MAIN MACROECONOMIC INDICATORS

Macroeconomic indicators are statistics that indicate the current status of the economy of a state depending on a particular area of the economy (industry, labor
market, trade, etc.). They are published regularly at a certain time by governmental agencies and the private sector.

Markets.com provides an Economic Calendar for the dates of critical fundamental announcements and events. When properly used, these indicators can be an
invaluable resource for any Forex trader.

In truth, these statistics help Forex traders monitor the economy's pulse; thus it is not surprising that these are religiously followed by almost everyone in the financial
markets. After publication of these indicators we can observe volatility of the market. The degree of volatility is determined depending on the importance of an
indicator. That is why it is important to understand which indicator is important and what it represents.

Interest Rates Announcement

Interest rates play the most important role in moving the prices of currencies in the foreign exchange market. As the institutions that set interest rates, central banks are
therefore the most influential actors. Interest rates dictate flows of investment. Since currencies are the representations of a countrys economy, differences in interest
rates affect the relative worth of currencies in relation to one another. When central banks change interest rates they cause the forex market to experience movement
and volatility. In the realm of Forex trading, accurate speculation of central banks actions can enhance the trader's chances for a successful trade.

Gross Domestic Product (GDP)

The GDP is the broadest measure of a country's economy, and it represents the total market value of all goods and services produced in a country during a given year.
Since the GDP figure itself is often considered a lagging indicator, most traders focus on the two reports that are issued in the months before the final GDP figures: the
advance report and the preliminary report. Significant revisions between these reports can cause considerable volatility.

Consumer Price Index

The Consumer Price Index (CPI) is probably the most crucial indicator of inflation. It represents changes in the level of retail prices for the basic consumer basket.
Inflation is tied directly to the purchasing power of a currency within its borders and affects its standing on the international markets. If the economy develops in
normal conditions, the increase in CPI can lead to an increase in basic interest rates. This, in turn, leads to an increase in the attractiveness of a currency.

Employment Indicators
Employment indicators reflect the overall health of an economy or business cycle. In order to understand how an economy is functioning, it is important to know how
many jobs are being created or destructed, what percentage of the work force is actively working, and how many new people are claiming unemployment. For
inflation measurement, it is also important to monitor the speed at which wages are growing.

Retail Sales

The retail sales indicator is released on a monthly basis and is important to the foreign exchange trader because it shows the overall strength of consumer spending and
the success of retail stores. The report is particularly useful because it is a timely indicator of broad consumer spending patterns that is adjusted for seasonal variables.
It can be used to predict the performance of more important lagging indicators, and to assess the immediate direction of an economy.

Balance of Payments
The Balance of Payments represents the ratio between the amount of payments received from abroad and the amount of payments going abroad. In other words, it
shows the total foreign trade operations, trade balance, and balance between export and import, transfer payments. If coming payment exceeds payments to other
countries and international organizations the balance of payments is positive. The surplus is a favorable factor for growth of the national currency.

Government Fiscal and Monetary policy

Stabilization of the economy (e.g., full employment, control of inflation, and an equitable balance of payments) is one of the goals that governments attempt to
achieve through manipulation of fiscal and monetary policies. Fiscal policy relates to taxes and expenditures, monetary policy to financial markets and the supply of
credit, money, and other financial assets.

Conclusion: There are many economic indicators, and even more private reports that can be used to evaluate the fundamentals of forex. It's important to take the time
to not only look at the numbers, but also understand what they mean and how they affect a nation's economy.

MEASURING MACROECONOMIC INSTABILITY: A CRITICAL SURVEY

ILLUSTRATED WITH EXPORTS SERIES
Authors

Jol Cariolle,

1.

Michal Goujon

1.

First published: 9 July 2013Full publication history

DOI: 10.1111/joes.12036View/save citation
Cited by: 0 articles Check for new citations

Abstract
For at least 40 years, the analysis of the causes and consequences of macroeconomic instability has greatly deepened our
understanding of the handicaps faced by developing countries. This concern on economic instability is evidenced by a broad
spectrum of indicators, based on the deviation of observed values of a given economic aggregate from its reference or trend
value. In general, the choice of this or that indicator is not discussed advocating that the resulting instability indicators are closely
correlated. Focusing on measurements of instability in export revenue data for 134 countries from 1970 to 2005, this paper finds
that this assertion may be true for variance-based indicators, measuring the average magnitude of deviations from the trend.
However, great discrepancies may arise between different measures of the asymmetry or of the occurrence of extreme deviations
around the trend when different trend computation methods are used. Our purpose is, therefore, to invite further discussions
regarding the use of these indicators, and to highlight the different dimensions of instability, which have been so far unheeded by
the economic literature.

Provide feedback or get help

1. Introduction
Global economic crises in the 20th century have made macroeconomic
instability a key issue in the analyses of economic growth and
development. Most empirical researches conclude that instability has a
negative impact on growth (see Hnatkovska and Loayza, 2005; Koren
and Tenreyro, 2007; Loayza et al., 2007. Indeed, instability reduces
consumption (Aizenman and Pinto, 2005; Wolf, 2005), investment and
factor productivity (Azeinman and Marion, 1999; Dehn, 2000), and
deteriorates the quality of economic policies (Fatas and Mihov, 2007;
Afonso and Furceri, 2010). Instability may, however, positively affect
the return on investment, and growth, but only in good institutional
context (Hnatkovska and Loayza, 2005; Imbs, 2007; Rancire et
al., 2008).
Furthermore, if good institutions reduce instability
(Acemoglu et al., 2003; Mobarak, 2005; Fatas and Mihov, 2006) or allow
a better absorption of shocks (Rodrik, 2000), financial development has
in contrast an ambiguous role in transmitting or attenuating instability
(Beck et al., 2001; Aghion et al., 2004; Aghion et al., 2005). These channels
explain why developing countries are more vulnerable to
macroeconomic instability, by being more exposed to shocks and less
able to absorb them (Loayza and Raddatz, 2007; Guillaumont, 2009a,b;
Malik and Temple, 2009; di Giovanni and Levchenko, 2010).
Instability is a complex and multidimensional phenomenon as witnessed
by the large array of methods by which it is measured. Economic
instability refers to the notion of economic disequilibrium, which is also
used for instances in the analyses of output gap and exchange rate
misalignment (see Egert et al., 2006): its measurement is then generally
based on the extent to which observed values of an economic variable

deviate transitorily from the trend or reference value. Therefore,

measuring instability requires in a first stage to look carefully at the data
and to choose the appropriate method of calculating the trend around
which a series fluctuates. Measures of instability that are discussed in
this paper must be distinguished from measures of uncertainty or risk
aimed at reflecting unpredictable variations only (Wolf, 2005), and
which are based on conditional variance GARCH (Servn, 1998;
Dehn, 2000; Dehn et al.,2005; Chua et al., 2011).
The second stage consists in summing deviations from this trend.
Traditionally, indicators of instability are confined to the average
amplitude of deviations, such as the standard deviation. However, this
masks other important dimensions of instability, such as the asymmetry
of deviations (predominance of positive or negative shocks) or the
occurrence of extreme deviations, which are expected to have specific
consequences (Alderman, 1996; Dercon, 2002; Rancire et al., 2008).
Usually there is little discussion about alternative methods for measuring
instability, which are then indiscriminately applied to economic series
with different patterns of evolution (see annex 1). Amongst available
measures of instability, variance-based indicators are the most common
ones. However, confining the analysis of instability to the analysis of a
variable's variance may mask other important dimensions of economic
instability. In fact, other important dimensions of economic instability
are the asymmetry of fluctuations and the likelihood of crisis or booms,
and are addressed in this paper.
The most common measure of instability is the standard deviation of the
growth rate of a variable (see annex 1), which assumes, sometimes without
any prior testing, that the variable is stationary in first difference. Other
measures consist in calculating the standard deviation of the residuals of a

of the variable on a deterministic and/or stochastic trend

(Servn, 1998; Pritchett, 2000; Lensink and Morrissey, 2006; Chauvet
and Guillaumont, 2009). Alternatively, the reference value can be
computed as a moving average (Dawe, 1996), aHodrick-Prescott filter (Becker and
Mauro, 2006; Chauvet and Guillaumont, 2009), or aBaxter-King
filter (Hnatkovska and Loayza, 2005; Afonso and Furceri, 2010).
regression

