session 1: introduction - university of warwick in economics from trinity college dublin msc, mres...

Session 1: Introduction

Michael McMahon

University of Warwick

McMahon (University of Warwick) Introduction 1 / 22

To Cover

1 The Abouts2 Syllabus3 Assignments and Grading4 Some philosophical points about Economics


The Abouts

About you and about me

About the course in general

Resources


The AboutsAbout you and about me

Dr. Michael McMahon

[email protected]

Biographical information:

IrishBA in Economics from Trinity College DublinMSc, MRes and PhD in Economics from London School of EconomicsWorked as a professional economist for the Bank of England, ECB, andCentral Bank of IrelandTaught for Warwick, LSE, NYU and University of Chicago.

Your name, your area of work, your economics and statisticsbackground, and your chance to ask me any questions about me.


The AboutsAbout the course

1 Lectures (40%)2 Class Discussion of Relevant Literature and current situation (40%)3 Computer Based Work (20%)



Economist Article....the measure of your learning.



There are 3 main themes to be covered in the course:

1 Macroeconomic Theory;2 Money, banking and the �nancial system (including the current crisisand policy responses);

3 Macroeconomic indicators and an introduction to the use of statisticsin macroeconomics.


The AboutsAbout the timetable - Day 1

The Basic Real Business Cycle Model (Theme 1)

Introducing the RBC modelDiscussing the costs of business cycles and the argument forstabilisation policyMeasuring business cycles using data (computer based work)Understanding and interpreting macroeconomic indicators

The economics of sectoral diversi�cation (Theme 1)

Bene�ts of diversi�ed economiesHow di¤erent sectors behave over the business cycle and economicadjustment across sectors

A brief introduction to forecasting and factor models (Theme 3)

the eurocoin indicator of the CEPR



Evening / Overnight Problem: Full problem set details to beprovided but expected to be done in 30 - 90 minutes.

Money and Banking (Theme 2)

The basics of the �nancial system including the main instrumentsWhy the current crisis? Focus on the concept of leverage and balancesheet e¤ects.Fiscal policy response including a discussion of how �scal policyoperates in the economy.A brief discussion of monetary policy including models of monetarypolicyClass discussion of the liquidity trap, the zero nominal bound oninterest rates and �quantitative easing�



Evening / Overnight Problem: To be decided - likely to considerthe issues of oil and monetary unions (full problem set details to beprovided but expected to be done in 30 - 90 minutes).Basic approaches to the labour market (Theme 1)

Search and Matching Models of the Labour MarketE¤ects of labour market regulation

The Macroeconomic E¤ects of Oil (Theme 1)How oil a¤ects the economyAnalysing the e¤ects of oil in the US and EUThe e¤ects of oil in oil-exporting countries

The Economics of Customs Unions (Theme 1)The bene�ts of customs unionsA brief introduction to the new economic geography (Krugman andVenables)Discussion of the GCC


The AboutsResources

Lectures

Course pack

http://www2.warwick.ac.uk/fac/soc/economics/staff/faculty/mcmahon/adced_course

Macroeconomic data

The Economist magazine, The FT, WSJ, etc...




Grading

Students will be assessed on 5 criteria:

Understanding of the theoretical components of the course

Understanding of the practical components of the course

Ability to combine the theoretical and practical elements of the coursein answers/project;

Communication of ideas � these should be both concise and clear;

Attitude to fellow students and the lecturer.


GradingGrade expectations

Top Marks: The student has an excellent understanding of the materialcovered within the lectures of the course. Moreover, the student has shownan ability to think beyond the basic material and to relate the sessions toone another in a highly perceptive way. Moreover, this student understandsthe limitations of the analytical tools developed in the course. The studentpresents ideas, both written and orally, in a clear and concise manner. Inboth class participation and also other interactions with the student, (s)hedisplays an excellent attitude; receptive to comments and eager to learn.



Middle Grade: Good understanding with some lack of clarity on eithertheoretical or empirical issues. Presentation of ideas may be moreambiguous or less well structured. Nonetheless, the student is clearlymotivated and works well within the course.



Lower Grade: The student has only partial understanding of some areasof the course. This leads to greater ambiguities in his/her presentation ofideas.


Grading

Each participant will be individually assessed via the following:

1 Two "within course" problem sets (20 points)

these short problems will involve the use of data from large databasessuch as FRED database at the St. Louis Fed, IMF and World Bankdatabases, or ILO data.each problem will be submitted by individual participants and will beworth up to 10 points.

2 2-week individual research exercise (80 points)

Assignment provided at the end of the course


Basics Concepts You Must Be Aware OfGDP - Real versus Nominal Concepts

Nominal GDP in year t =N

∑i=1pit .qit (1)

Real GDP (year K prices) in year t =N

∑i=1piK .qit (2)

De�ning the price level as a de�ator

Base Z de�atort =nominal GDPt

(real GDP in year Z prices)t(3)

Calculating growth

Growth between t and t + 1 =

�Xt+1Xt

� 1�

(4)

� lnXt+1 � lnXtMcMahon (University of Warwick) Introduction 17 / 22

Philosophical Points about Economics

Economics is a way of thinking - not a set of laws and rules

This framework for thinking can then be applied to many topics

Freakonomics

Empirical work is very tricky

LSE (my alma mater!) - motto is "Rerum Cognoscere Causas"

The following material is just for fun, but the frame of mind that theexercise highlights is all important for economics.



drc

tan

bur

eth

ngr

mlw

maluga

mad

car

rwa

nigmoz

bur

togben

gam

zam

ken

sil

cha

gha

angsennepcam

lescdihaichi

ind

guizimbannic

bol

honpng

pak

jam

phisri

guy

idnperrom

ecu

colmorpar

domegycos

gua

thaels

panfijtur

syrbot

iratun

ven

jorbra

uru

mexrsa

trihunchl

argmlymau

bar

por

gre

nwzkorcyp

jap

spa

ice

unkfin

swe

isrost

canden

swi

fra

netausireita

belhks

sin

nor

uns

3035

4045

Ave

rage

age

of t

he w

orkf

orce

0 20000 40000 60000Per capita GDP


Questions?


END


Session 2: Business Cycle Theory

Michael McMahon


McMahon (University of Warwick) Business Cycles 1 / 53

To Cover

1 Measuring business cycles using data2 Introducing the business cycle theory3 Discussing the costs of business cycles and the argument forstabilisation policy

4 Understanding and interpreting macroeconomic indicators.


Business Cycles

De�nitionCycle is made of an expansion (boom) and a contraction (recession).During the expansion all good things (GDP, employment, productivity,. . . )tend to go up, or grow faster than �normal,� and bad things(unemployment) tend to fall. During the contraction good things go downand bad things go up.

To be contrasted with economic growth.


The Cycle?

trough

peak

trough

expansion contraction

Business cycle

trough

peak

trough

expansion contraction

Business cycle


Features of Business Cycles I

1 Business cycles are NOT regular: They do not behave anything likesine waves. This is important because it tells us that they are not dueto some regular underlying component.

2 Expansions tend to be long and mild. They are measured in years.The 1990�s expansion and the most recent expansion = the longestever (my PhD).

