1973:Q1 (seasonal adjustment applied for 1973–1989); U.S. GDP: starts 1947:Q1; mainland China constructed series begins 1978:Q1.
  - Real consumption: headline PCE for Hong Kong SAR and U.S.; mainland quarterly consumption interpolated using real retail sales and annual consumption.
- Decomposition and tests:
  - State-space (UC) model decomposes output and consumption into permanent (τ_it) and transitory components; τ_it follows a random walk with drift μ_i.
  - AR(1) assumed for transitory components u_iyt and u_ict.
  - Unit-root and Johansen cointegration tests, VECM estimation, and tests of Hall (1978) implication for predictability of consumption by anticipated transitory income.
  - Hierarchical SVAR (Genberg-style) used to trace transitory and permanent shock transmission across US, CN, HK blocks.
  - Alternative robustness: Kim and Piger (2000) approach and DOLS/Stock-Watson DOLS cointegration estimates.

### IV. Tests of permanent income hypothesis and short-run dynamics
- Unit-root and cointegration:
  - Cannot reject unit root for output and consumption for Hong Kong SAR and mainland China at 5 percent; first differences reject at 1 percent → I(1).
  - Output-consumption gaps reject unit root at 5 percent → cointegration and common stochastic trend for both economies.
  - Johansen tests: null of no cointegrating vectors rejected at 5 percent for Hong Kong SAR; for mainland China rejected at 5 and 10 percent depending on statistic; null of one cointegrating vector not rejected.
- Estimated output-consumption gap formulas:
  - Hong Kong SAR: ẑ_t = ŷ_t − 0.07 − 0.97 ĉ_t.
  - Mainland China: ẑ_t = ŷ_t − 1.18 − 1.7 ĉ_t.
- VECM error-correction speeds:
  - Consumption adjusts in Hong Kong SAR at a quarterly frequency equivalent to 7 to 8 percent of the deviation from long-run consumption.
  - Consumption adjusts in mainland China at a speed of 9 to 10 percent of the consumption deviation.
  - Speeds faster than those in Cochrane (1994) and Morley (2007).
- Consumption responsiveness to lagged transitory output:
  - Hong Kong SAR: after controlling for long-term relationship, 1 percentage point increase in lagged transitory output predicts about 0.2 percentage point increase in consumption.
  - Mainland China: reported effect approximately 0.16 to 0.18 percentage point increase in consumption (source reports 0.16 to 0.18 percentage point).
  - Interpretation: coefficients ≤ 0.2 imply a smaller share of liquidity-constrained households compared with Campbell and Mankiw (1989) U.S. estimate of 0.5.

- VECM fit diagnostics (selected):
  - Hong Kong SAR R2 entries: 0.4313, 0.4028, 0.4458, 0.394 (across specifications).
  - Mainland China R2 entries: 0.7805, 0.6875, 0.7839, 0.6841 (across specifications).
  - Representative coefficient magnitudes preserved from source tables (examples listed in source).

### V. Unobserved component estimates of permanent income
- UC model equations (as in source):
  - i t y = τ i t + u i yt
  - i t c = c i + γ i τ i t + u i ct
  - τ i t follows random walk with drift μ i and shock v i t.
- γ_i interpreted as marginal propensity to consume out of permanent income.
- Figures 10–12 in source plot qoq growth of GDP, consumption, and estimated permanent income for US, CN, HK.
- Transitory income constructed as real output minus estimated permanent income.