The aim of this paper is therefore to discuss and apply popular

techniques for the measurement of instability to export revenue data (in
constant value), for a sample of 134 countries over the period 1970
2005. We focus on exports instability since it is a widely debated source
of output fluctuations, with strong destabilizing effects on growth, tax,
and redistribution policy (Bevan et al., 1993; Easterly et al., 1993;
Guillaumont, 2009a, b). First, we present and compare suitable
parametric and non-parametric univariate approaches for calculating the
trend. Then, on the basis of these trends, we compute three indicators of
instability for the magnitude, the asymmetry and the occurrence of
extreme deviations, respectively. Results show that (1) indicators of the
magnitude of export instability are strongly correlated with each other;
(2) correlations between indicators of asymmetry, and between
indicators of occurrence of extreme deviations, are low; (3) whatever the
trend calculation method, the average magnitude of instability is found
to be unrelated to its asymmetry and to the occurrence of extreme
deviations; while (4) the two latter are strongly associated.
The next section discusses parametric and non-parametric approaches
for calculating the trend component of export series. Based on these
trend computation methods, the third section outlines the various ways
of summing deviations from the trend, presents and compares indicators
reflecting the average magnitude of economic fluctuations, their
asymmetry and the occurrence of extreme variations.

2. Trend Computation Methods

Measures of instability typically rely on the extent to which observed
values of time series deviate transitorily from their permanent (or longrun) state. However, when modelling the permanent state by the simple
average of the observed values (a constant), most economic aggregates
show deviations that are not temporary, that is, series are not stationary.
As a consequence, it is then necessary to specify a more complex
permanent state, like a trend, around which fluctuations are stationary.1
A preliminary diagnostic of the pattern of change in export series is very
informative when trying to deduce an appropriate form of the trend. A
useful tool is the spectrum density of a series which represents the
contributions of each frequency of variation to the total variation of the
series. Because low-medium (medium-high) frequency variations are
likely to reflect the permanent (transitory) component, observing series
spectrum would help to point what should be an appropriate form of the
trend.
Figure 1 displays the evolution of exports in six countries, and their
associated spectrum densities. A peak at a given frequency (or
periodicity2) indicates that a significant proportion of the total variance
in a series can be explained by the variations at this frequency.

Figure 1.

Open in figure viewer

Download Powerpoint slide

Export Series and Spectrum Densities for South Korea, Argentina, Venezuela,
Kenya, Ivory Coast and Burundi.

The Figure 1 suggests that export series spectrum is located at long

periodicities of around 20 years (with a peak in density at a frequency of
around 0.05), except for South Korea.3 Variations of 5 to 10-year
periodicity (a frequency between 0.1 and 0.2) also represent an
important part of total variability in exports for this sample of countries.
The spectrum of export series points to the existence of density peaks at
high-frequency variations (between 0.3 and 0.5), corresponding to
periodicities of around 23 years, in all countries except South Korea.
Thus, while export series are dominated by low-frequency variations,
short or medium-run export movements are also visible for this sample
of developing countries.
In what follows, we compute four popular trends that differ in their
ability to separate low-frequency (permanent) from high-frequency
(transitory) variations in export revenue data. Two trends are computed
following a standard parametric univariate approach, the two others with
a univariate filter approach.

2.1 Parametric Approaches4

Univariate parametric approaches are commonly used to model
deterministic and/or stochastic trends (Harvey, 1997; White and
Granger, 2011).
2.1.1 Export Fluctuations around a Linear Deterministic
Trend
The traditional approach consists in using a deterministic trend,
assuming that, if not tested, the variable is trend-stationary. Like many

macroeconomic variables, export series are dominated by low-frequency

variations which can be modelled by a deterministic trend (linear here,
while a polynomial or exponential trend would be other options). This
takes the form:
(1)

where y is the observations of the economic variable, a constant, t a

linear trend, and a zero mean error term. The estimated trend or
reference value is then
and the deviations are
t

With having no permanent effects on y . This implies that (1) trend

values increase at a constant rate, (2) the long-term change of the series
is perfectly predictable and (3) all deviations from the trend are
transitory.
t

Beveridge and Nelson (1981) and Harvey (1997) highlighted the

limitations of this approach. To illustrate this, Figure 2 shows the actual
change in exports and their trend in Belize between 1980 and 2004, and
in Argentina between 1970 and 2005 (residuals are reported in Figure 3).
Although the trend in Belize seems linear, some deviations from the
trend are somewhat longstanding. This is more obvious for Argentina,
where the trend is clearly nonlinear. Assuming a linear trend would then
lead to overstate instability as residuals may wrongly include a part of
the non-constant trend.

Figure 2.

Open in figure viewer

Download Powerpoint slide

Linear Trends of Export Series in Belize and Argentina.

Figure 3.

Open in figure viewer

Download Powerpoint slide

Residuals of Linear Trends in Belize and Argentina.

2.1.2 Export Fluctuations around a Global Mixed Trend

Export variables may then fluctuate around a trend that varies over time.
In this case, a reference value may include a stochastic trend,
represented by the following first-order autoregressive AR(1) process
(Harvey, 1997),
(2)

which can be rewritten as:

Changes in y are determined by a successive history of random shocks: a
shock occurring in the distant past has the same effect than a current
shock. The trend component then follows a random walk process. A time
series with such a variation pattern exhibits a unit root process and is
said difference-stationary:
t

(3)

This hypothesis underpinned indicators of instability based on the

standard deviation in the series growth rate, a (too) strong assumption
given the often observed low-frequency variations in macroeconomic
variables like GDP, exports or public spending (Nelson and Kang, 1981).
White and Granger (2011) reconcile the two approaches by stressing that
trends commonly observed in macroeconomic variables display both
stochastic and deterministic movements. This takes the form:
(4)

Using 134 country exports series, we perform the Maddala-Wu panel

unit root test on (p-value are reported in Table 1). We also compute
Fisher statistics for the joint null hypothesis on, and (Table 1, second
column). Maddala-Wu unit root test suggests that the series contain a
unit root, and F-test statistics reject the joint null hypothesis on
coefficients. Modelling export series by equation (4) hence seems an
t

appropriate approach, even if the above tests have low power in cases of
near unit-root, structural breaks, or limited time-length (Sarris, 2000).
Table 1. Specification and Unit Root Test on Panel Data

H : The Series Is Non-Stationary

0

1.

Prob > Chi

F-test

Countries (Observations): 134(3693).

1.000

47.47

0.000

36.61

As illustrated in Figure 4 and annex 2 this method produces a trend that

is a slightly smoothed version of the observed change in real export
values. Correlograms of residuals for Belize and Argentina in
Figure 4 suggest that deviations from the trend are transitory, since
residuals are not significantly autocorrelated (for other country
examples, see annex 2). However, this trend seems to reproduce in t the
change in exports observed in t 1, which is not consistent with the
conclusion of White and Granger (2011), who stress that one desirable
property of trend values is to be pointedly smoother than the gross
series. We address this issue in the next sub-section.

Figure 4.

Open in figure viewer

Download Powerpoint slide

Export Revenue, Global Mixed Trend and the Correlogram of Residuals in Belize
and Argentina.