3 Recessions tend to be short but more dramatic. They are measured inmonths.

4 However, in recessions the decline per period is relatively large.


Features of Business Cycles II

Series St. Dev Rel. Dev Autocorrelation Corr. with GDPY 1.49 1.00 0.58 1.00C 1.59 1.06 0.73 0.68I 6.69 4.48 0.68 0.74G 1.50 1.00 0.63 0.15NX 0.90 0.60 0.51 �0.16Hours 1.60 1.07 0.70 0.58Employ 1.05 0.70 0.77 0.39

YE 1.43 0.95 0.34 0.40P 2.1 1.45 0.88 �0.57M1 3.11 2.05 0.74 0.33


Business Cycle Dating Game

ProblemAnswer the following:

1 How would you identify periods of recession in the US economy since1940?

2 What criteria would you use to de�ne the periods?3 Apply this criteria to the data on US GDP and US GDP growth (andother important variables) using Excel.


Should We Care About Business Cycles?


Should We Care About Business Cycles?

Picture suggests that business cycles are minor blips and not worth muchattention. However:

1 What would level of GDP be without business cycles?

Some average of what we observe now?Or by eliminating recessions something that is on average higher?

2 Although GDP doesn�t do �much� in a recession, this doesn�t meansome �rms and individuals are not hurt (e.g. bankruptcies are verycyclical)

3 Business cycles could a¤ect growth rates


Business Cycle v. Output GapsDe�nitionOutput gap - A measure of the di¤erence between actual activity andpotential activity.

Is the economy under-performing or over-performing?

The two concepts are related but conceptually distinct.

The economy could be in a statistical recession without thiscorresponding to an output gap - e.g. natural disaster.Likewise, the economy could be in a boom but still beunder-performing.Output gap may be more useful as a concept but it is much harder tomeasure - one way is the production function technique:

Yt = AtLαtK

1�αt


Growth Accounting Framework

Using the production function:

Yt = AtLαtK

1�αt

Combined with data on inputs and income shares...

we can decompose GDP growth into its driving forces:

Contributions to Aggregate GrowthY TFP K L

1970s 1.55 0.56 0.89 0.111980s 2.58 1.77 0.76 0.071990s 2.39 1.56 0.74 0.092000s 2.41 1.42 0.84 0.15


A Basic Business Cycle TheoryAS-AD Model

A simple diagrammatic exposition to examine the behaviour of outputand in�ation.

Contains 2 elements:1 Aggregate demand curve - captures the national income identity:

AD = C + I + G + X �M

2 Aggregate supply curve - captures the total quantity supplied at everyprice level.

The aim is to examine the interaction of these two relationships andto develop a framework to think about the causes of business cycles.


A Basic Business Cycle Theory IIAD Curve I

Aggregate demand curve - captures the national income identity:

AD = C + I + G + X �MShows the level of aggregate demand at every price level.Slopes down in Y-P space:

Real money balances e¤ect - As P#, the real value of money balancesheld " )purchasing power of consumers ".Prices and interest rates - CB cuts nominal interest rates as the pricelevel falls ) Lower nominal interest rates increases in consumerdemand and planned investment.International competitiveness - Lower UK prices makes the UK morecompetitive (for a given exchange rate) ) competitiveness " and X ".

Shifts in AD:the government may increase its own expenditure (�scal policy)the monetary authority may change interest ratesconsumer sentiment may change.


A Basic Business Cycle Theory IIIAD Curve II

Y

P

AD


A Basic Business Cycle Theory IVAS Curve I

Aggregate supply is the total quantity of goods and services produced(supplied) at every price level.It re�ects productivity capacity of the economy and production costs.The curve shows the level of aggregate supply at every price level.In Y-P space, the AS curve is :

vertical in the long-run. The economy operates at its potential level inthe long-run and prices will adjust to ensureslopes up in the short-run. As prices increase, �rms are willing toincrease the amount that is supplied. This may be achieved byemploying overtime sta¤ or increasing utilization of machines.

Shifts in AS curve:changes in size & quality of the labour force or capital stocktechnological progress and the impact of innovationchanges in unit wage costs (incl. taxes and subsidies)changes in in�ation expectations


A Basic Business Cycle Theory IVAS Curve II

Y

P

SRAS

equilibriumlevel ofincome

LRAS

Y*

SRASflex

SRASfixed


Price Adjustment is Key

Why don�t �rms adjust prices immediately - the Classical School (AScurve is vertical):

1 Most �rms seem to revise prices only occasionally (perhaps as rarelyas once a year). This means that some �rms will delay price revisionby as long as one year.

2 Firms are concerned with their market share: even those whose turn isto revise prices will do so slowly if they know many others are notrevising any time soon.

3 May be uncertain about nature of shock (is it only me or is iteveryone?) and reactions of competitors.

4 Wages are often �xed for long periods (annual contracts). Somarginal cost fairly stable. In competitive markets price cannot movetoo far from marginal cost.


Frequency of Price Reviews

More than 12 512 34 2 1 00

5

10

15

20

25

30

35

40


How React to Initial Changes in Demand

0 10 20 30 40 50 60 70

Overtime

More Workers

Increase Prices

Increase Capacity

Increase Subcontractors

Increase delivery lags

Other


Price Adjustment is Key

But eventually:

More and more �rms get chance to revise prices, meaning thataggregate prices creep up

Workers get a chance to renegotiate wages, meaning that costs startincreasing, and �rms need to increase prices

Also, start running out of unemployed workers, which also pusheswages and hence prices upwards


A Basic Business Cycle Theory VEquilibrium

Y

P

SRAS

ADequilibrium

level ofincome

LRAS

Y*


What causes business cycles?

Economy humming along its long-run growth path but constantly bu¤etedby:

Demand Shocks: changes in demand for good and services;

Supply Shocks: essentially changes in �rms�costs of production;

Both demand and supply shocks can generate �uctuations.


Demand Shock IShifts in the AD curve:

Consumption

Investment

Government spending

Exports

Changes in desired amounts of all these a¤ect the demand faced by �rms

1 Say initial shock pushes demand up;2 Firms�initial response is to leave prices unchanged, and accommodatethe surge in demand by producing more (GDP, investment,employment go up). This is the boom.

3 In a second stage �rms begin to increase prices and reduce output.GDP etc. return to their long-run trend.

And a recession is the same thing in reverse.McMahon (University of Warwick) Business Cycles 28 / 53

Demand Shock II

Y

P

SRAS

ADequilibrium

level ofincome

LRAS

Y*


Demand Shock II

Y

P

SRAS


LRAS

Y*

AD’

Y1


Demand Shock II

Y

P


LRAS

Y*

AD’

Y1

SRAS’’


Supply Shock IAccelerations or decelerations in technical changeChanges in commodity pricesChanges in regulationAny other changes to the cost structure

The cycle?

Positive supply shock: because of lower costs, �rms can producemore, hire more, and lower pricesNegative supply shock: because of higher costs, �rms reduceproduction, hire less, and increase prices

But:

Often observe recessions with no clear supply side �culprit�;Booms seem to be accompanied by increasing - not falling - prices;Technological �regress� is implausible.