### VI. Transmission results — transitory shocks (SVAR)
- Hierarchical SVAR with blocks for US, CN, HK; sample splits to assess evolution:
  - Subperiods: 1985:Q1–1997:Q2 and 2003:Q4–2013:Q2 (Asian financial crisis and SARS excluded).
- Variance-decomposition highlights for Hong Kong SAR (transitory shocks):
  - Short-run: Hong Kong SAR’s own shocks dominate its output and inflation variations.
  - 20-quarter horizon:
    - Earlier period (1985:Q1–1997:Q2): U.S. transitory shocks explain about 57 percent of Hong Kong SAR’s output fluctuations.
    - Later period (2003:Q4–2013:Q2): U.S. transitory shocks explain about 48 percent.
  - Inflation long-run influences (latter period):
    - Mainland shocks account for about 36 percent of Hong Kong SAR inflation variations.
    - U.S. influence accounts for about 37 percent of inflation variations.
- Impulse responses: transitory output shocks behave like demand shocks (transitory output and inflation move in same direction for own transitory output shock).
- Alternative measures (Section IV.A / Table C3) — selected numeric blocks preserved in source:
  - 1985:Q1–1997:Q2 output horizon entries include numeric values such as US 1: 17.91, CN 93.25, HK 78.82 (table-format blocks as presented).
  - 1999:Q1–2012:Q4 output horizon entries include values such as US 1: 15.58, CN 21.68, HK 82.73.
- Qualitative summary:
  - U.S. transitory shocks historically and persistently a major force behind Hong Kong SAR’s business cycle fluctuations; their impact diminished somewhat in later samples.
  - Mainland transitory shocks’ effect on Hong Kong SAR’s real economy dropped over time.

### VII. Transmission results — permanent shocks (SVAR on trends)
- SVAR applied to estimated permanent income series τ̂_it.
- Variance-decomposition (Hong Kong SAR permanent income) — selected outcomes:
  - U.S. long-run influence declined from about 36 percent (1985:Q1–1997:Q2) to about 30 percent (2003:Q4–2013:Q2).
  - Mainland China long-run influence increased from about 15 percent (1985:Q1–1997:Q2) to about 65 percent (2003:Q4–2013:Q2).
- Alternative measures (Table C4) preserved numeric blocks as in source:
  - Examples include trend-share blocks such as 1985:Q1–1997:Q2 entries 17.91, 93.25, 78.823 and later period blocks including 11.015, 23.924, 65.061 (as presented).
- Qualitative interpretation:
  - Mainland China became a far more dominant force behind Hong Kong SAR’s trend growth.
  - U.S. influence remained dominant for cyclical frequencies but declined in relative importance for trend/permanent shocks.

### VIII. Robustness and alternative specifications
- DOLS and OLS estimates for cointegrating coefficients (selected entries preserved):
  - OLS: Mainland China 1.0473 (0.9956, 6.79 6.84); U.S. 0.9057 (0.9978, 2.93, 2.98); Hong Kong SAR 1.0327 (0.9718, 6.28, 6.33).
  - DOLS (1 to 7 lags/leads) show Mainland China coefficients increasing across specifications: e.g., 1.0578, 1.0626, 1.0664, 1.0699, 1.0751, 1.079, 1.0828; U.S. coefficients around 0.9113–0.9177; Hong Kong SAR coefficients around 1.0473–1.0917 (numerical tuples preserved as in source).
- Appendix C alternative measure: consumption proxied as stochastic trend and Stock-Watson DOLS cointegrating vectors close to theory-implied (1, −1); robustness supports main findings that mainland’s importance for Hong Kong SAR trend increases in later sample.

### IX. Main findings and policy implications
- Main empirical findings:
  - U.S. transitory shocks remain the dominant drivers of Hong Kong SAR’s business cycle fluctuations.
  - Mainland China transitory shocks played a less important role for Hong Kong SAR’s cycles and their cyclical influence dropped over time.
  - Mainland China permanent shocks explain a much larger portion of Hong Kong SAR’s trend volatility in output in the later sample (rise from ~15 percent to ~65 percent influence).
  - U.S. influence on Hong Kong SAR permanent income declined modestly (from about 36 percent to about 30 percent across sample splits).
- Policy implications:
  - Exchange rate regime: because Hong Kong SAR is more synchronized with the United States at business-cycle frequencies, the LERS that pegs the Hong Kong dollar to the U.S. dollar remains appropriate for the foreseeable future (cyclical synchronization matters most for exchange rate choice).
  - Structural/economic transformation: Hong Kong SAR has benefited from mainland China’s rise through trade, FDI, transformation to a service economy, and exchange of human capital and knowledge supporting longer-term productivity progress.

*Source: _wp1582 (excerpted content provided).*

### References .............................................................................................................

### _wp1582 - References .............................................................................................................