2.1.3 Estimate Based on a Rolling Mixed Trend

The above mixed trend is global, in the sense that estimated
coefficients of equation (4) are assumed to be constant over the whole
period of data availability. It therefore excludes the possibility for
regime changes in the deterministic and stochastic components of the
trend. Although tests of structural breaks in time series do exist (for

example, CUSUM, Max Chow tests), an alternative and more practical

solution with panel dataset may consist in estimating a mixed trend on a
rolling basis (Guillaumont, 2007), allowing the estimated coefficients
to change from year to year, thus reflecting recent changes in the trend
regime. This rolling mixed trend is estimated each year for each
country over the period T = [t k; t], rather than the whole period of data
availability:
(5)

Figure 5 plots the estimated rolling trend when k = 12 against the

global mixed trend in Argentina (see annex 2 for more country results).
The rolling mixed trend is visibly smoother and does not exhibit a
saw-tooth profile that results from time-constant parameters in the case
of global trend. Similarly, correlograms in Figures 5 and 6, and in
annex 2, suggest that the resulting residuals are stationary, for this
sample of countries.

Figure 5.

Open in figure viewer

Download Powerpoint slide

Export Revenue, Global and Rolling Mixed Trends, and the Correlogram of
Residuals in Argentina.

Figure 6.

Open in figure viewer

Download Powerpoint slide

Export Revenue, Global and Rolling Mixed Trends, and Correlograms of Residuals
in Burundi.

Nonetheless, this technique presents some drawbacks. First, it reduces

the time coverage of instability indicators, since the first trend value is
only available from t = 1 + k. Second, estimates based on a rolling mixed
trend do not fully account for a structural break in the trend, since the
prior trend regime may still exhibit inertia after the break. Moreover, by
estimating the trend over a shorter period than the global mixed trend, it
is more sensitive to medium-periodicity than to long-periodicity
fluctuations, with the risk of including the latter in the residuals. This
point is illustrated by the behaviour of the global and rolling mixed
trends in exports from Burundi between 1985 and 1995, in Figure 6. In
this example, the rolling trend tends to underestimate the trend downfall
after 1985 relative to the global mixed trend.
Choosing an appropriate timeframe for the rolling trend is then a critical
step because of its serious implications. Instead of a 12-years estimation
period chosen for the sake of methods comparison, a rolling trend

calculated over a period of 15 or 20 years would be a reasonable basis.

As suggested by spectrum densities in Figure 1, variations of such
periodicity strongly contribute to the overall variation of the series.
Thus, the estimation period is an arbitrary choice intending to reflect at
best the evolution of a series for a sample of countries. Such a choice
possibly induces misspecification problems when looking at particular
countries export series and trends.

2.2 The Filter Approach

Statistical filters are used to isolate the deviations as a cyclical
component by removing the trend components of series. The standard
deviation of the resulting cycle is then a popular measure of economic
instability (Hnatkovska and Loayza, 2005; Becker and Mauro, 2006;
Chauvet and Guillaumont, 2009). Unlike the parametric approach, the
filter approach does not require a priori assumptions on the form of the
trend and is sensitive to structural breaks. The Hodrick-Prescott (HP)
filter (Hodrick-Prescott, 1997) is amongst the most popular (another
example is the Band Pass filter of Baxter and King, 1999). The filter
breaks down the change in a series into a trend component ( y ), and a
cyclical component (y ):
P
t

C
t

(6)

The HP filter isolates the cyclical component by optimising the

following programme:
(7)

giving the deviation

. This method is close to a symmetrical
moving average filter with an infinite time horizon. is a smoothing
parameter which can be either estimated or determined ad hoc. The first
term of equation (7) minimises the variance in the cyclical component

whilst the second term smoothes the change in the trend component.
When tends to infinity, the variance in the growth of the trend
component converges to zero, which implies that the trend component
or the filtered series is close to a simple linear trend. Conversely, when
tends to zero, the filtered series is close to the original series.
Consequently, the lower the value of , the shorter the periodicity of
isolated cyclical fluctuations is. The choice of the value of is still
debated in the literature. While Hodrick and Prescott (1997) advocate a
parameter equal to 100 for annual data, Baxter and King (1999)
suggest a value up to 400, while Maravall and Del Rio (2001) choose a
value down to 6.
We, hereafter, compare results obtained with set at 100 (longperiodicity trend, or HP100) and 6.5 (medium-periodicity trend, or
HP65). As expected, Figure 7 shows that the HP65 trend fluctuates more
than the HP100 trend. Annex 3 reports correlograms of the isolated
cycles for these countries. These results suggest that extracted cycles are
stationary and not significantly autocorrelated, for this sample of
countries.

Figure 7.

Open in figure viewer

Download Powerpoint slide

Export Revenue Data Smoothed by the HP Filter.

The filtering approach nonetheless presents some drawbacks compared

to the other approaches. First, it relies on the hypothesis that the cyclical
and the trend components are independent, which seems rather
restrictive (see Ramey and Ramey, 1995 for the impact of
macroeconomic fluctuations on long-term growth). Moreover, the HP
filter may suffer fromcompression effects when the smoothing parameter is set at
a low value: part of short-periodicity cyclical variations can wrongly be
attributed to trend variations. The resulting cyclical component upon
which instability measures are based may be understated. Conversely,
choosing a high-smoothing parameter may cause leakage effects: some of the
long-periodicity variations may be included into the cyclical component,
and instability may be overstated. In our case, the contribution of
medium-periodicity fluctuations in the total variance, seen in Figure 1,
suggest the choice of a lower smoothing parameter (it would also be
relevant for samples of developing countries with shorter business cycles
and more variable trend components, see Rand and Tarp, 2002; Aguiar
and Gopinath, 2007).

2.3 Interim Discussion

Parametric and filter approaches of computing the trend are conceptually
different. While the former models the process followed by trends on the
basis of the past evolution of the data, two-sided filter methods use both
past and future data, logically affecting measurements of instability.
On the one hand, the filter approach is probably more efficient in
isolating temporary deviations by being more sensitive to regime
changes, and is more flexible when series exhibit very diverging patterns
between countries. On the other hand, because trend values are predicted
according to past data, instability indicators based on the parametric
approach better reflects the impact of unusual economic events such as

economic crisis or booms, which is a key feature for the study of the
consequences of instability on economic outcomes.
Conversely, the two-sided filter approach using both past and future data
makes the trend to be more sensitive to outliers and may hence produce
deviations that are less sensitive to extreme events. This approach was
originally used to identify business cycles with the underlying
assumption of economic agents form rational expectations. By contrast,
parametric approaches are popular to analyse the impacts of shocks on
economic agents decisions that follow extrapolative or adaptive
expectations.
As a final word, both approaches imply arbitrary choices, regarding the
estimation period of the trend in the parametric approach, and the value
of the smoothing parameter in the filter approach. These choices should
at best reflect broad trend movements in economic time series, but
inevitably induce misspecification problems when applied to wide panel
datasets (this problem is discussed in another context by Egert et
al., 2006).

3. Quantifying Instability: Ways

Summing Deviations from the Trend

of

Variance-based indicators are the most common measures of instability.

However, they mask other important dimensions of instability like the
asymmetry of deviations and the occurrence of extreme deviations,
which can be quantified by exploiting moments two, three and four of
the series. Annex 4 plots the comparative evolution of export deviations
from trends values in Argentina, Belize, Burundi, Ivory Coast, South
Korea and Venezuela.

3.1 Calculation Period

Because indicators of instability are intended to reflect a past history of
shocks, it is first necessary to set the period over which deviations are
summed.5 Wolf (2005) suggests aligning the period to the approximate
duration of the episodes of instability observed in the series, if any.

3.2 The Magnitude of Export Fluctuations

The standard deviation is the most common indicator to estimate the
average deviation from the reference (trend) value:
(8)

with T = 1, , t. is the observed value, and is the reference value.