Supply Shock II - Temporary

Y

P

SRAS

ADequilibrium

level ofincome

LRAS

Y*


Supply Shock II - Temporary

Y

P


LRAS

Y*

AD

Y1

SRAS’’


Supply Shock III - Permanent

Y

P

SRAS

ADequilibrium

level ofincome

LRAS

Y*


Supply Shock III - Permanent

Y

P


LRAS’’

Y*

AD

Y1

SRAS’’


Propagation of Shocks - I

Above we have thought about business cycles in terms of demand andsupply shocks. This is informative but does not answer following question:

How do shocks to the economy become transmitted over time?

Remember that business cycles have a duration of 4-10 years, so the factthat the �uctuations are persistent is a key feature.

To think about this, it is useful to consider a distinction between the initialshock to the economy �the impulse �and the mechanism that transmitsshocks over time � the propagation mechanism.


Propagation of Shocks - II

ImpulseImpulse PropagationMechanismPropagationMechanism

BusinessCycles

BusinessCycles

If there was no propagation mechanism, impulses would quickly dieout unless the impulses are repeated over time.

Propagation mechanisms are aspects that gives rise to persistente¤ects.


Propagation of Shocks - III

Propagation mechanisms:

investment and capital accumulation. e.g. credit markets;

intertemporal substitution in consumption and labor supply

nominal rigidities

Neoclassical economists stress the �rst two.

Keynesian economists stress the last.


Propagation of Shocks - IVImpulse Response Functions

Suppose that we are in equilibrium

Allow there to be a once-occurring shock (e.g. technology shock)

Impulse-response functions tell you how the variables of the modeladjust over time.


Propagation of Shocks - V

2 0 2 4 6 8 10 12 14 16 18100

100.5

101

101.5Response of TFP (% of SS)

2 0 2 4 6 8 10 12 14 16 1898

100

102

104Response of GDP (% of SS)

2 0 2 4 6 8 10 12 14 16 1899

100

101Response of Consumption (% of SS)

2 0 2 4 6 8 10 12 14 16 1890

100

110Response of Investment (% of SS)

2 0 2 4 6 8 10 12 14 16 1899

100

101Response of Capital (% of SS)

2 0 2 4 6 8 10 12 14 16 1898

100

102Response of Hours (% of SS)

2 0 2 4 6 8 10 12 14 16 181

0

1Response of Inventory Investment ( ∆ from SS (% of GDP))

2 0 2 4 6 8 10 12 14 16 1895

100

105Response of Inventories (% of SS)

2 0 2 4 6 8 10 12 14 16 1880

100

120Response of ι (% of SS)

2 0 2 4 6 8 10 12 14 16 181

0

1Response of Storage (% of SS)

Model 0 (RBC)Model 2 (Baseline)


The Timeline of Modern Macroeconomic Theory

The basic framework we have examined is useful as a framework to thinkabout macroeconomics, but modern macroeconomics is more complex.

It is also an ever evolving �eld:

Traditional Keynesian models (60�s)

RBC models (early 80�s onwards)

New-Keynesian Economics (mid/late 80�s onwards)

DSGE models (mid 90�s onwards)


The DSGE Model IThe RBC model was the �rst of a class of models now known as DSGE:

D Dynamic - we look at variables over timeS Stochastic - the model is subject to random shocksGE General Equilibrium - everything e¤ects everything else (if

households work more, the wage rate is driven down)

The RBC model assumes:Perfect competitionno market imperfectionfully �exible nominal prices

Saving determines capital accumulationTechnology shocks and their propagation create �uctuations aroundthe equilibrium of all macro variablesThe model is fully microfoundedFluctuations re�ect optimal reactions to shocks


The DSGE Model IIContributions of the basic RBC model:

Combines growth and business cycles

Uses a relatively simple model and can capture the key features ofbusiness cycles

Hugely important methodological contribution

Objections to the RBC model:

Ignores demand shocks which are known to contribute to output�uctuations

Needs very persistent exogenous shocks to generate reasonable output�uctuation persistence

Cannot generate the highly variable employment numbers, unless weassume very high elasticity of labour supply (not true)

Weak propagation mechanism


The DSGE Model IIIBasic Equilibrium

Equilibrium in DSGE models is given by rational expectations.

Need to combine:

Optimality conditions for equilibrium consumption, savings and laboursupply from household�s maximisation problemOptimality conditions for equilibrium prices of production factors(rental and wage rate) from �rm�s maximisation problemMarket clearing conditions (GE)

Households derive utility from consumption and from leisure and so itfaces a trade-o¤:

Need to work (provide labour) to earn labour income for current andfuture consumptionNeed to have some leisure

Will have a set of equilibrium conditions which will determine theevolution of all variables over time


The DSGE Model IVFrom RBC to alternative DSGE models I

1 Capital market imperfections / Credit Frictions2 Externalities, multiple equilibria and self-ful�lling expectations(Farmer)

3 Sticky information (Mankiw-Reis, Mackowiak-Wiederholt)4 Bounded rationality


The DSGE Model IVFrom RBC to alternative DSGE models II

Capital market imperfections / Credit Frictions

RBC models assume perfect capital markets

But borrowers are better informed than lenders about the risk andreturns on their projects (information asymmetries)

Lenders guard themselves against this extra risk by:

requiring an external �nance premium that is negatively related to thevalue of the borrowers�collateral (Bernanke-Gertler-Gilchrist)limiting the amount of lending to a fraction of the value of thecollateral (Kiyotaki-Moore)

Both lead to a ��nancial accelerator�.


The DSGE Model IVFrom RBC to alternative DSGE models III

Bounded rationality

=) expectations are not always and completely rational

Agents need to �nd out what is going on by learning (Evans &Honkapohja)

While learning, bounded rationality may amplify business cycle andgenerate asymmetric e¤ects (Giannitsarou)

While a reasonable assumption, non-rationality could be anything -too many ways to model bounded rationality


The DSGE Model VIThe future?

Good modelling idea in principleBut these models get very complicated and di¢ cult to solve

Must resort to approximate solutions which are often computationallycomplicated

Assumption of rational expectations and representative agent is oftencriticised (see recent Economist article).

My View:Existing approaches are not perfect, and are certainly not de�nitivetheories of business cyclesHowever, we have gone a long way in explaining lots of things andshown that the methodology is �exible enough to deal with a lot ofthe current challenges:

though they will be hard from a computational point of view

A very exciting and challenging area of current research!McMahon (University of Warwick) Business Cycles 45 / 53

Macroeconomic Indicators

Di¤erent countries have a large (and varied) range of macroeconomicindicators apart from GDP statistics:

Industrial Production Indexes

Prices series

Money Aggregates

Interest Rates

Financial Variables

Demand Indicators (such as retail sales, auto sales, etc)

Surveys (Industrial, Construction, Consumer)

Trade Variables

Labour Market Series

Daily stock markets series


Leading and Co-incident IndicatorsIdentifying business cycles is important information for the public. But:High quality data become available only with a delay �nobody knowsexactly what current GDP is due both to revisions and to delay in datacollection.

For that reason leading and co-incident indicators are valuable:

Leading indicators: Indicators that have predictive power for businesscycle turning points and phases.

