## 1. Apple Inc. and Samsung Electronics Smartphone Release Dates

## Source details

**Canonical URL:** [1. Apple Inc. and Samsung Electronics Smartphone Release Dates](https://www.imf.org/-/media/files/publications/wp/2018/wp1822.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2018/wp1822.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2018/wp1822.pdf.json)

---

### Introduction and importance of the new tech cycle
- The new tech cycle is defined as the production and export of electronics and software associated with the release of new smartphones by global producers.
- Since 2012, global sales shifted from PCs to smartphones; in 2016 global smartphone sales reached close to 1.5 billion units, equivalent to one for every fifth person on earth.
- At peak (2016), exports of electronic components for smartphone production accounted for:
  - more than one third of total exports in Taiwan Province of China,
  - 15 percent in Singapore and South Korea,
  - 11 percent in Malaysia.
- In 2016, China exported USD107 billion worth of smartphones to the rest of the world, equivalent to 5 percent of its total exports.
- OECD data cited: CEO sectors accounted for 7.4 percent of total value added in South Korea in 2013; 2.1 percent in Japan and 2.0 percent in Ireland in 2012.

### Structure of the tech cycle: pre-release and post-release components
- The cycle is subdivided into:
  - Pre-release cycle: exports of components from several Asian countries to China for final assembly.
  - Post-release cycle: shipments of finished smartphones from China to the rest of the world.
- The cycle is driven by smartphone release dates (Apple and Samsung), not by standard seasonal patterns.

### Pre-release cycle: composition, geography, and magnitudes
- Pre-release shipments include:
  - memory chips (DRAM and flash memory) mainly from South Korea and Vietnam;
  - system chips (CPUs and app processors) and multichip packages (MCPs) from Japan, Malaysia, Singapore, South Korea and Taiwan Province of China;
  - displays from Japan and South Korea.
- Operating system sources:
  - Google Android: open source from the US (free);
  - Apple's operating system: from Ireland (Apple Inc.’s intellectual property location per cited source).
- Geography (2016): main contributors were Japan, Malaysia, Singapore, South Korea, and Taiwan Province of China; newcomers include the Philippines, Thailand, and Vietnam.
- Aggregate magnitudes:
  - The pre-release cycle accounted for USD68.2 billion of imports of electronics components to China in 2016, equivalent to about 4.3 percent of total Chinese merchandise imports.
  - Estimated value added from smartphone manufacturing and shipments in mainland China: USD20 billion, equivalent to 0.2 percent of 2016 GDP.

### Post-release cycle: destinations and shares (2016)
- Chinese exports of smartphones in 2016: USD106.9 billion (5.1 percent of total exports).
- Main destinations and values/shares:
  - US & Canada: USD29.9 Billion, 28.0%
  - Europe: USD24.3 Billion, 22.7%
  - South East Asia & India: USD16.4 Billion, 15.4%
  - South Korea: USD8.0 Billion, 7.5%
  - Japan: USD7.6 Billion, 7.1%
  - Latin America: USD6.3 Billion, 5.9%
  - MENA: USD7.1 Billion, 6.6%
  - Russia & CIS: USD3.0 Billion, 2.8%
  - Sub-Sahara: USD2.2 Billion, 2.1%
  - Australia and New Zealand: USD1.9 Billion, 1.8%
- Post-release shipments originate from China to regional distribution centers and then to consumers; shipments to South Korea reflect distribution of a portion of Samsung smartphones.

### Why standard seasonal adjustment fails to capture the tech cycle
- Reasons given:
  - Recent nature of the cycle (about six years) limits statistical capture by standard seasonal adjustment needing longer time series.
  - Rapidly evolving and increasingly complex supply chains across countries.
  - Cycle critically depends on smartphone release dates which vary within the calendar year and differ in impact depending on product success.