We compute indicators of the average magnitude of export instability
over the period 19822005, based on the standard deviation, for each
reference values. Results are shown in Tables 2and 3, and Figure 8.
Standard deviations of export around mixed trends and HP-filtered
values display high correlations exceeding 0.8, with the HP100 filter
standard deviation being the less correlated with the others. Standard
deviations around the global mixed trend and the HP100 filter tend to be
higher than the others in average, while standard deviations around the
HP65 filtered values are lower in average than the others. Thus, the
choice of the reference value does not seem to have serious implications
for the measurement of the average magnitude of export instability.

Figure 8.

Open in figure viewer

Download Powerpoint slide

Graphical Illustrations of Correlations between Standard Deviations of Exports

around Different Trend Values.
Table 2. Correlations between Standard Deviations of Exports around Different Trend Values

(1)

Global
trend

1.

mixed

(2)

(3)

(4)

Rolling mixed trend

HP6.5

HP100

*Significant at 5%. Observations: 134. Standard deviations calculated over the period 19822005.

(1)

1.00

(2)

0.92*

1.00

(3)

0.96*

0.95*

1.00

(4)

0.87*

0.80*

0.87*

1.00

Table 3. Descriptive Statistics of Standard Deviations of Exports around Different Trend Values

Global
trend

1.

Mixed

Rolling
trend

Mixed

HP6.5

HP100

9.2

13.3

Observations: 134. Standard deviations calculated over the period 19822005

Mean

13.6

11.5

Global
trend

Standard deviation

8.7

Mixed

Rolling
trend

7.8

Mixed

HP6.5

HP100

6.1

7.8

3.3 The Asymmetry of Export Fluctuations

Variance-based indicators of instability do not reflect the asymmetry of
shocks, a very important dimension of instability given the expected
asymmetry in agents responses to adverse and favourable shocks
(Dercon, 2002; Wolf, 2005; Elbers et al., 2007). In this regard, the
coefficient of asymmetry in a series or skewness is an indicator of
the predominance of adverse or favourable shocks, which can be
calculated as follow (in% of the trend):
(9)

with T = 1, , t. A symmetrical distribution displays a coefficient of

skewness close to 0%, while a positive (negative) skewness indicates
that instability is dominated by positive (negative) shocks. Figure 9 is a
graphical illustration of a positively skewed, a centred, and a negatively
skewed distribution of export in Argentina, Algeria and Mexico,
respectively. We can observe that a positive (negative) skewness
increases with the size or the frequency of positive (negative) shocks.

Figure 9.

Open in figure viewer

Download Powerpoint slide

Kernel Densities of the Distribution of Exports around a 12-year Rolling Mixed

Trend and Its Corresponding Moments in Argentina, Algeria and Mexico.

Table 4 and annex 5 suggest that the coefficients of skewness display

important discrepancies when reference values differ, except between
HP-based measures. Consequently, the choice of the reference value is
likely to influence the generated picture of shocks.
Table 4. Correlations between Coefficients of Skewness Calculated over the Period 19822005

(1)

Global
Trend

1.

(2)

Mixed

Rolling
Trend

Mixed

*Significant at 10%. Sample = 134 countries. Skewness calculated over the period 19822005.

(3)

(4)

HP6.5

HP100

(1)

Global
Trend

(2)

Mixed

Rolling
Trend

Mixed

(3)

(4)

HP6.5

HP100

Skewness calculated over the period 19822005

(1)

(2)

0.23*

(3)

0.08*

0.14*

(4)

0.29*

0.02

0.65*

Moreover, for a similar magnitude of instability, asymmetry may

strongly differ. Figure 10illustrates the fairly weak correlation between
the standard deviations and the skewness of exports for each of the four
reference value, particularly once some outliers are excluded.

Figure 10.

Open in figure viewer

Download Powerpoint slide

Graphical Illustrations of Correlations between the Standard Deviation and the

Skewness of Exports, by Trend Value.

3.4 The Occurrence of Extreme Shocks in Exports

A third dimension of instability is the occurrence of extreme, but
infrequent, deviations in a series, measured by the fourth moment of the
distribution around the reference value, or the kurtosis, computed as
follows (as a percentage of the reference value):
(10)

The kurtosis is a measure of the relative peakedness or tails fatness of a

statistical distribution. A low value indicates that the distribution tends to

be uniformly distributed around the mean, while a high value indicates a

distribution pointed around the mean with thick tails. Figure 11illustrates
three types of flattening of distributions of exports around their trend
platykurtic (kurtosis <300%), mesokurtic (kurtosis = 300%), and
leptokurtic (kurtosis >300%).

Figure 11.

Open in figure viewer

Download Powerpoint slide

Graphical Illustrations of Platykurtic, Mesokurtic, and Leptokurtic Distributions of

Exports around a Rolling Mixed Trend.

As underlined by Rancire et al. (2008), a high value of kurtosis should be

interpreted with caution since it may result from either extreme
variations or a cluster of observations around the mean. Table 5 shows
descriptive statistics for kurtosis-based indicators of instability for
exports series. Distributions around reference values are slightly
leptokurtic on average (higher than 300%). Distributions around the HP
filter display a similar degree of flattening, which is lower than those
displayed by distributions around mixed trends.
Table 5. Descriptive Statistics of Coefficients of Kurtosis Calculated over 19822005

Global
Trend

1.

Mixed

Rolling
Trend

Mixed

HP6.5

HP100

Total sample: 134 countries. Kurtosis calculated over the period 19822005.

Mean (%)

352.7

367.9

320.0

312.2

Standard deviation

155.7

173.0

139.0

146.2

Table 6 and annex 6 report correlations between kurtoses based on the

four reference values. These correlations are slightly stronger than the
correlations between asymmetry coefficients, the calculated occurrence
of infrequent large-scale positive or negative deviations being logically
less influenced by the choice of the reference value. However,
correlations between rolling mixed trend-based and HP-based kurtoses
are weak, suggesting again that the different trend computation methods
generate different histories of shocks, even in the case of sharp and
infrequent shocks.
Table 6. Correlations between Coefficients of Kurtosis Calculated over 19822005

(1)

Global
Trend

1.

(2)

Mixed

Rolling
Trend

Mixed

(3)

(4)

HP6.5

HP100

*Significant at 5%. Sample = 134 countries. Kurtosis calculated over the period 19822005.

(1)

(2)

0.39*

(3)

0.38*

0.28*

(4)

0.49*

0.22*

0.62*

As suggested by Figure 12, whatever the reference values, the average

magnitude of deviations seems quite unrelated to the occurrence of
extreme ones. In Figure 13, a U-shaped relationship can be observed
between asymmetry and kurtosis, with a turning point located close to a
null skewness. Table 7 shows a strong positive correlation between these
two indicators when the sample is limited to countries with a positive
asymmetry, and a strong negative correlation for countries with a
negative asymmetry. Moreover, it appears that high values of kurtosis
are often associated with high positive values of skewness. For an
asymmetry between 0 and 100% in absolute value, these two dimensions

of instability are however less related. Therefore, a low asymmetry

would reflect deviations characterized by a smaller magnitude and a
higher frequency, while a high value would reflect large-magnitude lowfrequency deviations.

Figure 12.

Open in figure viewer

Download Powerpoint slide

Graphical Illustrations of the Correlation between the Standard Deviation and the
Kurtosis of Export Distributions, by Trend Value.

Figure 13.

Open in figure viewer

Download Powerpoint slide

Graphical Illustrations of the Correlation between the Skewness and the Kurtosis
of Export Distributions, by Trend Value.
Table 7. Correlations between the Skewness and the Kurtosis of Export Distributions, by Trend Value, 19822005

1.

