Macroeconomic Forecasting in South Korea

1.

Macroeconomic
Forecasting in South
Korea
ALISHER ABU, ECFUC-21

2.

Exploring South Korea's
Economic Forecasting
1.
Research Focus:
Topic: South Korea's Economic Growth.
Context: An in-depth analysis of how
key indicators shape the nation's
economic trajectory.
2.
Key Economic Indicators:
GDP Growth: Trends and predictions in
South Korea's GDP.
Inflation Rate: Analyzing the impact of
inflation on the economy.
Unemployment Rate: Understanding its
correlation with economic health.

3.

South Korea's Economic
Transformation
1.
Post-War Recovery (1950s-60s):
Rebuilding from the Korean War devastation.
Early focus on agriculture and light industries.
2.
Industrialization Boom (1970s-80s):
Government-led push towards heavy industries: steel, shipbuilding,
and electronics.
Rapid urbanization and workforce expansion.
3.
Tech-Led Growth (1990s-Present):
Emergence as a global tech hub, led by brands like Samsung and
LG.
Focus on high-tech industries, IT services, and innovation.
4.
Recent Challenges:
Adapting to a digital, service-based economy.
Aging population and workforce sustainability.
Balancing export-led growth with domestic consumption.

4.

Future Outlook of South
Korea's Economy
South Korea's economic future presents a complex picture,
influenced by both internal and external factors. Researchers
anticipate moderate growth recovery, contingent on structural
reforms and global economic conditions. The nation's average
GDP growth could maintain around 3% through 2030, but without
these reforms, growth could fall to as low as 1%. This forecast
highlights the pivotal position of South Korea's economy at a
global level.
The impact of the COVID-19 pandemic has been significant,
leading to structural changes across various sectors. The
manufacturing sector demonstrated resilience, while the services
sector experienced a more gradual recovery. The pandemic's
long-term effects are expected to suppress potential output,
dropping growth below 2%, with a medium-term recovery
anticipated around 2¼%.

5.

Future Outlook of
South Korea's
Economy
Economic policies, including South
Korea's universal transfer program,
have influenced the economy.
While supporting household
consumption and small businesses,
the exclusion of online and large
businesses raises concerns about
long-term economic efficiency
and fiscal sustainability.
Recent trends show a slight
acceleration in economic growth
in 2022, influenced by net trade
dynamics. However, challenges
such as a rapid decline in exports,
particularly to China, and
reduced government spending,
signal potential short-term
economic challenges.

6.

Future Outlook of South
Korea's Economy
The transformation of the e-commerce landscape, led by
corporations like Coupang, Gmarket, and WeMakePrice, is
reshaping South Korea's economy. This rapid development in
cross-border e-commerce technologies emphasizes
environmental considerations and has led to shifts in the labor
market towards digital professions. Research by academics like
Cheong, Yoo, and Park et al. provides insights into the
macroeconomic implications of e-commerce on factors like
employment and wages.
Looking ahead, GDP growth is expected to ease to 1.4% in
2023, with a rebound to 2.3% in 2024 and 2.1% in 2025. Factors
such as elevated interest rates, energy prices, and a slowdown
in global demand, particularly from China, are impacting
private consumption and exports. The aging population is
projected to increase fiscal pressures, necessitating reforms in
fiscal policy and strategies to address productivity gaps, labor
market dualism, and environmental concerns

7.

Forecasting
Methodologies in
Macroeconomic Analysis
1.
2.
Mean Forecast Method:
Utilizes the average of historical data for future
predictions, especially effective for data showing
a consistent pattern without significant
fluctuations.
Involves statistical functions (like meanf in R) to
generate forecasts, considering parameters such
as forecast periods and confidence levels.
Naïve Method:
A straightforward approach that assumes future
values will mirror the last observed value.
Despite its simplicity, it can show decent out-ofsample predictive accuracy but may not fully
capture data nuances, as indicated by high
Mean Absolute Percentage Error (MAPE) in certain
cases.

8.