The spread between the yield on short and long bonds; New privatehousing starts; Un�lled orders in manufacturing; Number of people onpart-time work.

Co-incident indicators: Indicators that have predictive power forcurrent state of the economy.

Industrial production; Real personal sector income; Real manufacturingtrade and sales; Total non-agricultural payroll and hours


Leading and Co-incident Indicators

6

4

2

0

2

4

6

8

Mar69

Mar71

Mar73

Mar75

Mar77

Mar79

Mar81

Mar83

Mar85

Mar87

Mar89

Mar91

Mar93

Mar95

Mar97

Mar99

Mar01

Mar03

% Real annual GDP growth

Yield curve

RecessionCorrect 2 Recessions

Correct

RecessionCorrect Yield curve accurate

in recent forecast

RecessionCorrect

AnnualGDP growth

or Yield Curve

Data though 3/5/03

6

4

2

0

2

4

6

8

Mar69

Mar71

Mar73

Mar75

Mar77

Mar79

Mar81

Mar83

Mar85

Mar87

Mar89

Mar91

Mar93

Mar95

Mar97

Mar99

Mar01

Mar03

% Real annual GDP growth

Yield curve

RecessionCorrect 2 Recessions

Correct

RecessionCorrect Yield curve accurate

in recent forecast

RecessionCorrect

AnnualGDP growth

or Yield Curve

Data though 3/5/03


Surveys I

Many bodies (o¢ cial and uno¢ cial) conduct surveys in order to gatherinformation on the economy:

They are very timely - often can get them within the period of interest

Expectations of agents matter

But:

there is an issue of how to interpret them

we need to be careful

they can be volatile and may not be representative

sample size is important


Surveys IIEuropean Surveys

All EU countries conduct a common survey of manufacturers, servicecompanies, construction �rms, retailers and consumer:

survey has both backward-looking and forward-looking components;

most questions asked monthly (some quarterly);

respondents choose from a menu of qualitative answers:,

�+ +�(most positive answer)�+�(positive answer)�=�(neutral answer)�-� (negative answer)�- -� (most negative answer)�Don�t know".

Some surveys are combined to give "Economic Sentiment Indicator"


Surveys IIIEuropean Surveys

Example from consumer survey:

1 How do you expect the general economic situation in this country todevelop over the next 12 months? It will...

+ + get a lot better+ get a little better= stay the same- get a little worse- - get a lot worseN don�t know.

Answer is reported as a balance between positives and negativeswhere ++ is worth double +.

�Don�t knows�are discarded.

Often compare to long-term average.


Final Comments on Indicators

When looking at macroeconomic indicators, two additional issues need tobe considered:

1 Seasonal adjustment2 Revisions to statistics


END


Session 3: Economic Diversi�cation

Michael McMahon


McMahon (University of Warwick) Diversi�cation 1 / 18

To Cover

1 Di¤erent Economic Sectors2 How di¤erent sectors are a¤ected by the business cycle3 The bene�ts of diversi�cation


Di¤erent Economic Sectors

We have discussed how GDP can be split into di¤erent expenditurecomponents (C, I, etc...)

But we can also look at how di¤erent industries contribute to nominalGDP:

GDPt = Ind1t + Ind2t + Ind

3t ...+ Ind

Nt

where Ind jt is the contribution of industry j to nominal GDP (which"adds up").

We cannot just use industry output as the contribution - we must usea concept called "Gross Value Added (GVA)".

this avoids any double-counting of industry output


Gross Value Added - An Example I

Consider an economy that has 2 sectors

1 Manufacturing sector:

Imports raw materials from abroad and then creates output from that.

2 Retail and Distribution sector:

Buys the output from manufacturers and sells the goods to consumers.

GVAt = Outputt � intermediate_inputst


Gross Value Added - An Example IIIntermediate goods Output SalesDomestic Imported Intermed. Final

Manufacturing 0 50 100 100 0Retail & Dist. 100 0 150 0 150

1 Expenditure Approach

GDP(E ) = C + I + G + X �M= 150+ 0+ 0+ 0� 50= 100

2 GVA - must yield the same as expenditure approach

GDP(VA) = VAman + VAret

= (100� 50) + (150� 100)= 100


Adding Up Real GDP

It is not possible to simply "add up" real values

Real GDP is often simply reported as an index

Calculation of real GDP uses a method called "chain-weighting"

Following Jorgensen, we calculate real GDP (gYt ) growth as aTornqvist index of growth in each sector:

gYt = ∑i

vi ,t .gVAi ,t (1)

where vi ,t =12

�pi ,0.qi ,0

∑ipi ,0.qi ,0

+ pi ,t .qi ,t∑ipi ,t .qi ,t

�is the average nominal share of

sector i in periods t and t � 1.The main implication of this is that we cannot take real shares - wemust calculate nominal shares.


Overnight Problem

ProblemContinuing to look at the overnight problem - discussion of the issuesencountered and some time to work on the data with my help.


What is Diversi�cation?

De�nitionDiversi�cation is the process whereby an economy expands:

1 the range of sectors it produces in;2 the markets that are served; and/or3 its sources of income.

The objective is to create a stable and sustainable level of income at a(relatively) high level.

Economic diversi�cation is a structural transformation and is oftenaccompanied by economic development.


Why Diversi�cation?

There are well-known e¢ ciency bene�ts from specialization...

..but also risks of too little diversi�cation.

The reasons for diversi�cation typically discussed in the literature (seeZhang) are:

1 Trends in terms of trade2 Price instability in primary commodity markets3 Depletion of mineral resources4 Economies of scale and external economies especially associated withmanufacturing

5 Reduction of portfolio risk


How to Measure Diversi�cation? I

1 share of sector in nominal GDP2 share of sector in exports3 the % of the labour force employment in the sector


How to Measure Diversi�cation? II

1 UN Council on Trade and Development�s (UNCTAD) ExportConcentration Index (ECI):

A Her�ndahl-Hirschman Index

ECI =

K∑k=1

(sk )2 � 1

K

1� 1K

Ranges from1 (low level of economic diversi�cation) to 0 (high level ofdiversi�cation).The calculation takes into account 239 products.

2 UNCTAD�s Economic Vulnerability Index (EVI), a composite indexbased on the share of manufacturing in GDP, the share of labourforce in industry, annual per capita commercial energy consumptionand the ECI.


How do Countries Encourage Diversi�cation?

Foreign-direct investment

creating �nancial incentives for foreign investmentfavourable legislationsidenti�cation of investment opportunities and marketing theopportunities

privatization of industries

may lead to greater e¢ ciency and thus expansion of marketsAbu Dhabi utilities

liberalization of �nancial markets

use of subsidies

other industrialization policies

education and training capacities.


Diversi�cation in the GCC I

Many of the GCC economies stress the need for economicdiversi�cation

E¤ort has been made with:

downstream industries (such as petrochemicals, fertilizers and metals)agriculture (albeit at a high cost in subsidies)services (in the form of �nancial institutions and tourism).