Global Mixed

Rolling Mixed

Trend

Trend

*Significant at 5%. Sample = 134 countries. Instability indicators calculated over the period 19822005.

HP6.5

HP100

Global Mixed

Rolling Mixed

Trend

Trend

HP6.5

HP100

Total sample

+0.65*

+0.51*

+0.12*

+0.35*

Positive
asymmetry
skewness > 0%

+0.85*

+0.83*

+0.84*

+0.91*

Negative
asymmetry
skewness < 0%

0.58*

0.48*

0.70*

0.45*

Weak
skewness
100%]

+0.24*

+0.16*

+0.33*

+0.35*

asymmetry
[100%;

4. Conclusion
The literature on macroeconomic instability is extensive and uses a wide
range of instability indicators. Approaches of measuring instability vary
according to the choice of the reference value and the way deviations
around it are summed. Moreover, beyond the standard measure of the
average magnitude of economic instability, the asymmetry of
fluctuations and the occurrence of extreme shocks are separate
dimensions, largely overlooked by the applied literature. We have used
exports data on 134 countries to illustrate these points, following three
main steps: (1) the examination of export series patterns of evolution,
(2) the computation of the trend component of export series, and (3) the

computation of instability indicators based on the moments of the

distribution of exports around their trend.
We applied four popular trend calculation methods. The parametric
approach allows us selecting an appropriate form of the trend, a mixed
trend including a deterministic and a stochastic component, which we
declined in two versions: a global mixed trend and a rolling mixed
trend. In our view, when data coverage is long enough, a rolling trend
should be preferred, as estimated coefficients better account for
structural breaks. The filter approach is also declined in two versions,
with a low- or a high-smoothing parameter. This approach does not
require an a priori choice of the form of the trend, which is also more
sensitive to structural breaks than parametric approaches.
Each trend computation method is not exempt from misspecification
problems. The trend estimation period or smoothing parameter value
must be chosen in accordance with the typical trend pattern in the
country sample. As developing countries compose a significant part of
our sample, a rolling trend estimated over a 1520 year period or a
smoothing parameter set between 6 and 10 seemed appropriate initially,
since the trend component in these countries is less stable than in
developed countries.
Three types of indicators of exports instability are computed using the
four forms of the trend: the magnitude of fluctuations (measured by the
standard deviation from the trend), asymmetry (skewness) and the
occurrence of extreme shocks (kurtosis). First, results show that the
indicators of magnitude are strongly correlated. Second, in contrast, the
indicators of asymmetry are weakly correlated, as are the indicators of
the occurrence of extreme deviations. Third, whatever the form of the
trend, the average magnitude of instability seems unrelated to

asymmetry and to the occurrence of extreme deviations. And fourth,

inversely, asymmetry is strongly associated to the occurrence of extreme
events. Thus, while low absolute values of skewness reflects the
occurrence of asymmetric, frequent but small fluctuations, high absolute
values of skewness reflect the occurrence of asymmetric, infrequent but
large fluctuations. As stressed by Alderman (1996) and Dercon (2002)
besides asymmetry, the frequency and intensity of economic fluctuations
clearly shape economic agents responses to them. 6 Then, using
skewness-based measure along with variance-based of instability should
therefore allow researchers to go further in the study of the consequence
of economic instability on economic behaviours and economic outcomes
(Cariolle, 2012; Rancire et al., 2008).

Acknowledgements
Authors are grateful to Patrick Guillaumont, Jean-Louis Combes,
Laurent Wagner and an anonymous referee for helpful comments and
suggestions. This paper benefited from the support of the FERDI
(Fondation pour les tudes et recherches sur le dveloppement
international) and of the Programme d'investissements d'avenir of the
French government.

Notes
1.

1
If the series is poorly stationarized, variations which are attributable to a long-term (or permanent) change may be included in the residual.

2.

2
The calculation for switching from frequency (F) to periodicity (T) is as follows: T = 1/F.

3.

Like many western countries, South Korea's export series exhibits a Granger profile (Granger, 1966), with most of the power of the spectrum
located at zero frequency. This pattern suggests that a strictly decreasing or increasing trend in exports represent the principal source of total
export variance.
4.

4
Along this section, we estimate trend values using data in logarithm, and then rescale them using an exponential transformation.

5.

5
This period does not necessarily correspond to the period chosen for trend estimation.

6.

6
As Dercon (2002, p. 2) points out, other characteristics of income risk include the frequency and intensity of shocks, and the persistence of their
impact (). Relatively small but frequent shocks are more easily to deal than large, infrequent negative shocks.

Annexes
Table Annex 1. Overview of Indicators of Instability and Their Applications in the Literature

Indicators

Authors

to Reflect

Variables
Concerned
(y )
t

Growth rate/first difference

Standard deviation or coefficient of variation of a variable's growth rate

Servn
(1997),
Acemoglu
et
al.
(2003),
Mobarak
(2005),
Koren and
Tenreyro
(2007),
Raddatz
(2007), di
Giovanni
and
Levchenk
o (2010),
Malik and
Temple
(2009),
Ploeg and
Poelhekke
(2009)

Variability

GDP/inhabita
nt, inflation,
terms
of
trade, actual
exchange
rate, blackmarket
premium,
international
interest rate
(LIBOR),
development
aid/inhabitant
,
public
spending,
ratio of wheat
cultivation
yields
to
national yield

Indicators

Authors

to Reflect

Variables
Concerned
(y )
t

5-year variance of a variable's growth rate

Koren and
Tenreyro
(2007)

Variability

Annual
growth rate
of
work
productivity

Decline in GDP: decrease of more than 1% of the (annual) series log

smoothed by the HP filter (lambda = 1000)

Becker
and
Mauro
(2006)

Variability

GDP

Standard deviation of the cyclical component, i.e. the standard deviation of

the difference between series smoothed by the HP or BK filter and actual
series.

Hnatkovs
ka
and
Loayza
(2005),
Chauvet
and
Guillaum
ont
(2009),
Afonso
and
Furceri
(2010)

VariabilityVariabi
lity

Aid, export
revenues,
GDPGDP,
budget
variables
(transfers,
subsidies,
public
spending, tax
revenues,
etc.)

Average over five years of the ratio of the absolute deviation between the
observed value of export revenues (X) and the value filtered using the ratio of
export five-year moving average process over GDP (Y):

Dawe
(1996)

Variability

Export
revenues

Filters

Indicators

Authors

to Reflect

Variables
Concerned
(y )
t

Forecasts/Estimates

Ramey
and
Ramey
(1995)

Uncertainty

Growth
of GDP

rate

Pritchett
(2000),
Mobarak
(2005)

Variability

Growth rate
of GDP per
inhabitant,
growth rate
of capital per
worker

Rolling standard deviation or average absolute deviation of the

residual obtained based on a regression of y on a rolling mixed trend,

Guillaum
ont
(2007),
Servn
(1998)

Variability/
uncertainty

Development
aid, exports,
terms
of
trade,
inflation rate,
relative price
of
capital,
actual
exchange
rate, growth
rate of GDP.

Standard deviation of the error in a regression of FDI over three lags and a
temporal
trend:

Lensink
and
Morrissey
(2006)

Uncertainty

FDI/GDP,
FDI

The standard deviation of the residual , is seen as a measure of volatility

reflecting uncertainty. This approach is adopted for the whole of the sample
and for all countries, to produce an estimate of coefficients specific to each
country.

The standard deviation of the residual , obtained using a regression of y on a

t

linear trend:

Indicators

Authors

to Reflect

Variables
Concerned
(y )
t

They estimate volatility measures for each country based on the following
GARCH
(1,1)
model:
where t = 1,T, and D the vector of mute quarterly variables. By imposing
the
following
constraint
on
conditional
variance
The
represents the uncertainty of y .
it

estimated

value

of

Dehn
(2000),
Servn
(1998)

Uncertainty

Prices of raw
materials,
inflation rate,
price
of
capital
relative
to
actual
exchange rate

Annex 2.