### I. INTRODUCTION
- Research question: Are Hong Kong SAR’s business cycles more synchronized with mainland China or the United States?
- Context and motivation:
  - Hong Kong SAR: linked exchange rate system (LERS) with the Hong Kong dollar pegged to the U.S. dollar.
  - Geographic role: gateway between mainland China and the rest of the world.
  - Structural shift: transformation from manufacturing to a service-based economy; manufacturing migrated to Guangdong province.
  - Key hypothesis: mainland China is more important for Hong Kong SAR’s trend growth; the United States remains dominant for cyclical fluctuations.
- Conceptual distinctions:
  - Stochastic trend shocks (permanent): linked to productivity and have permanent impact on income.
  - Transitory shocks (cyclical): include transitory productivity shocks and demand shocks, causing temporary fluctuations.
- Identification strategy:
  - Use real GDP and consumption data to estimate stochastic income trends for Hong Kong SAR, mainland China, and the United States.
  - Approach built on Fama (1992), Cochrane (1994), Kim and Piger (2000), Aguiar and Gopinath (2007).
  - Avoids conventional filtering techniques (e.g., HP filter); employs a theory-guided state-space model using consumption and output to identify permanent income.
  - Confirms whether real consumption in Hong Kong SAR and mainland China satisfy the permanent income hypothesis.
  - Uses a hierarchical, structural vector autoregression (SVAR) model to analyze transmission of permanent and transitory shocks from mainland China and the United States to Hong Kong SAR and their evolution over time.

### II. STYLIZED FACTS ON ECONOMIC INTEGRATION
- Headline trade and financial flow observations:
  - Headline figures suggest increasing mainland China influence on Hong Kong SAR through tighter trade and financial linkages.
  - Caveat: headline export destination data do not necessarily indicate where final demand originates; substantial re-exporting from mainland China to advanced economies is important.
- Specific comparative shares reported (headline vs value-added):
  - Headline measure (2012): share of Hong Kong SAR’s merchandise and service exports to mainland China = 51 percent; U.S. share = 21 percent.
  - Value-added adjustment (2012): share of merchandise exports to mainland China in value-added terms = 22 percent.
  - Value-added adjustment (2012): U.S. share of merchandise exports in value-added terms = 25 percent.
- Implications:
  - Final-demand attribution differs materially when accounting for import content and value-added.
  - Hong Kong SAR’s role as a “bridge” suggests its cyclical conditions are tied to flows between the mainland and major trading partners; the bridge role is expected to remain important even as flows more closely tied to mainland domestic demand gain significance.

### III. DATA, METHODS, AND MODELING APPROACH
- Data used:
  - Real GDP and consumption series for Hong Kong SAR, mainland China, and the United States.
- Decomposition methodology:
  - State-space model applied to consumption and output to derive permanent and transitory components of income for each economy.
  - Tests of the permanent income hypothesis for real consumption in Hong Kong SAR and mainland China.
- Time-series econometric steps mentioned:
  - Unit root and cointegration testing for output and consumption (Tables referenced).
  - Vector error correction models (VECM) estimated for output and consumption for Hong Kong SAR and mainland China (Tables referenced).
  - Hierarchical structural VAR used to trace effects of permanent and transitory shocks across economies.
- Rationale for methodological choices:
  - Preference for theoretically guided decomposition over filtering (e.g., HP filter) to avoid artificial business cycle dynamics.

### IV. MAIN FINDINGS
- Role of the United States:
  - Transitory shocks from the United States remain a major driving force behind Hong Kong SAR’s business cycle fluctuations.
- Role of mainland China:
  - Hong Kong SAR and mainland China share a strong co-movement in long-run trend growth.
  - Permanent shocks originating from the mainland have substantial influence on Hong Kong SAR’s trend growth, reflecting ongoing economic and social integration.
- Synthesis of trends vs cycles:
  - Evidence supports the hypothesis that different external forces drive trend and cyclical components:
    - Mainland China: larger role in Hong Kong SAR’s trend (permanent) growth.
    - United States: larger role in Hong Kong SAR’s cyclical (transitory) fluctuations.

### V. PAPER ORGANIZATION (as provided)
- Section II: Stylized facts on economic integration among Hong Kong SAR, mainland China and the United States.
- Section III: Estimation of stochastic trends for each economy.
- Section IV: Analysis of how permanent and transitory shocks from mainland China and the United States affect Hong Kong SAR.
- Section V: Conclusion.