Forecasting
Methodologies in
Macroeconomic Analysis
1.
2.
ARIMA (AutoRegressive Integrated Moving Average):
ARIMA is a widely used method that captures
various aspects of time series data like trends,
cycles, and seasonality.
Suitable for more complex forecasting where data
patterns are not straightforward.
Exponential Smoothing Method:
Incorporates Winters Seasonal Method, adding a
seasonal component to Holt's Linear Trend
method.
Ideal for data with trends and seasonality, these
models are simple, easy to use, and require fewer
data inputs, often used for short-term economic
forecasting.

9.

Key Economic
Indicators in South
Korea
GDP Growth:
Gross Domestic Product (GDP) growth
measures the increase in value of all goods
and services produced by an economy
over a period, reflecting economic
performance and health.
Measures the increase in the value of all
goods and services produced over a
specified period.
Expressed as a percentage, indicating the
economy's growth relative to a previous
period.
South Korea's GDP growth was primarily
influenced by net trade dynamics, growing
by 0.6% in Q2 2022, with fluctuations in
exports and imports impacting the overall
economic outlook.

10.

Key Economic
Indicators in South
Korea
Inflation Rate:
The inflation rate quantifies the
increase in prices of goods and
services
The coefficient for the Inflation
consumer price (annual%) is
approximately -16.92 billion US
dollars, suggesting that an
increase in inflation is associated
with a decrease in GDP.

11.

Key Economic
Indicators in South
Korea
Unemployment Rate:
Definition: This rate represents the
percentage of the labor force
that is unemployed and actively
seeking work, indicating labor
market health.
In South Korea, the labor market
has shown resilience with
historically high employment rates
and low unemployment.

12.

Key Economic
Indicators in South
Korea
Gross National Expenditure (GNE):
Definition: GNE represents the
total spending on final goods and
services by a nation's residents
and businesses, reflecting
domestic economic activity.
In South Korea, GNE is closely
linked to GDP growth, with
increments in GNE leading to
proportional increases in GDP.

13.

Analyzing GDP
Growth with the
ARIMA(0,1,1)
Model
Model Overview:
ARIMA(0,1,1): A model that captures the
random walk characteristic of South Korea's
GDP growth rates, balancing simplicity with
accuracy.
Residual Analysis:
The model shows no clear patterns or trends
in residuals, indicating effective capture of
the underlying data process.
Autocorrelation of Residuals:
The absence of significant autocorrelation in
residuals, as evidenced by the ACF plot,
confirms a comprehensive model fit.
Ljung-Box Test Results:
A p-value of 0.1356, suggesting no significant
autocorrelation and an acceptable fit,
indicating the model's errors are random
and not systematic.

14.

Analyzing GDP
Growth with the
ARIMA(0,1,1) Model
Forecast Accuracy Metrics:
Training Set: RMSE: 2.278164, MAE: 1.543108,
MAPE: 52.43103, MASE: 0.9414537
Test Set: RMSE: 1.947709, MAE: 1.466258, MAPE:
145.70134, MASE: 0.8945672
Lower RMSE and MAE values suggest better
predictive accuracy.
Predictive Power:
Despite high MAPE values, the model's lower
MASE values, particularly in the test set,
demonstrate its superiority over a naive
benchmark.
In conclusion, the ARIMA(0,1,1) model's balance of
simplicity, fit to historical data, and accuracy in
forecasting unseen data makes it the optimal choice
for analyzing South Korea's GDP growth.

15.

Analyzing Inflation Rate
with the ARIMA Model
Model Suitability:
The ARIMA model shows no significant autocorrelation in the
residuals, as indicated by the Ljung-Box test results. A high pvalue of 0.934 confirms that the model's errors are random,
enhancing confidence in its predictive capabilities.
Forecast Accuracy Metrics:
Training Set: RMSE: 0.9944768, MAE: 0.7543551, MAPE: 34.15738,
MASE: 0.9413792
Test Set: RMSE: 2.0104454, MAE: 1.6664832, MAPE: 142.98489,
MASE: 2.0796473
These metrics reveal the model's performance in historical data
(training set) and its predictive accuracy in unseen data (test
set).
Considerations:
The high MAPE value in the test set suggests that the model,
while effective in certain aspects, may have limitations in
capturing the full complexity of inflation rate changes,
especially in predicting future values.