Open in figure viewer

Download Powerpoint slide
Mixed Trends and Correlograms of Residuals.

Annex 3.

Open in figure viewer

Download Powerpoint slide

Correlograms of Export Revenue Cycles Smoothed by the HP Filter.

Annex 4.

Open in figure viewer

Download Powerpoint slide
Comparative Evolution of Deviations (Residuals or Cycles), in % of Trend Values.

Annex 5.

Open in figure viewer

Download Powerpoint slide
Graphical Illustration of Correlations between Coefficients of Skewness of Export Distributions Around Different Trend Values.

Annex 6.

Open in figure viewer

Download Powerpoint slide

Graphical Illustration of Correlations between Coefficients of Kurtosis of Export Distributions Around Different Trend Values.

Ancillary
References

o
o

o
o

o
o

Acemoglu, D., Johnson, S., Robinson, J. and Thaicharoen, Y. (2003) Institutional causes, macroeconomic symptoms: volatility, crises
and growth. Journal of Monetary Economics 50(1): 49123.
CrossRef |
Web of Science Times Cited: 233
Afonso, A. and Furceri, D. (2010) Government size, composition, volatility and economic growth. European Journal of Political
Economy 26(4): 517532.
CrossRef |
Web of Science Times Cited: 16
Aghion, P., Bacchetta, P. and Banerjee, A. (2004) Financial development and the instability of open economies Journal of Monetary
Economics 51: 10771106.
CrossRef |
Web of Science Times Cited: 65
Aghion, P., Banerjee, A., Angeletos, G. and Manova, K. (2005) Volatility and growth: credit constraints and productivity-enhancing
investment. NBER Working Paper Series No. 11349.


o
o

o
o

o
o

o
o

o
o

o
o

o
o

o
o

o
o

o
o

o
o

o
o

o
o

Aguiar, A. and Gopinath, G. (2007) Emerging market business cycles: the cycle is the trend. Journal of Political Economy 115: 69
102.
CrossRef |
Web of Science Times Cited: 143
Alderman, H. (1996) Saving and economic shocks in rural Pakistan. Journal of Development Economics 51(2):343366.
CrossRef |
Web of Science Times Cited: 30
Azeinman,
J. and Marion,
N. (1999) Volatility
and
investment:
interpreting
evidence
from
developing
countries. Economica 66(262): 157179.
Wiley Online Library |
Web of Science Times Cited: 76
Azeinman, J. and Pinto, B. (eds). (2005) Managing Economic Volatility and Crises: A Practitioner's Guide. NBER Working Paper No.
10602. Cambridge, MA: Cambridge University Press.
CrossRef
Baxter, M. and King, R.G. (1999) Measuring business cycles. Approximate band-pass filters for economic time-series. Review of
Economics and Statistics 81(4): 575593.
CrossRef |
Web of Science Times Cited: 477
Beck, T., Lundberg, M. and Majnoni, G. (2001) Financial intermediary development and economic volatility: does finance dampen or
magnify shocks?, World Bank Policy Research Working Paper No. 2707, The World Bank.
Becker, T. and Mauro, P. (2006) Output drops and the shocks that matter. IMF Working Paper 06/172, International Monetary Fund.
Bevan, D., Collier, P. and Gunning, J.W. (1993) Trade shocks in developing countries, consequences and policy responses. European
Economic Review 37(23): 557565.
CrossRef |
Web of Science Times Cited: 12
Beveridge, S. and Nelson, C.R. (1981) A new approach to decomposition of economic time series into permanent and transitory
components with particular attention to measurement of the business cycle.Journal of Monetary Economics 7: 151174.
CrossRef |
Web of Science Times Cited: 569
Cariolle, J. (2012) When corruption is a response to export instability. Working Paper P48, Foundation for Research and Studies on
International Development (FERDI), Clermont-Ferrand, France.
Chauvet, L. and Guillaumont, P. (2009) Aid, volatility, and growth again: when aid volatility matters and when it does not. Review of
Development Economics 13(3): 452463.
Wiley Online Library |
Web of Science Times Cited: 8
Chua, C.L., Kim, D. and Suardi, S. (2011) Are empirical measures of macroeconomic uncertainty alike?.Journal of Economic
Surveys 25: 801827.
Wiley Online Library |
Web of Science Times Cited: 3
Dawe, D. (1996) A new look at the effects of export volatility on investment and growth. World Development 24(12): 19051914.
CrossRef |
Web of Science Times Cited: 29
Dehn, J. (2000) Commodity price uncertainty in developing countries. World Bank Policy Research Working Paper No. 2426, The
World Bank.
Dehn, J., Gilbert, C.L. and Varangis, P. (2005) Agricultural commodity price volatility. In J. Azeinman and B. Pinto (eds), Managing
Economic Volatility and Crises (pp. 137185). Cambridge, MA: Cambridge University Press.
Dercon, S. (2002), Income risk. Coping strategies, and safety nets, Discussion Paper No. 2002/22, UNU/WIDER.
di Giovanni, J. and Levchenko, A. (2010) The risk content of exports: a portfolio view of international trade. NBER Working Paper
Series No. 16005.
Easterly, W., Kremer, M., Pritchett, L. and Summers, L.H. (1993) Good policy or good luck?: country growth performance and
temporary shocks. Journal of Monetary Economics 32(3): 459483.
CrossRef |
Web of Science Times Cited: 207
Egert, B., Halpern, L. and MacDonald, R. (2006) Equilibrium exchange rates in transition economies: taking stock of the
issues. Journal of Economic Surveys 20(2): 257324.
Wiley Online Library |
Web of Science Times Cited: 40
Elbers, C., Gunning, J.W. and Kinsey, B.H. (2007) Growth and risk: methodology and micro evidence. World Bank Economic
Review 21(1): 120.
CrossRef |
Web of Science Times Cited: 27
Fatas, A. and Mihov, I. (2006) The macroeconomic effects of fiscal rules in the US states. Journal of Public Economics 90(1-2): 101
117.
CrossRef |
Web of Science Times Cited: 45