*Source: _wp1582 - References .............................................................................................................*

### 2012. The U.S. share would be even more significant if we take into account its influence on other

### _wp1582 - 2012. The U.S. share would be even more significant if we take into account its influence on other

### External demand and export shares
- Merchandise exports (value-added) and services exports excluding tourism (value-added) show the U.S. and mainland China shares over 2000–2012 with series plotted in Figures 1 and 2.
- The U.S. final demand shocks could still be the dominating force that drives fluctuations in external demand for goods in Hong Kong SAR.
- The U.S. share in Hong Kong SAR’s services export excluding tourism services continued to be larger than mainland’s share (Figure 2), implying that final demand from the United States in nontourism services exports remained larger than demand from mainland China.

### Financial services exports
- Within Hong Kong SAR’s financial services export, the U.S. demand accounted for 33 percent of the total in 2012, whereas the mainland share was 4 percent (Figure 3).
- Hong Kong SAR has been transforming into an international financial centre by:
  - providing intermediation services between users of funds and global investors;
  - expanding the supply of financial products, including initial public offerings (IPOs) and renminbi bonds, to raise funds for mainland corporations;
  - targeting overseas investors seeking exposure to mainland-related financial assets.
- Overseas investors have been driving demand for mainland-related financial products supplied through Hong Kong SAR.
- Ongoing financial integration between Hong Kong SAR and mainland is a supply-side factor contributing to rising productivity of the Hong Kong SAR economy.

### FDI, stock market capitalization, and offshore renminbi bonds
- Figures 4 and 5 show an increasing importance of the mainland as a source of inward foreign direct investment (FDI) and as a destination of outward foreign direct investment (ODI); by comparison, U.S. shares have been stable or declining in recent years.
- Share of mainland companies in Hong Kong stock market capitalization has become larger since 2005 (Figure 6).
- Issuance of offshore renminbi bonds has increased substantially since 2010 as Hong Kong SAR develops into a major offshore renminbi centre (Figure 7).

### Tourism services and structural transformation
- Tourism services exports to the mainland (value-added, stripping out import content) rose from 28 percent of total tourism services exports in 2000 to over 66 percent in 2012; the U.S. share dropped from 11 percent to 4 percent over the same period (Figure 8).
- The secular trend in tourism services exports to mainland China reflects mainland’s demand for Hong Kong’s tourism services and has contributed to Hong Kong’s transformation toward a service-based economy from a supply-side perspective.
- The size of Hong Kong’s service sector has been rising along with the increase in labor productivity as shown in Figure 9.
- The rise in labor productivity has been underpinned by strong total factor productivity (TFP) growth as a result of expansion of services export sector; a large part of TFP growth might have been boosted by increasing financial linkages between Hong Kong SAR and the mainland, especially given the high value-added content of financial services.
- These facts point to the possibility that mainland China has been a major force affecting Hong Kong SAR’s trend growth.

### Estimating trends and cycles — Model motivation
- Adopt a theory-based method to identify trends and cycles in output data using a standard stochastic growth model.
- Output yt produced by capital Kt and labor Lt, subject to trend productivity At and transitory productivity shock zt:
  - y_t = z_t * K_t^α * (A_t L_t)^(1-α)  (equation (1) as presented)
- Productivity trend At follows a random walk with shocks g_t:
  - A_t = A_0 * Π_{s=1}^{t} e^{g_s}  (equation (2) as presented)
- A realization of g_t affects A_t permanently, so output is nonstationary containing a stochastic trend.
- Representative agent maximizes lifetime utility U(C_t, L_t−ε_t) with discounting β (equation (3) as presented); ε_t denotes consumption shock, a demand shock with temporary effect.
- Guidance: treat the common stochastic trend shared by output and consumption as a measure of permanent income. The difference between actual output and stochastic trend is transitory income.
- Use a simple state-space model to identify the common stochastic trend between output and consumption (“permanent income”).
- As robustness, follow Kim and Piger (2000) using consumption as proxy for stochastic trend and decompose output into trends and cycles (Appendix C).