16.

Analyzing Unemployment
Rate with the
ARIMA(2,0,0) Model
Model Overview:
The Auto ARIMA model, specifically ARIMA(2,0,0),
includes two autoregressive terms without
differencing or moving average components.
It effectively captures historical data patterns in
unemployment rates.
Forecast and Residuals Analysis:
The model forecasts future unemployment rates,
remaining stable around a mean value of 3.2528%.
Residuals show no significant autocorrelation, as
indicated by the Autocorrelation Function (ACF) plot
and a histogram of residuals distribution.
Ljung-Box Test Results:
Q* statistic: 1.3376, p-value: 0.7202.
The high p-value indicates an absence of significant
autocorrelation in the residuals, confirming a good
model fit.

17.

Analyzing
Unemployment Rate with
the ARIMA(2,0,0) Model
Accuracy Metrics:
Training Set: RMSE: 0.1710822, MAE:
0.1551080, MAPE: 4.796085, MASE: 0.8190521
Test Set: RMSE: 0.3799773, MAE: 0.3347572,
MAPE: 9.701779, MASE: 1.7676947
These metrics indicate a good fit with
historical data, though errors increase when
forecasting future values.
Conclusion:
The ARIMA(2,0,0) model robustly forecasts
the unemployment rate in South Korea, with
adequate capturing of historical patterns.
However, prediction intervals suggest
uncertainty in future rates, and forecasts
should be interpreted with caution.

18.

Analyzing GNE with
the ARIMA Model
Model Overview:
The ARIMA model is used to predict future values in
the GNE time series by analyzing trends and random
'shocks' in the data.
It includes an integrated part for differencing, and a
moving average component.
Residuals Analysis and Ljung-Box Test:
The residuals' density plot approximates a normal
distribution, indicating well-behaved errors.
Ljung-Box test results:
Q* value: 5.1593
P-value: 0.1605
These suggest no significant autocorrelation within
the residuals, implying a good fit of the model to the
data.

19.

Analyzing GNE with
the ARIMA Model
Accuracy Metrics:
Training Set:
Test Set:
RMSE: 63,101,012,229
MAE: 56,535,515,452
MAPE: 3.401125%
MASE: 0.5547571
These metrics indicate the model's good fit with
historical data and its effectiveness in predicting
future values.
Conclusion:
The ARIMA model's ability to effectively capture
the historical pattern in GNE data makes it a
suitable choice for forecasting.
The model's performance indicates reliability in
capturing GNE trends, but as with any forecast,
there is inherent uncertainty in predictions.

20.

Output and Future
Economy Forecast for
South Korea
GDP Forecast:
The ARIMA model projects an upward trend, reflecting historical economic
growth.
While the forecast shows stability, it suggests that future GDP growth may
experience slight fluctuations rather than volatility.
GNE Forecast:
Predicts an upward trajectory, aligning with an expanding economy.
Indicates potential increases in investment and consumption, as seen in the
rising national expenditure component.
Unemployment Rate Forecast:
Shows a general downward trend, signaling improving employment
conditions.
Forecast uncertainty implies potential impacts from factors not included in
the model.
Inflation Rate Forecast:
Projects a relatively stable inflation rate with increasing uncertainty over
time.
Implies that while inflation may stay within a certain range, long-term
predictions carry more uncertainty.

21.

Output and Future
Economy Forecast for
South Korea
Conclusions on South Korea's Economy:
Growth Trajectory: Likely continuation of growth, as
indicated by GDP and GNE forecasts.
Employment Conditions: Positive trends in the labor market,
though long-term forecasts carry uncertainties.
Price Stability: Inflation expected to remain stable in the
short term, but long-term forecasts suggest potential
external impacts.
Overall Outlook:
Short-term economic outlook appears positive, with growth,
improved employment, and stable inflation.
Longer-term forecasts show increasing uncertainty,
emphasizing the importance of external factors and global
economic changes.