There are problems associated with the diversi�cation process - e.g.the scarcity of water resources in agriculture

Success has followed with reliance on oil falling


Diversi�cation in the GCC II

Diversi�cation does not only mean new industries in the GCC:

GCC countries have natural advantages in solar, wind, geothermal,hydropower and biomass energies:

Existing infrastructure can be adapted for other energy purposes;training may be easier for existing energy workers

These sectors are less vulnerable than oil:

renewable�green�

Thus, measuring just energy sector share does not capture the fullextent of possible diversi�cation


Diversi�cation in the GCC III

But early e¤orts (1970s, 1980s) were not so successful:

development of heavy industries such as petrochemicals, aluminiumand steel such as SABIC and Dubai Aluminium.

Heavy cost in terms of subsidies and tax exemptionsCapital and labour (incl. management) were largely imported

Also, the speed of diversi�cation lead to issues of in�ation, esp. inreal estate - domestic labour force becoming rentier class

Public sector accounted for most employment of local population


Diversi�cation in the GCC IV

Now:

Additional focus on new services such as IT and tourism

diversi�cation within energy sector (lique�ed natural gas)

Creation of further free zones

Real estate ownership opening to nonresidents (e.g. Dubai)

Opening-up the economy to private sector FDI

Encouraging domestic enterprise

Structural and institutional reforms


Diversi�cation in the GCC VSuccess - Non-oil GDP in Nominal GDP (%)


END


Session 4: An Introduction to Forecasting

Michael McMahon


McMahon (University of Warwick) Forecasting 1 / 22

To Cover

1 Basic Statistics and Regression2 Forecasting3 euro-coin indicator


Probability and Classical StatisticsRandom Variables and Probability Distribution I

De�nitionA random variable (RV) is a function that associates a unique numericalvalue with every outcome of an experiment. The value of the randomvariable will vary from trial to trial as the experiment is repeated. Arandom variable, which can be either continuous or discrete, has anassociated probability distribution.

For example:

a toss of a coin (discrete RV)

resultt =�heads w .p. 12tails w .p. 12

a roll of a dice


Probability and Classical StatisticsRandom Variables and Probability Distribution II

A common probability distribution is the normal distribution:

Xvalues0

Probability

Try the repeated coin toss game.McMahon (University of Warwick) Forecasting 4 / 22

Probability and Classical StatisticsRandom Variables and Probability Distribution III

Any observation on a RV is the outcome of a statistical experiment:

Occasionally we know enough about the experiment to predict thefrequency of di¤erent outcomes.Usually, we observe a sample of outcomes, and try to infer the natureof the underlying experimental mechanism from this sample.In both cases we cannot predict the exact outcome of an individualexperiment.

We can imagine two types of economic experiments:1 using a set of characteristics (e.g. wage, education, experience, sex) foran individual unit (e.g. an employee) selected at random from a muchlarger population.

2 analysing an economic time series (e.g. GDP, or individual �rm sales)


Probability and Classical StatisticsRandom Variables and Probability Distribution IV

Economics is not like other science experiments:

We do not (usually) control any of the observed characteristics(unlike a physicist who might control the temperature of a reaction)

we need to adopt statistical methods which cope with situations wheremany characteristics vary simultaneously.

We cannot measure all the relevant characteristics precisely (e.g. wecannot directly observe risk aversion)

we have to assume that the e¤ects of the unobserved characteristicscan be represented by a random disturbance.Therefore, �averaging�over many observations will eliminate theire¤ects and allow us to make inferences about how the observedcharacteristics a¤ect individual behaviour.


Statistics and ProbabilityRegression - The Basics I

You want to model the e¤ect of X on Y where the true relationship is:

Yt = α+ βXt + εt (1)

Yt �! variable to explain (dependent variable)

α �! constant

β �! regression coe¢ cient

Xt �! explanatory variable

εt �! economic shocks in the relationship.


Statistics and ProbabilityRegression - The Basics II

But we don�t know the true relationship so we estimate it using the datathat we have:

Yt = α+ βXt + et (2)

Yt �! variable to explain (dependent variable)

α �! estimated constant

β �! estimated regression coe¢ cient

Xt �! explanatory variable

et �! estimated residuals in the relationship.

The aim of Least Squares regression is to make the line �t as best aspossible (minimise the square of the errors).


Statistics and ProbabilityRegression - The Basics III

Question for the group to discuss: What explains this plot?y-axis = average number of days it takes to start a new businessx-axis = GDP per worker (adjusted for purchasing power di¤erences)

arg

aus

bel

bol

bra

can

chi

chl

col

den

domecu

egy

fin

fragha

gre

hks

hun

idn

ind

ire

isr

ita

jam jap

jor

ken

kor

mex

mlwmly

mor

moz

netnig

nor

nwz

ost

pak

pan

per

phi

por

rom

rsa

sen

sin

spa

sri

swe swi

taitan

thatuntur

uga

unk uns

uru

ven

zam

zim

050

100

150

0 20000 40000 60000real GDP per worker in international dollars version 6.1 of the Penn World Tab

# days to register Fitted values


Statistics and ProbabilityMultivariate Regression

You want to model the e¤ect of X on Y , but we know that Z also plays arole; the true relationship is:

Yt = α+ βXt + ϕZt + εt (3)

We want to make sure that the correlation that we capture is notcapturing something else:

Imagine the true relationship was Yt = α+ ϕZt + εt (β = 0), butcov(X ,Z ) > 0

regression of Yt = α+ βXt + εt would �nd β 6= 0


Statistics and ProbabilityA Closer Look At Residuals

The residual term (et) captures everything not explained by X :Using equations (2), we get:

et = Yt � α� βXt

If the true relationship is (1):

et = α+ βXt + εt � α� βXt= εt + (α� α) +

�β� β

�Xt

which means that so long as our coe¢ cients are unbiased, then theestimated residuals match up with the economic shocks.But if the true relationship is (3):

et = α+ βXt + ϕZt + εt � α� βXt= εt + ϕZt + (α� α) +

�β� β

�Xt

which means that the omitted variable turns up in the residual.


Statistics and ProbabilityCare with Statistical Analysis

As already discussed, you must be careful with how you interpret simplestatistics:

Selection bias - see below

Omitted variable bias - wrongly attribute e¤ects to X rather than to Z

Measurement error

Speci�cation error - we have been talking about linear models, whatabout non-linear models?


Selection Bias

Consider again the regression hospitalization on health outcomes:

y = α+ βD + ε

Using 12,000 + German households in 1994, where health is a measurefrom 1 (ill�health) to 5 (best health), we get the following results:

health = 2.52 � 0.58 �Hospital(0.09) (0.03)

=) Hospital makes people sicker!

=) Need to Randomise our experiment.


Controls and Conditional Randomisation

y = earnings; D = 1 if have Warwick MSc; and X = UG grades.

Question: what is the e¤ect of a Warwick MSc on earnings?If admission to Warwick is more likely for those with high grades, butrandom within grades, then the regression will get us our answer solong as we control for X .

y = α+ βD + γX + ε

But, if admission also depends on recommendations from anundergrad prof, and the letter carries information for earnings, thenour regression will fail!

The regression will wrongly attribute higher earnings to Warwick�s MScwhen it was really something else (letter) that determined both.