### Decomposing trend from cycle: methodology
- Monthly dummies before and after release dates capture the cycle.
- Time and time^2 capture a potentially non-linear trend.
- A Chinese New Year dummy (equal to the number of vacation days in each month) captures the roughly 15-day production shutdown that shifts about 11 days each year in the Gregorian calendar.
- The methodology is applied to:
  - China smartphone exports (units),
  - South Korea semiconductor exports (units),
  - Taiwan Province of China electronic export orders (value).

### Empirical results — China smartphone exports
- Main findings:
  - The cycle is mainly determined by iPhone release dates; Samsung release dates were not statistically significant.
  - Coefficients on one month before the release date and up to four months after are statistically significant.
  - Trend (time and time^2) is non-linear with a peak in September 2015.
  - Chinese New Year dummy coefficient is negative and highly significant.
  - Decomposition (cycle + trend + controls) explains about 92 percent of Chinese smartphone exports.
- Caution noted: lack of data for the second half of 2017 may overstate recent downturn in the trend.
- Regression results (Dependent Variable: Volume of China Smartphone Exports (in log)):
  - M-3: 0.02238; Std. Error 0.029; t 0.77; P>|t| 0.45
  - M-2: 0.03710; Std. Error 0.029; t 1.27; P>|t| 0.21
  - M-1: 0.05686; Std. Error 0.032; t 1.80; P>|t| 0.01
  - M 0.15978; Std. Error 0.032; t 5.10; P>|t| 0.00
  - M+1 0.17295; Std. Error 0.032; t 5.50; P>|t| 0.00
  - M+2 0.29625; Std. Error 0.031; t 9.41; P>|t| 0.00
  - M+3 0.24075; Std. Error 0.029; t 8.36; P>|t| 0.00
  - M+4 0.14691; Std. Error 0.029; t 5.16; P>|t| 0.00
  - Constant -70.04554; Std. Error 9.640; t -7.27; P>|t| 0.00
  - Time 0.26577; Std. Error 0.029; t 9.06; P>|t| 0.00
  - Time2 -0.0002; Std. Error -0.000; t -8.94; P>|t| 0.00
  - Chinese New Year dummy* -0.03106; Std. Error 0.002; t -13.96; P>|t| 0.00
  - R-squared 0.922
  - Adjusted R-squared 0.907
  - F-statistic 60.35
  - Durbin-Waston d-statistic 2.434
  - Note: *This dummy takes the value of the number of Chinese New Year holidays in a month.

### Empirical results — South Korea semiconductor exports
- Main findings:
  - Coefficients on dummies up to one month ahead of the release date and one month after are significant, indicating production ramps up just before release.
  - Trend terms (time and time^2) are both positive though not statistically significant, suggesting demand for South Korean semiconductors continued to rise during 2011-17.
  - Chinese New Year dummy is negative and statistically significant.
  - Regression explains about two thirds of variation in South Korea’s semiconductor exports.
  - Contextual evidence: Samsung Electronics overtook Intel Corporation as the largest semiconductor producer in the world in the second quarter of 2017 (Figure 9: Intel and Samsung sales in Billions of USD).
- Regression results (Dependent Variable: Volume of South Korea Semiconductor Exports (in log)):
  - M-3 0.11220; Std. Error 0.060; t 1.86; P>|t| 0.07
  - M-2 0.03441; Std. Error 0.060; t 0.57; P>|t| 0.57
  - M-1 0.16054; Std. Error 0.602; t 2.67; P>|t| 0.01
  - M 0.16968; Std. Error 0.060; t 2.82; P>|t| 0.01
  - M+1 0.12721; Std. Error 0.060; t 2.12; P>|t| 0.04
  - M+2 0.06735; Std. Error 0.060; t 1.12; P>|t| 0.27
  - M+3 -0.03859; Std. Error 0.055; t -0.70; P>|t| 0.49
  - M+4 -0.13387; Std. Error 0.055; t -2.45; P>|t| 0.02
  - Constant 40.34612; Std. Error 19.815; t 2.04; P>|t| 0.62
  - Time -0.07255; Std. Error 0.060; t -1.20; P>|t| 0.24
  - Time2 0.00006; Std. Error 0.000; t 1.29; P>|t| 0.20
  - Chinese New Year dummy* -0.01446; Std. Error 0.004; t -3.41; P>|t| 0.00
  - R-squared 0.673
  - Adjusted R-squared 0.606
  - F-statistic 10.09
  - Durbin-Waston d-statistic 1.038
  - Note: *This dummy takes the value of the number of Chinese New Year holidays in a month.