### Test for Permanent Income Hypothesis — data and procedure
- Data: real GDP and real consumption for Hong Kong SAR, mainland China, and the United States in quarterly frequencies; all data end on the second quarter of 2013.
- Seasonally adjusted real GDP data:
  - Hong Kong starts from the first quarter of 1973 (note: seasonally adjusted real GDP data for Hong Kong SAR only goes back to the first quarter of 1990, so X12-ARIMA was used to perform seasonal adjustment on Hong Kong real GDP data between 1973 and 1989).
  - United States series starts from the first quarter of 1947.
  - Mainland China constructed series begins from the first quarter of 1978 (construction used multiple series as documented).
- Real consumption:
  - Use headline real personal consumption expenditure (PCE) for Hong Kong SAR and the United States.
  - For mainland China, quarterly real consumption is estimated by interpolation procedures (Chow and Lin (1971); Bloem, Dippelsman and Maehle (2001)) based on real retail sales and annual real consumption data; the real retail sales series starts from the first quarter of 1985.
- Testing approach to verify permanent income hypothesis for Hong Kong SAR and mainland China:
  1. Test stationarity of the gap between output and consumption, where gap ẑ_t = ŷ_t − α̂ − β̂ ĉ_t with α̂ and β̂ OLS estimates from regressing output on consumption in logs.
     - Reported output-consumption gaps: Hong Kong SAR ẑ_t = ŷ_t − 0.07 − 0.97 ĉ_t; mainland China ẑ_t = ŷ_t − 1.18 − 1.7 ĉ_t.
  2. Conduct Johansen cointegration tests.
  3. Estimate a vector error correction model (VECM) to test if consumption is not predictable by anticipated transitory income change (Hall (1978) implication).
- Unit-root and cointegration test outcomes:
  - Cannot reject unit root for output and consumption for both Hong Kong SAR and mainland China at the 5 percent level; first differences reject unit root at the 1 percent level, implying I(1) processes.
  - Reject unit root for the output-consumption gaps at the 5 percent level, implying cointegration and existence of a stochastic trend between output and consumption for both economies.
  - Johansen tests: null of no cointegrating vectors rejected at 5 percent for Hong Kong SAR, and at 5 and 10 percent for mainland China depending on test statistic; null of one cointegrating vector is not rejected.
- VECM results and implications:
  - Error correction coefficients indicate consumption adjusts in Hong Kong SAR at a quarterly frequency equivalent to 7 to 8 percent of the deviation from long-run consumption.
  - Consumption adjusts in mainland China at a speed of 9 to 10 percent of the consumption deviation.
  - Speeds are much faster than those found in Cochrane (1994) and Morley (2007).
  - Testing Hall (1978) implication: consumption does exhibit some responses to changes in transitory income.
    - Coefficient on first lagged output growth in the consumption growth equation is positive and significant at the 5 percent level under the restricted VECM for Hong Kong SAR.
    - Coefficient is positive and significant in both restricted and unrestricted VECMs for mainland China.
    - After controlling for long-term relationship, 1 percentage point increase in lagged transitory output can still predict about 0.2 percentage point increase in Hong Kong SAR’s consumption (and about [text truncated in source at this point for mainland China]).

*Source: CEIC; C&SD; Hong Kong Stock Exchange; Newswires; authors’ estimates (content as provided in the supplied PDF excerpt).*

### 0.16 to 0.18 percentage point increase in mainland’s consumption.

### _wp1582 - 0.16 to 0.18 percentage point increase in mainland’s consumption.

### Consumption behavior and interpretation
- Reported effect: 0.16 to 0.18 percentage point increase in mainland’s consumption.
- Empirical interpretation:
  - Cointegration tests suggest a common stochastic trend exists between output and consumption in both Hong Kong SAR and mainland China, consistent with a permanent productivity/permanent income trend.
  - Coefficient on lagged output growth is positive and significant but with a magnitude of 0.2 or less.
  - Comparison: an estimate of 0.5 (Campbell and Mankiw (1989) for U.S. data) implies roughly half of U.S. consumers are “rule-of-thumb” (consume current income rather than permanent income); by contrast, the smaller coefficients here imply a smaller share of liquidity-constrained households.
- Two-type consumer explanation:
  - Liquidity-constrained households respond to temporary income shocks; unconstrained households respond to permanent income.
  - Evidence from other studies: constrained households share estimated at 20–40 percent of total in Benito and Mumtaz (2012) and Faruqui and Torchani (2012); after grouping, constrained households’ consumption responds positively to lagged income growth while unconstrained do not.
  - Interpretation: small coefficients on lagged income growth capture behavior of liquidity-constrained households; the rest follow permanent income hypothesis.