A good observational study convinces the reader that,after controlling for X, variation in D is random


Forecasting I

We want to know what will happen in the future to variable Y which isdriven by the true relationship:

Yt+1 = α+ βXt+1 + ϕZt + εt+1 (4)

Yt+1 �! variable of interest to forecast

α �! constant

β, ϕ �! regression coe¢ cients

Xt+1 �! explanatory variables that correlate with Yt+1Zt �! explanatory variables that have a lagged e¤ect

εt+1 �! economic shocks (residuals in the relationship)


Forecasting II

But we may have to use best guesses of Xt+1 and Zt , which we denoteXt+1(= Xt+1 + ηt+1) and Zt (= Zt + νt ), and an estimated equation:

E(Yt+1) = α+ βXt+1 + ϕZt + et+1 (5)

where E(Yt+1) denotes the best guess at time t of what value variable Ywill take at time t + 1 - �expectation�.

The key decisions for the forecaster to make are:

what value for Xt+1 (and Zt) to use?

what value of et+1 to use?


Forecasting III

Forecast errors result from many sources as is highlighted by combiningequations (4) and (5):

error = E(Yt+1)� Yt+1= α+ βXt+1 + ϕZt + et+1 � (α+ βXt+1 + ϕZt + εt+1)

= (α� α) +�

βXt+1 � βXt+1�+�

ϕZt � ϕZt�+ (et+1 � εt+1)

= (α� α) +�

β� β�Xt+1 + βηt+1 + (ϕ� ϕ)Zt + ϕνt + (et+1 � εt+1)

Which means that there are, potentially, lots of sources of error!


Forecasting IV

4

3

2

1

0

1

2

3

4

5

6

1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004

Per centResiduals from a US Business Investment Equation

4

3

2

1

0

1

2

3

4

5

6

1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004


What next?


Forecasting IV

4

3

2

1

0

1

2

3

4

5

6

1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004


4

3

2

1

0

1

2

3

4

5

6

1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004


4

3

2

1

0

1

2

3

4

5

6

1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004


Choices to make...


Eurocoin - A "Nowcasting" tool

De�nitione-coin is a real-time, monthly estimate of area-wide underlying GDPgrowth, computed each month by the sta¤ of the Banca d�Italia.It provides a single number summarizing the current economic outlook forthe Euro area.e-coin is obtained by collecting a large set of statistics (such as industrialproduction, surveys, stock market and �nancial data, demand indicators )and extracting from it the information that is relevant to forecast futureGDP trend.It tracks GDP growth anticipating o¢ cial GDP releases by several months.It is a �smooth version�of GDP growth...


Eurocoin - Macroeconomic Indicators UsedThe database uses 141 data series:

industrial Production Indexes (41 series)

Prices (24)

Money Aggregates (8)

Interest Rates (11)

Financial Variables (6)

Demand Indicators (14)

Surveys (25)

Trade Variables (9)

Labour Market Series (7)

Daily stock markets series

See other slides for (i) what it does, (ii) its real-time performance, and (iii)why use so much data.McMahon (University of Warwick) Forecasting 20 / 22

Eurocoin - Recent Performance


END


Money, Banking, the Current Crisis and the PolicyResponse

Michael McMahon

McMahon () Current Crisis 1 / 116

Money, Banking, the Current Crisis and the PolicyResponse

The basics of the �nancial system including the main instruments

Why the current crisis? Focus on the concept of leverage and balancesheet e¤ects.

A brief discussion of monetary policy including models of monetarypolicy

Class discussion of the liquidity trap, the zero nominal bound oninterest rates and �quantitative easing�

Fiscal policy response including a discussion of how �scal policyoperates in the economy.

A discussion of exchange rate policy

McMahon (Warwick) Current Crisis 2 / 116

The Basic Concept

De�nitionA �nancial system is de�ned as a set of markets for �nancial instruments,and the individuals and institutions who trade in those markets, togetherwith the regulators and supervisors of the system.[Howells and Bain, 2003]

The basic function of the �nancial system is to transfer resources fromthose with excess funds for investment, to those who require more fundsfor investment.


The Basic Concept


The PeopleBorrowers

Firms

Households

The government

Foreigners - varies


The PeopleLenders

Households

The government

Firms

Foreigners - varies

Same people - but usually in a di¤erent order.


Why do Lenders Lend?

Lenders will worry about:

the return that they can get.

the risk surrounding this return:

default risk;income risk;capital risk;in�ation risk;

liquidity

If we wish to encourage more lending, one way is to increase the returnwhich we o¤er to lenders.


What determines borrowers behaviour?

Borrowers will worry about:

the return that they must pay to get the funds;

the terms of this return;

for example, debt is not state-contingent while equity is.

the length and �exibility of the borrowing;

�rms will not wish to have debts which are too easily recalled.


The Financial Account

Like in double-entry book-keeping, all de�cits (borrowing) must have ano¤setting surplus (lender)In the national accounts, it is possible to look at the �nancial accounts ofdi¤erent sectors over time.In Europe:

Private Non-Financial Corporations (PNFCs)

Monetary and Financial Institutions (MFIs)

Government

Households (and NPISH)

Foreigners - this is the current account


The Financial Account

8.0

6.0

4.0

2.0

0.0

2.0

4.0

6.0

8.0

1999 2000 2001 2002 2003 2004 2005 2006

Households PNFCsMFI GovtNet with ROW

% of GDP


How the Matches Occur

Lending can be:

1 Completely direct - lenders seeking out other agents who need toborrow.

2 Direct lending through a market.3 Through �nancial intermediation:

direct; orthrough a market.


The Di¤erent Types of Market

There are di¤erent �nancial markets for each di¤erent type of instrument

Primary versus secondary markets

OTC versus exchange markets

Dealers versus Brokers

Money-market versus capital market (maturity)


Financial Intermediation

De�nitionInstitutions that borrow funds from people who have saved and in turnmake loans to others. [Mishkin, p.7]

So �nancial intermediates include banks, but will also include pensionfunds, insurance companies, etc...Why go through a �nancial intermediate:

Reduced transaction costs

Risk diversi�cation

Maturity transformation

Reduce Asymmetric Information

Adverse SelectionMoral Hazard


Financial Intermediation

Asymmetric Information (also very important in insurance)

1 Adverse Selection

Occurs before the transactionOnly choose those who are most risky - therefore potentially notchoose anyone at all

2 Moral Hazard

Occurs after the transactionOnce they have the service, their behaviour changes in undesirableways.