### Empirical results — Taiwan Province of China electronic export orders
- Main findings:
  - The release month and subsequent three months are statistically significant in explaining the cycle.
  - Results indicate Taiwan Province of China electronic export orders are contemporaneous with the China smartphone export cycle, implying short lead-times for component production.
  - Time coefficients (time and time^2) are borderline statistically significant and suggest export orders may be in the process of peaking; this may reflect limited data for the second half of 2017.
  - Data caveat: Taiwan data are in value terms because volume data are not available.
- Regression results (Dependent Variable: Value of Taiwan Province of China Electronic Export Orders (in log)):
  - M-3 -0.01319; Std. Error 0.025; T -0.52; P>|t| 0.61
  - M-2 0.01082; Std. Error 0.025; T 0.42; P>|t| 0.67
  - M-1 0.01387; Std. Error 0.027; T 0.50; P>|t| 0.62
  - M 0.11080; Std. Error 0.027; T 4.03; P>|t| 0.00
  - M+1 0.12798; Std. Error 0.027; T 4.66; P>|t| 0.00
  - M+2 0.13684; Std. Error 0.027; T 4.98; P>|t| 0.00
  - M+3 0.09003; Std. Error 0.025; T 3.59; P>|t| 0.00
  - M+4 0.02103; Std. Error 0.025; T 0.84; P>|t| 0.40
  - Constant -6.52886; Std. Error 8.403; T -0.78; P>|t| 0.44
  - Month 0.04467; Std. Error 0.026; T 1.75; P>|t| 0.09
  - Month2 0.00003; Std. Error 0.000; T -1.63; P>|t| 0.11
  - Chinese New Year dummy* -0.01427; Std. Error 0.014; T -7.36; P>|t| 0.00
  - R-squared 0.809
  - Adjusted R-squared 0.771
  - F-statistic 21.54
  - Durbin-Waston d-statistic 1.158
  - Note: *This dummy takes the value of the number of Chinese New Year holidays in a month.

### Concluding observations and key implications
- The new tech cycle has a large cyclical impact on export and growth performance across several Asian countries; peak effects reached up to 150 percent growth for Chinese smartphone exports in cycle spikes (as illustrated in Figure 2).
- Evidence suggests global smartphone shipments may have peaked in late 2015, indicating possible market saturation, while demand for semiconductors (notably from South Korea) continued to rise through 2017.
- Pre- and post-release cycles overlap and intertwine, amplifying dynamics across Asian supply chains.
- The release date of iPhone products has a significant impact on related electronics exports in China, South Korea, and Taiwan Province of China.
- Decomposition of trend and cycle suggests:
  - Global smartphone market may have saturated in late 2015.
  - Demand for other electronics products continues to boost semiconductor production, particularly in South Korea.
- Concluding assessment: Asia continues to gain in importance as the global tech manufacturer.

*Source: wp1822 (2017) — authors’ regression results and analyses.*

### 1. Apple Inc. and Samsung Electronics Smartphone Release Dates .......................................18

### 1. Apple Inc. and Samsung Electronics Smartphone Release Dates .......................................18

### Introduction and importance of the new tech cycle
- The new tech cycle is defined as the production and export of electronics and software associated with the release of new smartphones by global producers.
- Since 2012, global sales shifted from PCs to smartphones; in 2016 global smartphone sales reached close to 1.5 billion units, equivalent to one for every fifth person on earth.
- At peak (2016), exports of electronic components for smartphone production accounted for:
  - more than one third of total exports in Taiwan Province of China,
  - 15 percent in Singapore and South Korea,
  - 11 percent in Malaysia.
- In 2016, China exported USD107 billion worth of smartphones to the rest of the world, equivalent to 5 percent of its total exports.
- OECD data cited: CEO sectors accounted for 7.4 percent of total value added in South Korea in 2013; 2.1 percent in Japan and 2.0 percent in Ireland in 2012.