o
o

o
o

o
o

o
o

o
o

o
o

o
o

o
o

o
o

o
o

o
o

o
o

o
o

Fatas, A. and Mihov, I. (2007) Fiscal discipline, volatility, and growth. In P.G. Servn and R. Suescun (eds),Fiscal Policy,
Stabilization, and Growth (pp. 4374). Washington, DC: Latin American Development Forum, The World Bank.
Granger, C.W.J. (1966) The typical spectral shape of an economic variable. Econometrica 34(1): 150161.
CrossRef |
Web of Science Times Cited: 221
Guillaumont, P. (2007) Economic vulnerability, persistent challenge to African development Revue Africaine du
Dveloppement 19(1): 123162.
Web of Science Times Cited: 1
Guillaumont, P. (2009a) Caught in a Trap: Identifying the Least Developed Countries. Paris: Economica.
Guillaumont, P. (2009b) An economic vulnerability index: its design and use for international development policy. Oxford
Development Studies 37(3): 193228.
CrossRef
Harvey, A. (1997) Trends, cycles, and autoregressions. The Economic Journal 107(440): 192201.
CrossRef |
Web of Science Times Cited: 56
Hnatkovska, V. and Loayza, N. (2005) Volatility and growth. In J. Azeinman and B. Pinto (eds), Managing Economic Volatility and
Crises. Cambridge, MA: Cambridge University Press.
Hodrick, R. and Prescott, E. (1997), Postwar US business cycles: an empirical investigation. Journal of Money, Credit and
Banking 29(1): 116.
CrossRef |
Web of Science Times Cited: 896
Imbs, J. (2007) Growth and volatility. Journal of Monetary Economics 54(7): 18481862.
CrossRef |
Web of Science Times Cited: 30
Koren, M. and Tenreyro, S. (2007) Volatility and development. Quarterly Journal of Economics 122(1):243287.
CrossRef |
Web of Science Times Cited: 79
Lensink, R. and Morrissey, O. (2006) Foreign direct investment: flows, volatility, and the impact on growth.Review of International
Economics 14(3): 478493.
Wiley Online Library
Loayza, N.V. and Raddatz, C. (2007) The structural determinants of external vulnerability. World Bank Economic Review 21(3): 359
387.
CrossRef |
Web of Science Times Cited: 21
Loayza, N.V., Rancire, R., Servn, L. and Ventura, J. (2007) Macroeconomic volatility and welfare in developing countries: an
introduction. World Bank Economic Review 21(3): 343357.
CrossRef |
Web of Science Times Cited: 41
Malik, A. and Temple, J. (2009) The geography of output volatility. Journal of Development Economics 90:163178.
CrossRef |
Web of Science Times Cited: 15
Maravall, A. and Del Rio, A. (2001) Time aggregation and the Hodrick-Prescott filter, Working Paper No. 108.
Mobarak, A. (2005) Democracy, volatility, and economic development. Review of Economics and Statistics87(2): 348361.
CrossRef |
Web of Science Times Cited: 44
Nelson, C. and Kang, H. (1981) Spurious periodicity in inappropriately detrended time series. Econometrica49(3): 741751.
CrossRef |
Web of Science Times Cited: 156
Pritchett, L. (2000) Understanding patterns of economic growth: searching for hills among plateaus, mountains, and plains. The World
Bank Economic Review 14(2): 221250.
CrossRef |
Web of Science Times Cited: 117
van der Ploeg, F. and Poelhekke, S. (2009) Volatility and the natural resource curse. Oxford Economic Papers 61(4): 727760.
CrossRef |
Web of Science Times Cited: 27
Raddatz, C. (2007) Are external shocks responsible for the volatility of output in low-income countries?.Journal of Development
Economics 84(1): 155187.
CrossRef |
Web of Science Times Cited: 48
Ramey, G. and Ramey, V.A. (1995) Cross-country evidence on the link between volatility and growth. The American Economic
Review 85(5): 11381151.
Web of Science Times Cited: 397
Rancire, R., Tornell, A. and Westermann, F. (2008) Systemic crises and growth. Quarterly Journal of Economics 123(1): 359406.
CrossRef |

o
o

o
o

o
o

Web of Science Times Cited: 46

Rand, J. and Tarp, F. (2002) Business cycles in developing countries: are they different?. World Development30(12): 20712088.
CrossRef |
Web of Science Times Cited: 31
Rodrik, D. (2000) Participatory politics, social cooperation, and economic stability. The American Economic Review 90(2): 140144.
CrossRef |
Web of Science Times Cited: 33
Sarris, A.H. (2000) Has world cereal market instability increased?. Food Policy 25(3): 337350.
CrossRef |
Web of Science Times Cited: 7
Servn, L. (1997) Uncertainty, volatility, and irreversible investment: theory, evidence, and lessons for Africa. Journal of African
Economies 6(3): 229268.
Web of Science Times Cited: 17
Servn, L. (1998) Macroeconomic uncertainty and private investment in developing countries: an empirical investigation. Policy
Research Working Paper, The World Bank.
White, H. and Granger, W.J. (2011) Consideration of trends in times series. Journal of Time Series Econometrics 3(1): 19411928.
CrossRef
Wolf, H. (2005) Volatility: definitions and consequences. In J. Azeinman and B. Pinto (eds), Managing Economic volatility and
Crises (pp. 4564). Cambridge, MA: Cambridge University Press.

Types of Inflation: 4 Different Types Plus More

Inflation is when the prices of goods and services increase. There are four main types of inflation, categorized by their
speed: creeping, walking, galloping, and hyperinflation. There are also many types of asset inflation and, of course,
wage inflation. Many experts consider demand-pull and cost-push to be types of inflation, but they are actually causes
of inflation, as is expansion of the money supply.

People know that next year's car model will probably cost more. (Photo: Bill Pugliano/Getty Images)

1. Creeping Inflation

Creeping or mild inflation is when prices rise 3% a year or less. According to the U.S. Federal Reserve, when prices
rise 2% or less, it's actually beneficial to economic growth. That's because this mild inflation sets expectations that
prices will continue to rise. As a result, it sparks increased demand as consumers decide to buy now before prices rise
in the future. By increasing demand, mild inflationdrives economic expansion.

Ads

Prescient

www.prescient.co.za

Preservation of Capital Positive returns

Shorts & Bermudas - Kaymu

www.kaymu.pk

Discover All Our Shorts & Bermudas. From Rs 589, Buy It Now !

Health care costs rise faster than 3% a year. (Photo: Jason Greenspan/Getty Images)

2. Walking Inflation

This type of strong, or pernicious, inflation is between 3-10% a year. It is harmful to the economy because it heats
upeconomic growth too fast. People start to buy more than they need, just to avoid tomorrow's much higher prices. This
drives demand even further, so that suppliers can't keep up. More important, neither can wages. As a result, common
goods and services are priced out of the reach of most people.

Galloping inflation occurred during WWII. (Photo: U.S. National Archives and Records Administration)

3. Galloping Inflation

When inflation rises to ten percent or greater, it wreaks absolute havoc on the economy. Money loses value so fast that
business and employee income can't keep up with costs and prices. Foreign investors avoid the country, depriving it of
needed capital. The economy becomes unstable, and government leaders lose credibility. Galloping inflation must be
prevented.

Ads

HKTDC Research Platform

hktdc.com

Hub for the HK Product Industries, Research Studies and Data Analysis!

QMS ISO 9001 IRCA Course

www.proftechleading.com

Become Quality System Lead Auditor 5 days course. 95% success rate

Hans and Greta playing with worthless marks during the Weimar Republic. Photo: Albert Harlingue/Getty Images

4. Hyperinflation

Hyperinflation is when the prices skyrocket more than 50% -- a month. It is fortunately very rare. In fact, most
examples of hyperinflation have occurred when the government printed money recklessly to pay for war. Examples of
hyperinflation include Germany in the 1920s, Zimbabwe in the 2000s, and during the American Civil War. More

Federal Reserve Chairman Paul Volcker ended stagflation (Photo: Win McNamee/Getty Images)

5. Stagflation

Stagflation is just like its name says: when economic growth is stagnant, but there still is price inflation. This seems
contradictory, if not impossible. Why would prices go up when there isn't enough demand to stoke economic growth? It
happened in the 1970s when the U.S. went off the gold standard. Once the dollar's value was no longer tied to gold, the
number of dollars in circulation skyrocketed. This increase in the money supply was one of the causes of inflation.
Stagflation didn't end until then-Federal Reserve Chairman Paul Volcker raised the Fed funds rate to the double-digits -and kept it there long enough to dispel expectations of further inflation. Because it was such an unusual situation, it
probably won't happen again.More
Ads

Start Free Download

www.readingfanatic.com

Begin w/ Reading Fanatic for Free Access to Unlimited eBooks!