The Di¤erent Types of Instruments

The main two instruments to distinguish are:

Equity

May get an annual share of pro�ts as dividendOwns part of the company )voting rightPrice varies depending on supply and demand

Debt

Contractually �xed return

per period interestprincipal at maturity date

Preferential debtorNo voting right


No Single Interest RateThere is NOT 1 single interest rate:

0

5

10

15

20

25

Apr71 Apr81 Apr91 Apr01

Fed funds Nominal 10yearAAA BAAState & local Mortgage

%


The Di¤erent Types of Instruments - USMoney Market Instruments

Level in US $, billionsType of instrument 1980 1990 2000 2005US Treasury Bills 216 527 647 923Negotiable bank certi�cates of deposit 317 543 1053 1742Commercial Paper 122 557 1619 1544Repo agreements 57 144 366 518Eurodollars 55 92 195 438


The Di¤erent Types of Instruments - USMoney Market Instruments

Level (% of GDP)Type of instrument 1980 1990 2000 2005US Treasury Bills 8 9 7 7Negotiable bank certi�cates of deposit 11 9 11 14Commercial Paper 4 10 16 12Repo agreements 2 2 4 4Eurodollars 2 2 2 4


The Di¤erent Types of Instruments - USCapital Market Instruments

Level (% of GDP)Type of instrument 1980 1990 2000 2005Corporate Stocks 57 71 180 144Residential Mortgages 40 50 56 76Commercial and Farm Mortgages 13 14 12 15Corporate Bonds 13 17 23 24Government Bonds 18 22 29 36Consumer Loans 13 14 12 15


Regulation

Financial services are a highly regulated industry (and have been for a longtime). But why do we regulate �nancial services?

Banking relies on con�dence of public

fractional reserves systemliquidity mismatch between assets and liabilities

Contagion...

Consumer protection...not caveat emptor!


Regulation

Problems because of regulation:

Moral hazard

Compliance costs

Costs of entry and exit are higher - more monopoly power!


The Macroeconomic E¤ects of Financial Markets

ProblemWhy might the �nancial system have macroeconomic e¤ects?


How Banking Works - The Bank Balance Sheet


Bank Balance Sheets

Assets LiabilitiesReserves 4% Checkable Deposits 7%Securities 23% Time Deposits 59%Loans 66% Borrowings 26%Other assets (e.g. physical capital) 7% Bank Capital 8%

decreasing liquidity

Basic idea is to make more on the assets than you pay on the liabilities

) pro�ts


Bank T-accounts

New account opened with cash (£ 100)

Assets LiabilitiesReserves +£ 100 Checkable Deposits +£ 100

If opened with a cheque from another bank, pretty much the same - cashin process of collection.


Bank T-accounts

What if the new account holder then wrote a cheque for £ 50?

Assets LiabilitiesReserves -£ 50 Checkable Deposits -£ 50

.


Bank T-accounts

Creating new loans after the £ 100 deposit is received (ignore last chequetransaction)

Assets LiabilitiesRequired Reserves +£ 10 Checkable Deposits +£ 100Excess Reserves +£ 90

Reserves = RR+ ER

Assumes a 10% required reserves (% of deposits) - no interest on reserves


Bank T-accounts


Assets LiabilitiesRequired Reserves +£ 10 Checkable Deposits +£ 100Excess Reserves 0Loans +£ 90

Reserves = RR+ ER

...so use the excess to create new pro�ts - issue new loans


Bank T-accounts


Assets LiabilitiesRequired Reserves +£ 10 Checkable Deposits +£ 100Excess Reserves 0Loans 0Securities +£ 90

Reserves = RR+ ER

...so use the excess to create new pro�ts - issue new loans (or could buysecurities).


Bank Management (brie�y)

1 Liquidity and Reserve management2 Asset and liability management3 Capital adequacy - trade-o¤

Capital prevents failures (cushion against drop in value of assets);Capital lowers return to shareholders;

4 Credit risk and interest rate risk


Liquidity and Reserve management

Consider the Bank which created the £ 90 of extra loans and kept noexcess reserves (assume this is the whole balance sheet of the �rm):

Assets LiabilitiesRequired Reserves £ 10 Checkable Deposits £ 100Excess Reserves 0Loans £ 90

If there is a (relatively small) £ 10 withdrawal of deposits, deposits drop by£ 10 and reserves would be exhausted (-£ 10). The Bank is about to fail -they have a reserve shortage of £ 9...


Liquidity and Reserve management

Consider the Bank which created the £ 90m of extra loans and kept noexcess reserves (assume this is the whole balance sheet of the �rm):

Assets LiabilitiesRequired Reserves £ 0 Checkable Deposits £ 90Excess Reserves 0Loans £ 90

They need to:

Borrow the required reserves, at a cost, from other banks (borrowingliabilities +£ 9, reserves +£ 9);Borrow the required reserves, at the discount rate, from the CB;Try to recall loans worth £ 9 and put the money into reserves;If they had other assets (securities or even physical capital), theycould also try to sell £ 9m of this.McMahon (Warwick) Current Crisis 34 / 116

Bank Management (brie�y)

1 Liquidity and Reserve management2 Asset and liability management3 Capital adequacy - trade-o¤

Capital prevents failures (cushion against drop in value of assets);Capital lowers return to shareholders;

4 Credit risk and interest rate risk


Central Bank Balance Sheets

Like the banks, the Central Bank (Fed, BOE, ECB) also maintains abalance sheet:

Assets LiabilitiesGovernment Securities Currency in circulationDiscount loans to banks Reserves (from Banks)

Other assets

The liabilities side is known as the monetary base (MB = Cash+ reserves)The CB also engages in Open Market Operations to change the monetarybase


Central Bank Open Market Operations

Consider a $100 purchase of a government bond (from Bank A):

Central BankAssets Liabilities

Government Securities +100 Currency in circulation -Discount loans to banks - Reserves (from Banks) +100

Other assets -

Bank AAssets Liabilities

Required Reserves - Checkable Deposits -Excess Reserves +100Loans -Securities -100


The Subprime and Financial Crisis


The Subprime and Financial Crisis IUS House Prices Since 1890 (adjusted for in�ation)

Source: S&P, CaseShiller Index


The Subprime and Financial Crisis IIUS House Prices and Equity Markets Have Fallen Sharply


The Subprime and Financial Crisis IIIUS House Prices and Equity Markets Have Fallen Sharply


The Subprime and Financial Crisis IVThe Fall has led to a rise in foreclosures


The Subprime and Financial Crisis V

But bank losses due to mortgages in this crisis are relatively small

approximately US$1,000 mnthis is equivalent to a 3% fall in the stock market (at end 2006 prices)In October 1987, the S&P 500 fell 20 % in a single month with almostno macroeconomic repercussions

Therefore, we need to understand how the shock got transmitted tobecome the large scale, global crisis that we have seen in the last 2years.

I will focus here on the idea of leverage.


Leverage and the US Banking Crisis I

Leverage =assetscapital

Assets LiabilitiesReserves Checkable DepositsSecurities Time DepositsLoans BorrowingsOther assets (e.g. physical capital) Bank Capital


Leverage and the US Banking Crisis II


= 10

Assets LiabilitiesAssets 100 Debt 90

Bank Capital 10


Leverage and the US Banking Crisis II


= 9.18


Bank Capital 11

If debt constant, leverage is inversely proportional to asset values:

assets # ) leverage "

Holding debt constant ) counter-cyclical leverage


Leverage and the US Banking Crisis IIIHouseholds


Leverage and the US Banking Crisis IVPNFCs


Leverage and the US Banking Crisis VBanks


Leverage and the US Banking Crisis VIBanks


Leverage and the US Banking Crisis VIIImplications of Asset Price Fall with Procyclical Leverage


= 10 = TARGET


Bank Capital 10

Pro-cyclical leverage means that as asset prices go up, we increaseleverage by buying even more assets using debt.

But buying more assets pushes the value of assets " ) circular

When prices start to fall, we get a �resale of assets.