### VECM / Short-run dynamics (selected coefficients and fit)
- Hong Kong SAR VECM (Table 3) — selected diagnostics:
  - R2: 0.4313, 0.4028, 0.4458, 0.394 (reported across specifications).
  - Representative coefficient magnitudes (preserve reported signs/values in source tables): e.g., 0.198, 0.143, 0.122, 0.195, 0.089, 0.2, -0.321, 0.217, -0.282, 0.233, 0.120, 0.134, 0.142, 0.149, 0.297, 0.017, 0.309, 0.025, 0.00, 0.0, -0.001, 0.005 (standard errors reported in parentheses in source).
- Mainland China VECM (Table 4) — selected diagnostics:
  - R2: 0.7805, 0.6875, 0.7839, 0.6841 (reported across specifications).
  - Representative coefficient magnitudes (preserve reported signs/values in source tables): e.g., -0.091, -0.094, -0.096, -0.076, 0.177, -0.174, 0.155, -0.201, 0.105, -0.154, 0.09, -0.174, 0.142, -0.402, 0.131, -0.419, 0.167, 0.114, 0.164, 0.105, -0.145, 0.14, -0.121, 0.162, 0.204, 0.268, 0.228, 0.287, -0.024, 0.314, 0.002, 0.327, 0.064, 0.521, 0.082, 0.528, -0.004, 0.004, -0.005, 0.006 (standard errors reported in parentheses in source).

### Estimating permanent income: Unobserved component (UC) model
- Model structure (preserve notation and equations as in source):
  - Real output and consumption decomposed as:
    - i t y = τ i t + u i yt
    - i t c = c i + γ i τ i t + u i ct
  - τ i t follows a random walk with drift μ i and shock v i t.
  - Transitory components u i yt and u i ct follow finite-order AR processes; in estimation AR(1) assumed.
  - Notation: i ∈ {US, CN, HK}; i t y and i t c are logarithms of real GDP and consumption respectively; γ i is marginal propensity to consume out of permanent income.
- Estimation outputs and data:
  - Figures 10–12 (in source) plot qoq growth of GDP, consumption, and estimated permanent income for Hong Kong SAR, mainland China, and United States.
  - Transitory income measure constructed as real output minus estimated permanent income for each economy.

### Transmission of transitory shocks (SVAR set-up and key results)
- Model: hierarchical structural VAR following Genberg (2005) and Genberg, Liu, and Jin (2006); blocks for US, mainland China, and Hong Kong SAR with economy-specific variables; ordering captures growing importance of mainland China.
- Sample splits to assess post-handover evolution:
  - Subperiods estimated: 1985:Q1–1997:Q2 and 2003:Q4–2013:Q2 (skipping Asian financial crisis and SARS period).
- Key variance-decomposition results for Hong Kong SAR (transitory shocks):
  - Short-term horizons: shocks originating from Hong Kong SAR dominate Hong Kong SAR’s own output and inflation variations.
  - 20-quarter horizon:
    - U.S. transitory shocks explain about 57 percent of Hong Kong SAR’s output fluctuations in the earlier period (1985:Q1–1997:Q2).
    - U.S. transitory shocks explain about 48 percent in the latter period (2003:Q4–2013:Q2).
  - Hong Kong SAR inflation long-run influences in latter period:
    - Mainland shocks account for about 36 percent of inflation variations.
    - U.S. influence accounts for about 37 percent of inflation variations.
- Additional note from impulse responses: output shocks behave like demand shocks since impulse responses of transitory output and inflation move in same direction for an economy’s own transitory output shock.