22.

Regression Model and
Additional Forecasting for
South Korea's Economy
Model Overview:
The regression model used for additional
forecasting analyzes South Korea's GDP in current
US dollars.
It includes two predictor variables: Inflation
consumer price (annual%) and Gross National
Expenditure (GNE, current US$).
Statistical Significance and Fit:
The model demonstrates high explanatory power,
as indicated by R-squared values close to 1.
The F-statistic is 3212, with a highly significant pvalue (< 2.2e-16), affirming the model's statistical
significance.
Ljung-Box test on residuals: Q* statistic of 6.6741, pvalue of 0.08304, suggesting no significant
autocorrelation and confirming a good fit.

23.

Regression Model and Additional
Forecasting for South Korea's Economy
Model Coefficients and Impact:
The coefficients obtained are statistically significant.
GNE has a substantial positive impact on GDP, indicating that increases in
national expenditure significantly boost GDP.
In contrast, inflation has a negative impact, but its effect is comparatively
less pronounced than GNE.
Conclusion:
The regression model effectively captures the relationship between GDP
and its predictors (inflation and GNE), providing valuable insights for
forecasting.
Its adequacy in explaining GDP variation is evident from the lack of
autocorrelation in residuals and high explanatory power.
This model serves as a supplementary tool to time-series models, offering a
more nuanced understanding of economic dynamics.

24.

Challenges and
Limitations in Economic
Forecasting
Complex Economic Interplay:
South Korea's technology-driven economy presents a complex
web of factors influencing GDP growth, making accurate
forecasting challenging.
The intricate relationship between various economic sectors and
global influences adds layers of complexity to predictive
modeling.
Data Limitations:
Forecasting is often constrained by the availability and quality of
data.
Historical data may not fully capture future economic dynamics,
especially in a rapidly evolving economy like South Korea's.
Model Selection Challenges:
Choosing the most suitable model involves balancing simplicity
with predictive power.
Each model, whether ARIMA or regression, has inherent strengths
and weaknesses, affecting their forecasting accuracy.

25.

Challenges and
Limitations in Economic
Forecasting
Vulnerability to External Shocks:
Economic forecasts are susceptible to unforeseen
global events and external shocks, which can
significantly deviate actual outcomes from
predictions.
This includes technological innovations, geopolitical
changes, or global economic trends.
Rapid Technological Change:
The fast pace of technological advancement in
South Korea can rapidly alter economic conditions,
making it difficult for models to keep pace with
changes.
Conclusion:
While the forecasting models provide valuable
insights into South Korea's economic trajectory, they
are not without limitations.
Stakeholders, from policymakers to businesses, must
interpret these forecasts within the context of these
challenges and limitations.

26.

Conclusion: South Korea's
Economic Outlook
South Korea's economy is forecasted to continue growing, with GDP and GNE
trends indicating an expansion and increased national expenditure. The GDP
growth is expected to be stable, though with some fluctuations, suggesting
controlled volatility in the near future. Employment conditions look promising, as the
unemployment rate is projected to decrease, signaling an improving labor market.
However, there's an undercurrent of uncertainty in these forecasts, especially in the
long term.
Inflation rates are expected to remain relatively stable, but with increasing
uncertainty over time, indicating potential fluctuations in the future. This stability in
the short term is a positive sign, but the widening confidence intervals in the longterm forecasts suggest that external factors could impact this trend.
Overall, while the short-term outlook for South Korea's economy is positive, with
growth, better employment conditions, and stable inflation, the increasing
uncertainty in the longer-term forecasts indicates that predictions become less
reliable as they extend into the future. It's important to note that these forecasts are
based on past trends and do not incorporate external information or sudden
changes in economic conditions. Therefore, while the data currently suggests
stability and growth, external shocks or significant global economic changes could
impact these forecasts.