### Structure of the tech cycle: pre-release and post-release components
- The cycle is subdivided into:
  - Pre-release cycle: exports of components from several Asian countries to China for final assembly.
  - Post-release cycle: shipments of finished smartphones from China to the rest of the world.
- The cycle is driven by smartphone release dates (Apple and Samsung), not by standard seasonal patterns.

### Pre-release cycle: composition, geography, and magnitudes
- Pre-release shipments include:
  - memory chips (DRAM and flash memory) mainly from South Korea and Vietnam;
  - system chips (CPUs and app processors) and multichip packages (MCPs) from Japan, Malaysia, Singapore, South Korea and Taiwan Province of China;
  - displays from Japan and South Korea.
- Operating system sources:
  - Google Android: open source from the US (free);
  - Apple's operating system: from Ireland (Apple Inc.’s intellectual property location per cited source).
- Geography (2016): main contributors were Japan, Malaysia, Singapore, South Korea, and Taiwan Province of China; newcomers include the Philippines, Thailand, and Vietnam.
- Aggregate magnitudes:
  - The pre-release cycle accounted for USD68.2 billion of imports of electronics components to China in 2016, equivalent to about 4.3 percent of total Chinese merchandise imports.
  - Estimated value added from smartphone manufacturing and shipments in mainland China: USD20 billion, equivalent to 0.2 percent of 2016 GDP.

### Post-release cycle: destinations and shares (2016)
- Chinese exports of smartphones in 2016: USD106.9 billion (5.1 percent of total exports).
- Main destinations and values/shares:
  - US & Canada: USD29.9 Billion, 28.0%
  - Europe: USD24.3 Billion, 22.7%
  - South East Asia & India: USD16.4 Billion, 15.4%
  - South Korea: USD8.0 Billion, 7.5%
  - Japan: USD7.6 Billion, 7.1%
  - Latin America: USD6.3 Billion, 5.9%
  - MENA: USD7.1 Billion, 6.6%
  - Russia & CIS: USD3.0 Billion, 2.8%
  - Sub-Sahara: USD2.2 Billion, 2.1%
  - Australia and New Zealand: USD1.9 Billion, 1.8%
- Post-release shipments originate from China to regional distribution centers and then to consumers; shipments to South Korea reflect distribution of a portion of Samsung smartphones.

### Why standard seasonal adjustment fails to capture the tech cycle
- Reasons given:
  - Recent nature of the cycle (about six years) limits statistical capture by standard seasonal adjustment needing longer time series.
  - Rapidly evolving and increasingly complex supply chains across countries.
  - Cycle critically depends on smartphone release dates which vary within the calendar year and differ in impact depending on product success.

### Decomposing trend from cycle: methodology
- Monthly dummies before and after release dates capture the cycle.
- Time and time^2 capture a potentially non-linear trend.
- A Chinese New Year dummy (equal to the number of vacation days in each month) captures the roughly 15-day production shutdown that shifts about 11 days each year in the Gregorian calendar.
- The methodology is applied to:
  - China smartphone exports (units),
  - South Korea semiconductor exports (units),
  - Taiwan Province of China electronic export orders (value).

### Empirical results — China smartphone exports
- Main findings:
  - The cycle is mainly determined by iPhone release dates; Samsung release dates were not statistically significant.
  - Coefficients on one month before the release date and up to four months after are statistically significant.
  - Trend (time and time^2) is non-linear with a peak in September 2015.
  - Chinese New Year dummy coefficient is negative and highly significant.
  - Decomposition (cycle + trend + controls) explains about 92 percent of Chinese smartphone exports.
- Caution noted: lack of data for the second half of 2017 may overstate recent downturn in the trend.