### Transmission of permanent shocks (SVAR on estimated trends) — key findings
- SVAR using estimated permanent income series τ̂ i t as inputs (state-space UC estimates).
- Variance-decomposition results for Hong Kong SAR permanent income (Table 6) — selected outcomes:
  - U.S. long-run influence on Hong Kong SAR permanent income:
    - Declined from about 36 percent between 1985:Q1 and 1997:Q2 to 30 percent during 2003:Q4–2013:Q2.
  - Mainland China long-run influence on Hong Kong SAR permanent income:
    - Increased from about 15 percent in 1985:Q1–1997:Q2 to about 65 percent in 2003:Q4–2013:Q2.
- Interpretation:
  - Mainland China has become a more dominant force behind Hong Kong SAR’s trend growth.
  - U.S. influence remained dominant at business-cycle frequencies but has declined in relative importance for trend (permanent) shocks.

### Conclusions and policy relevance
- Main findings:
  - U.S. transitory shocks remain the dominant drivers of Hong Kong SAR’s business cycle fluctuations.
  - Mainland China’s transitory shocks played a less important role for Hong Kong SAR’s cycles.
  - In contrast, mainland China’s permanent shocks now explain a much larger portion of Hong Kong SAR’s trend volatility in output than U.S. permanent shocks.
- Policy implication for exchange rate regime:
  - Because Hong Kong SAR is more synchronized with the United States at business-cycle frequencies, the LERS (Link Exchange Rate System) that pegs the Hong Kong dollar to the U.S. dollar remains appropriate for the foreseeable future (cyclical synchronization matters most for exchange rate choice).
- Structural/economic transformation implication:
  - Hong Kong SAR has benefited from mainland China’s rise through trade, FDI, transformation to a service economy, and active exchange of human capital and knowledge propagating longer-term productivity progress across the border.

### Appendices and robustness checks (methodological notes)
- Appendix A: Construction of value-added exports
  - Merchandise exports in value-added terms = domestic exports + re-export margins + commissions from offshore trade; adjustments applied to domestic exports and re-exports using C&SD and China NBS data; tourism services adjusted by stripping import content in visitor shopping expenditure.
- Appendix B: Coefficient estimates of state-space (UC) models
  - Coefficients and standard errors reported; shocks’ standard errors multiplied by 100 in table note; ρ i denotes correlations between shocks.
  - Example parameter estimates and diagnostics reported in source (preserve reported values where presented).
- Appendix C: Alternative measure of permanent income (robustness)
  - Alternative specification: consumption itself as stochastic trend (consumption proxied by real retail sales for mainland China).
  - Cointegrating coefficients estimated via Stock-Watson DOLS; model selection via AIC and BIC.
  - Cointegration vectors for all three economies close to theory-implied value (1, -1).
  - Robustness results largely consistent with main findings: mainland’s importance for Hong Kong SAR trend increases in later sample.

*Source: _wp1582 - 0.16 to 0.18 percentage point increase in mainland’s consumption.*

### Section IV.A using the alternative measures of

### _wp1582 - Section IV.A using the alternative measures of

### Methodology and data
- Analysis uses alternative measures of transitory outputs, inflation rates of the three economies (Mainland China, U.S., Hong Kong SAR), and the three-month U.S. Treasury bill rates, at quarterly frequency.
- An alternative model described in Section IV.B is also estimated.
- DOLS (Dynamic OLS) estimates reported with varying numbers of lags/leads (1 through 7) alongside OLS; parameter estimates from DOLS are noted as super consistent, with the rate of convergence given by the sample size instead of its square root, implying no generated-regressors problem for subsequent VECM inference.

### Transmission of transitory shocks to Hong Kong SAR (variance decompositions, Table C3)
- Time samples and horizons covered include 1985:Q1 to 1997:Q2 and 1999:Q1 to 2012:Q4 (quarterly horizons reported).
- Findings on transitory shocks (output and price):
  - 1985:Q1 to 1997:Q2 — Output (Horizon in quarters): US 1: 17.91, CN 93.25, HK 78.82; Price (Horizon in quarters): US 32.42, CN 51.11, HK 86.46, CN 3.88, HK 5. (Note: table entries organized by horizon and shock shares; exact numeric blocks preserved below.)
  - 1999:Q1 to 2012:Q4 — Output (Horizon in quarters): US 1: 15.58, CN 21.68, HK 82.73; Price (Horizon in quarters): US 14.06, CN 30.32, HK 85.61.  
- Qualitative summary:
  - The U.S. transitory shocks had a large impact on Hong Kong SAR before 1997, explaining more than 50 percent of Hong Kong SAR’s output fluctuation at the 4-quarter horizon and an even larger share at longer horizons.
  - In the latter sample period, U.S. transitory shocks remained important for Hong Kong SAR’s business cycles, but the magnitude of their impact decreased at all frequencies.
  - Mainland China transitory shocks’ effect on Hong Kong SAR’s real economy also dropped over time, implying that Hong Kong SAR’s economic integration with the mainland did not increase output co-movement at business cycle frequencies.
  - Consistent with results using transitory income from state-space models, U.S. transitory shocks remained a dominant force behind Hong Kong SAR’s business cycle.