Food prices are not included in the core inflation rate. Photo: Monashee Frantz/Getty Images

6. Core Inflation

The core inflation rate measures rising prices in everything except food and energy. That's because gas prices tend to
escalate every summer, usually driving up the price of food and often anything else that has large transportation costs.
TheFederal Reserve uses the core inflation rate to guide it in setting monetary policy. The Fed doesn't want to
adjust interest rates every time gas prices go up -- and you wouldn't want it to. More

Deflation in housing prices trapped many homeowners.(Photo: Peter Dazeley/Getty Images)

7. Deflation

Deflation is the opposite of inflation -- it's when prices fall. It's caused when an asset bubble bursts. That's what
happened in housing in 2006. Deflation in housing prices trapped those who bought their homes in 2005. In fact, the
Fed was worried about overall deflation during the recession. That's because deflation can turn a recession into a
depression. During the Great Depression of 1929, prices dropped 10% -- a year. Once deflation starts, it is harder to
stop than inflation. More

Most U.S. workers have not experienced wage inflation. (Photo: Getty Images)

8. Wage Inflation

Wage inflation is when workers' pay rises faster than the cost of living. This occurs when there is a shortage of workers,
when labor unions negotiate ever-higher wages, or when workers effectively control their own pay. A worker shortage
occurs whenever unemployment is below 4%. Labor unions negotiated higher pay for auto workers in the 90s. CEOs
effectively control their own pay by sitting on many corporate boards, especially their own. All of these situations
created wage inflation. Of course, everyone thinks their wage increases are justified. However, higher wages are one
element of cost-push inflation, and can cause prices of the company's goods and services to rise.

In 2005, there was an asset bubble in housing. (Photo: Justin Sullivan/Getty Images)

9. Asset Inflation

An asset bubble, or asset inflation, occurs in one asset class, such as housing, oil orgold. It is often overlooked by
the Federal Reserve and other inflation-watchers when the overall rate of inflation is low. However, as we saw in
the subprime mortgage crisis and subsequent global financial crisis, asset inflation can be very damaging if left
unchecked. More

Inflation in gas prices affect people dramatically. (Photo: Mark Renders/Getty Images)

10. Asset Inflation -- Gas

Gas prices rise each spring in anticipation of the summertime vacation driving season. In fact, you can expect gas
pricesto rise ten cents per gallon each spring. However, political uncertainty in the oil-exporting countries drove gas
prices higher in 2011 and 2012. Prices hit an all-time peak of $4.17 in July 2008, thanks to economic uncertainty. For
more on that, see Gas Prices in 2008.

What do oil prices have to do with gas prices? A lot. In fact, oil prices are responsible for 72% of gas prices. The rest is
distribution and taxes, which aren't as volatile as oil prices. For more, see How Do Crude Oil Prices Affect Gas
Prices? More

Oil price inflation affects many other asset classes. (Photo: David McNew/Getty Images).

11. Asset Inflation -- Oil

Crude oil prices hit an all-time high of $143.68 a barrel in July 2008. This was in spite of a decrease in
global demand and an increase in supply. Oil prices are determined by commodities traders, both speculators and

corporate traders hedging their risks. Traders will bid up oil prices if they think there are threats to supply, such as
unrest in the Middle East, or an uptick in demand, such as growth in China. More

Ads

PetroSync O&G Training

www.petrosync.com

Professional Petroleum Training by Leading Industry Experts!

Crude Oil Blending

www.jiskoot.com

Crude oil blenders/mixing systems for terminals, refinery & pipeline

Food price inflation can cause food riots. (Photo: Elly Lange/ Getty Images).

12. Asset Inflation -- Food

Food prices soared 6.8% in 2008, causing food riots in India and other emerging markets. They spiked again in 2011,
rising 4.8% and leading to the Arab Spring, according to many economists. Food riots caused by inflation in this
important asset class could continue to reoccur. More

Inflation in gold prices occurred in 2011.(Photo: David McNew/Getty Images)

 13. Asset Inflation -- Gold

An asset bubble occurred when gold prices hit the all-time high of $1,895 an ounce on September 5, 2011. Although
many investors might not call this inflation, it sure was. That's because prices rose without a corresponding shift in
gold's supply or demand. Instead, investors drove up gold prices as a safe haven. They were concerned about
the declining dollar, hyperinflation in U.S. goods and services, and uncertainty about global stability. What spooked
investors? In August, thejobs report showed absolutely zero new jobs gains. During the summer, the eurozone debt
crisis looked like it might not get resolved and there was stress about whether the U.S. would default on its debt. Gold
prices go up in response to uncertainty, whether it's tohedge against inflation or its exact opposite, the resurgence
of recession. Article updated May 6, 2013 More

Different Types of Inflation

Tejvan Pettinger December 4, 2013 inflation

Inflation means a sustained increase in the general price level. However, this increase in the
cost of living can be caused by different factors. The main two types of inflation are
1.

Demand pull inflation this occurs when the economy grows quickly and starts to
overheat Aggregate demand (AD) will be increasing faster than aggregate supply (LRAS).

2.

Cost push inflation this occurs when there is a rise in the price of raw materials, higher
taxes, e.t.c

1. Demand Pull Inflation

This occurs when AD increases at a faster rate than AS. Demand pull inflation will typically
occur when the economy is growing faster than the long run trend rate of growth. If demand
exceeds supply, firms will respond by pushing up prices.
Simple diagram showing demand pull inflation

The UK experienced demand pull inflation during the Lawson boom of the late 1980s. Fuelled by
rising house prices, high consumer confidence and tax cuts, the economy was growing by 5% a
year, but this caused supply bottlenecks and firms responded by increasing prices.

This graph shows inflation and economic growth in the UK during the 1980s. High growth in
1987, 1988 of 4-5% caused an increase in the inflation rate. It was only when the economy went
into recession in 1990 and 1991, that we saw a fall in the inflation rate.

2. Cost Push Inflation

This occurs when there is an increase in the cost of production for firms causing aggregate
supply to shift to the left. Cost push inflation could be caused by rising energy and commodity
prices. See:Cost Push inflation
Simple Diagram showing cost push inflation.

Example of Cost push inflation in the UK

In early 2008, the UK economy entered a deep recession(GDP fell 6%). However, at the same
time, we experienced a rise in inflation. This inflation was definitely not due to demand side
factors; it was due to cost push factors, such as rising oil prices, rising taxes and rising import
prices (as a result of depreciation in the Pound) By 2013, cost push factors had mostly
disappeared and inflation had fallen back to its target of 2%.
Sometimes cost push inflation is known as the wrong type of inflation because this inflation is
associated with falling living standards. It is hard for the Central Bank to deal with cost push
inflation because they face both inflation and falling output.
3. Wage Push Inflation
Rising wages tend to cause inflation. In effect this is a combination of demand pull and cost
push inflation. Rising wages increase cost for firms and so these are passed onto consumers in
the form of higher prices. Also rising wages give consumers greater disposable income and
therefore cause increased consumption and AD. In the 1970s, trades unions were powerful in
the UK. This helped cause rising nominal wages; this was a significant factor in causing
inflation.
4. Imported Inflation.
A depreciation in the exchange rate will make imports more expensive. Therefore, the prices will
increase solely due to this exchange rate effect. A depreciation will also make exports more
competitive so will increase demand.
5. Temporary Factors.
The inflation rate can also increase due to temporary factors such as increasing indirect taxes. If
you increase VAT rate from 17.5% to 20%, all goods which are VAT applicable will be 2.5% more
expensive. However, this price rise will only last a year. It is not a permanent effect.

Core Inflation
One measure of inflation, is known as core inflation This is the inflation rate that excludes
temporary volatile factors, such as energy and food prices. The graph below shows inflation in
the EU. The headline inflation rate (HICP) is more volatile rising to 4% in 2008, and then falling
to -0.5% in 2009. However, the core inflation (HCIP energy, food, alcohol and tobacco) is more

constant.

Example of Inflation in UK

This shows that energy prices were very volatile in this period, contributing to cost push inflation
in 2008.
Different measures of inflation
There are different measures of inflation. RPI includes mortgage interest payments. In 2009,
interest rates were cut, therefore, RPI measure of inflation became negative. CPI excludes the
effect of mortgage interest payments. The ONS now produce a statistic CPIH, which is CPI

owner occupier costs.