### Empirical results — South Korea semiconductor exports
- Main findings:
  - Coefficients on dummies up to one month ahead of the release date and one month after are significant, indicating production ramps up just before release.
  - Trend terms (time and time^2) are both positive though not statistically significant, suggesting demand for South Korean semiconductors continued to rise during 2011-17.
  - Chinese New Year dummy is negative and statistically significant.
  - Regression explains about two thirds of variation in South Korea’s semiconductor exports.
- Contextual evidence: Samsung Electronics overtook Intel Corporation as the largest semiconductor producer in the world in the second quarter of 2017 (Figure 9: Intel and Samsung sales in Billions of USD).

### Empirical results — Taiwan Province of China electronic export orders
- Main findings:
  - The release month and subsequent three months are statistically significant in explaining the cycle.
  - Results indicate Taiwan Province of China electronic export orders are contemporaneous with the China smartphone export cycle, implying short lead-times for component production.
  - Time coefficients (time and time^2) are borderline statistically significant and suggest export orders may be in the process of peaking; this may reflect limited data for the second half of 2017.
- Data caveat: Taiwan data are in value terms because volume data are not available.

### Concluding observations from the chapter
- The new tech cycle has a large cyclical impact on export and growth performance across several Asian countries; peak effects reached up to 150 percent growth for Chinese smartphone exports in cycle spikes (as illustrated in Figure 2).
- Evidence suggests global smartphone shipments may have peaked in late 2015, indicating possible market saturation, while demand for semiconductors (notably from South Korea) continued to rise through 2017.
- Pre- and post-release cycles overlap and intertwine, amplifying dynamics across Asian supply chains.

*Source: wp1822 - 1. Apple Inc. and Samsung Electronics Smartphone Release Dates .......................................18*

### 2017. Finally, the dummy on the Chinese New Year effect is negative and statistically

### wp1822 - 2017. Finally, the dummy on the Chinese New Year effect is negative and statistically

### New tech cycle: dynamics and drivers
- The new tech cycle has materially affected export and growth performance of several Asian countries over the last six years through complex, evolving supply chains for smartphones.
- iPhone release dates have driven the tech cycle in the last five years across:
  - Chinese smartphone exports,
  - South Korean semiconductor exports,
  - Taiwan Province of China electronic export orders.
- Decompositions indicate:
  - Global demand for smartphones may have peaked in late 2015.
  - Global demand for South Korean semiconductors appears on an upward trend, implying other electronic products (e.g., automobile computers, smart appliances, wearable devices) are driving acceleration.
  - Underlying trend in Taiwan Province of China electronic export orders remains strong, albeit at a more moderate rate in the first half of 2017.

### Regression evidence: China smartphone exports (Table 2)
- Dependent Variable: Volume of China Smartphone Exports (in log)
- Months before/after phone release coefficients:
  - M-3: 0.02238; Std. Error 0.029; t 0.77; P>|t| 0.45
  - M-2: 0.03710; Std. Error 0.029; t 1.27; P>|t| 0.21
  - M-1: 0.05686; Std. Error 0.032; t 1.80; P>|t| 0.01
  - M 0.15978; Std. Error 0.032; t 5.10; P>|t| 0.00
  - M+1 0.17295; Std. Error 0.032; t 5.50; P>|t| 0.00
  - M+2 0.29625; Std. Error 0.031; t 9.41; P>|t| 0.00
  - M+3 0.24075; Std. Error 0.029; t 8.36; P>|t| 0.00
  - M+4 0.14691; Std. Error 0.029; t 5.16; P>|t| 0.00
- Other variables:
  - Constant -70.04554; Std. Error 9.640; t -7.27; P>|t| 0.00
  - Time 0.26577; Std. Error 0.029; t 9.06; P>|t| 0.00
  - Time2 -0.0002; Std. Error -0.000; t -8.94; P>|t| 0.00
  - Chinese New Year dummy* -0.03106; Std. Error 0.002; t -13.96; P>|t| 0.00
- Statistics:
  - R-squared 0.922
  - Adjusted R-squared 0.907
  - F-statistic 60.35
  - Durbin-Waston d-statistic 2.434
- Note: *This dummy takes the value of the number of Chinese New Year holidays in a month.