### Permanent shocks and Hong Kong SAR trend growth (variance decompositions, Table C4)
- Table C4 reports variance decompositions for Hong Kong SAR’s permanent income (trend).
- Key quantitative excerpts (trend shocks):
  - 1985:Q1 to 1997:Q2 — Trend_US 1: 17.91, Trend_CN 93.25, Trend_HK 78.823.
  - 1997:Q3 to 2003:Q4 — Trend_US 1: 15.58, Trend_CN 21.688, Trend_HK 82.730.
  - 2004:Q1 to 2013:Q2 — Selected entries: 11.015, 23.924, 65.061; 41.3183, 32.967, 53.581; 2011.606, 42.724, 46.67. (Preserved numeric blocks as presented.)
- Qualitative summary:
  - In contrast to transitory shocks, permanent shocks from mainland China were more important than the United States in determining Hong Kong SAR’s trend growth.
  - A permanent shock to the mainland leads to a notable shift in Hong Kong SAR’s output trend.

### DOLS and OLS parameter estimates (excerpted table)
- OLS and DOLS estimates for three economies (Mainland China, U.S., Hong Kong SAR) with confidence or related tuples shown:
  - OLS: Mainland China 1.0473 (0.9956, 6.79 6.84); U.S. 0.9057 (0.9978, 2.93, 2.98); Hong Kong SAR 1.0327 (0.9718, 6.28, 6.33)
  - DOLS with 1 lag/lead: Mainland China 1.0578 (0.9960, 6.66 6.78); U.S. 0.9113 (0.9980, 2.84, 2.97); Hong Kong SAR 1.0473 (0.9725, 6.27, 6.39)
  - DOLS with 2 lags/leads: Mainland China 1.0626 (0.9962,6.61 6.78); U.S. 0.9127 (0.9982, 2.84, 3.01); Hong Kong SAR 1.0567 (0.9728, 6.26, 6.43)
  - DOLS with 3 lags/leads: Mainland China 1.0664, U.S. 0.9140, Hong Kong SAR 1.0691 (0.9963, 6.56, 6.79)(0.9980, 2.85, 3.08)(0.9732, 6.24, 6.47)
  - DOLS with 4 lags/leads: Mainland China 1.0699 (0.9964, 6.50, 6.79); U.S. 0.9159 (0.9981, 2.86, 3.13); Hong Kong SAR 1.0805 (0.9733, 6.24, 6.52)
  - DOLS with 5 lags/leads: Mainland China 1.0751 (0.9968, 6.39, 6.73); U.S. 0.9177 (0.9981, 2.87, 3.19); Hong Kong SAR 1.0917 (0.9734, 6.24, 6.57)
  - DOLS with 6 lags/leads: Mainland China 1.079 (0.9969, 6.34, 6.74)
  - DOLS with 7 lags/leads: Mainland China 1.0828 (0.9970, 6.29, 6.75)

### Overall interpretation
- U.S. transitory shocks have historically been the primary driver of Hong Kong SAR’s business-cycle fluctuations, particularly before 1997, and remain important though somewhat diminished in the later sample.
- Mainland China’s permanent shocks play a larger role than U.S. permanent shocks in shaping Hong Kong SAR’s trend (permanent) output growth, indicating stronger influence of mainland events on Hong Kong SAR’s long-run trend than on cyclical fluctuations.

*Source: _wp1582 - Section IV.A using the alternative measures of*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2015/_wp1582.pdf_