### Regression evidence: South Korea semiconductor exports (Table 3)
- Dependent Variable: Volume of South Korea Semiconductor Exports (in log)
- Months before/after phone release coefficients:
  - M-3 0.11220; Std. Error 0.060; t 1.86; P>|t| 0.07
  - M-2 0.03441; Std. Error 0.060; t 0.57; P>|t| 0.57
  - M-1 0.16054; Std. Error 0.602; t 2.67; P>|t| 0.01
  - M 0.16968; Std. Error 0.060; t 2.82; P>|t| 0.01
  - M+1 0.12721; Std. Error 0.060; t 2.12; P>|t| 0.04
  - M+2 0.06735; Std. Error 0.060; t 1.12; P>|t| 0.27
  - M+3 -0.03859; Std. Error 0.055; t -0.70; P>|t| 0.49
  - M+4 -0.13387; Std. Error 0.055; t -2.45; P>|t| 0.02
- Other variables:
  - Constant 40.34612; Std. Error 19.815; t 2.04; P>|t| 0.62
  - Time -0.07255; Std. Error 0.060; t -1.20; P>|t| 0.24
  - Time2 0.00006; Std. Error 0.000; t 1.29; P>|t| 0.20
  - Chinese New Year dummy* -0.01446; Std. Error 0.004; t -3.41; P>|t| 0.00
- Statistics:
  - R-squared 0.673
  - Adjusted R-squared 0.606
  - F-statistic 10.09
  - Durbin-Waston d-statistic 1.038
- Note: *This dummy takes the value of the number of Chinese New Year holidays in a month.

### Regression evidence: Taiwan Province of China electronic export orders (Table 4)
- Dependent Variable: Value of Taiwan Province of China Electronic Export Orders (in log)
- Months before/after phone release coefficients:
  - M-3 -0.01319; Std. Error 0.025; T -0.52; P>|t| 0.61
  - M-2 0.01082; Std. Error 0.025; T 0.42; P>|t| 0.67
  - M-1 0.01387; Std. Error 0.027; T 0.50; P>|t| 0.62
  - M 0.11080; Std. Error 0.027; T 4.03; P>|t| 0.00
  - M+1 0.12798; Std. Error 0.027; T 4.66; P>|t| 0.00
  - M+2 0.13684; Std. Error 0.027; T 4.98; P>|t| 0.00
  - M+3 0.09003; Std. Error 0.025; T 3.59; P>|t| 0.00
  - M+4 0.02103; Std. Error 0.025; T 0.84; P>|t| 0.40
- Other variables:
  - Constant -6.52886; Std. Error 8.403; T -0.78; P>|t| 0.44
  - Month 0.04467; Std. Error 0.026; T 1.75; P>|t| 0.09
  - Month2 0.00003; Std. Error 0.000; T -1.63; P>|t| 0.11
  - Chinese New Year dummy* -0.01427; Std. Error 0.014; T -7.36; P>|t| 0.00
- Statistics:
  - R-squared 0.809
  - Adjusted R-squared 0.771
  - F-statistic 21.54
  - Durbin-Waston d-statistic 1.158
- Note: *This dummy takes the value of the number of Chinese New Year holidays in a month.

### Key implications and conclusions (Section IV)
- The release date of iPhone products has a significant impact on related electronics exports in China, South Korea, and Taiwan Province of China.
- Decomposition of trend and cycle suggests:
  - Global smartphone market may have saturated in late 2015.
  - Demand for other electronics products continues to boost semiconductor production, particularly in South Korea.
- Concluding assessment: Asia continues to gain in importance as the global tech manufacturer.

*Source: wp1822 (2017) — authors’ regression results and analyses.*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2018/wp1822.pdf_