= 11

TARGET = 9 (falls)


Bank Capital 9




= 11

NEW TARGET = 9 (falls)


Bank Capital 9

To achieve new target the bank must sell o¤ assets and use theproceeds to pay down debt

1% fall in asset prices ) 18% sale of assets which puts furtherpressure on prices.


Leverage and the US Banking Crisis VIIIImplications of Asset Price Fall with Procyclical Leverage


= 11

NEW TARGET = 9 (falls)

Assets LiabilitiesAssets 101.25 Debt 90

Bank Capital 11.25

To achieve new target the bank must sell o¤ assets and use theproceeds to pay down debt

1% fall in asset prices ) 18% sale of assets which puts furtherpressure on prices.

Could have raised capital and used the proceeds to buy assets.raised 2.25 in new capital and buys assets with this.


Leverage and the US Banking Crisis IXA Game of Two Halves

In the beginning of the crisis (up until August 2008), banks tried to coverthe declines in the values of their assets (marked-to-market) and reducethe �resale activity by raising new capital:

Source: Bloomberg

($Bn)

CitigroupWachoviaMerril LynchWashington MutualBank of America

UBSIKB DeutscheCredit SuisseDeutsche BankCredit Agricole

HSBCRBSBarclaysHBOSLloyds TSB

Mizuho FGNomura HoldingsBank of ChinaMitsubishi UFJBank Hapoalim

333.3

170.4

62.824.2

235.4

118.6

58.122.2

0

50

100

150

200

250

300

350

North America Continental Europe UK Asia

Sum of Writedown & Loss Sum of Capital Raised


Leverage and the US Banking Crisis XA Game of Two Halves

From September 2008, were unable to raise new capital (con�dence wasgone after Lehman) and so the governments stepped in to recapitalise thebanks and stop the �resale.

1

4

5

6

11

14

36

67

144

490

0 200 400 600

Iceland

Luxembourg

Switzerland

Russia

Belgium

France

Netherlands

Germany

UK

US•Fannie and Freddie (200), Wachovia (12), Citigroup (25),

JPMorgan (25), Bank of America (20), Merril Lynch (5), WellsFargo (25), Goldman Sachs (10), Morgan Stanley (10), New YorkMellon (3) and State Street (3), AIG (123), Bear Sterns (29)

•Bradford and Bingley (32), RBS (35), HBOS (20), Loyds TSB (10),Northern Rock (47)

•Hypo Real Estate (67)

•Fortis (23), ING (13)

•Dexia (5), Credit Agricole (4), BNP Paribas (3), Soc. Gen. (2)

•Fortis (6), Dexia (5)

•Globex (2), Svyaz Bank (2), Sobinbank (0.5), KIT Finance (1)

•UBS (5)

•Fortis (3), Dexia (1)

•Glitnir (1), Landsbanki (0), Kaupthing (0)

US$bn


Leverage and the US Banking Crisis XIA Game of Two Halves

Seems to be working.

Interest rate at which banks are willing to lend to each other


Northern Rock - A slightly di¤erent problem

ProblemFrom the provided articles, can you develop an understanding of why theNorthern Rock crisis occurred? What were the primary failings of thebank? What was the policymakers response?


Monetary and Fiscal Policy


Monetary Policy

Interest rates are discussed everyday in the papers and in �nancialmarkets...Why?

Monetary policy is important because interest rates determine:

The interest rate is the cost of borrowing;

It a¤ects investment decisions by �rms;

It determines consumption decisions by families;=) it a¤ects aggregate demand.

In a recession. . . if aggregate demand is low, can stimulate economy bylowering interest rates.In a boom. . . booms driven by unusually high aggregate demand tend tolead to in�ation (overheating) and we can prevent this in�ation by reigningin aggregate demand through higher interest rates.


Monetary Policy - contd

Ideal: Be �Tight� in booms and �Loose� in recessions

But:

�Long and variable lags� faced by Central Bankers;

Only get clear picture of where economy is after the fact (recognitionlag);

Takes time for interest rate changes to work their way into aggregatedemand and prices;

Very hard to tell if you are being �too loose,��too tight,� or �justright�at any particular moment;

Not always clear whether cycle is driven demand or supply shock

e.g. oils shocks of the 1970s lead to a �supply-side� recession;�stag�ation�.


Monetary Transmission Mechanism

De�nitionThe monetary transmission mechanism describes how policy-inducedchanges...in the short-term nominal interest rate impact on real variablessuch as aggregate output and employment. Speci�c channels of monetarytransmission operate through the e¤ects that monetary policy has oninterest rates, exchange rates, equity and real estate prices, bank lending,and �rm balance sheets.

We do not know exactly what the size of each channel is (with anycertainty), but we do have good ideas about which are the main channels.

"Monetary Policy acts with long and variable lags" - MiltonFriedman


Monetary Transmission Mechanism - Reality


Starting at the End - Consumer Price In�ation(CPI)

%∆CPIt =N

∑i=0wi .%∆pi ,t

where wi is the weight of good i in the consumers basket of goods, and%∆pi ,t is the change in prices of the ith good.

Splitting the N goods into:

domestically produced - determined more directly by domesticdevelopments/producers;

imported goods - determined by foreign producers, but also:

the exchange rate (directly) - e.g. price of a $10 book in UK £ ifexchange rate is $1=£ 1 versus if $2=£ 1?relative price shocks - e.g. oil prices.


Domestically Generated In�ation - Role of the Output GapDe�nitionWhen there is a positive (negative) output gap, i.e. demand is greater(lower) than the potential output measure, then there will be upward(downward) or in�ationary (de�ationary) pressures on prices.

What can monetary policy do:

Monetary policy a¤ects the demand side of the economy (AD curve);Policymakers will need to try to �gure out where the AS curve is - itis constantly changing.

"It is plain that if you are trying to hit an in�ation target, youhave to form a judgment about the level of demand in theeconomy relative to potential supply and how it is likely to move.This must, among other things, involve making judgments aboutgrowth prospects, not for their own sake but because they arevital when it comes to understanding the prospects for in�ation."


How Monetary policy a¤ects AD

AD = C + I + G + (X � Z )where C is �nal consumption by individuals and families mainly, I isinvestment in gross �xed capital (by �rms and households), G isgovernment spending, (X � Z ) is the trade balance (X is exports and Z isimports).We can broadly assume the G is not a¤ected too much by monetarypolicy- although there is some e¤ect from the business cycle on G .The other 3 will depend on how the interest rates a¤ect:

1 market interest rates;2 asset prices (equities, house prices, etc...);3 the exchange rate;

For all 3 e¤ects, expectations and how the move is interpreted plays a keyrole.


How Monetary policy a¤ects ADMarket interest rates

As we already saw, the short rate will closely follow the policy rate andthen the longer term rates will depend on the expectations of futureinterest rates.

If the interest rate is increased, people might think:

interest rates in the future will be higher (better growth prospects);

lower future interest rates.

Therefore, it is necessary to understand how decisions will be interpretedand how expectations will change (see later).


How Monetary policy a¤ects ADAsset Prices

If interest rates increase, and so bond yields also increase, bond prices willfall:=) this can generate a wealth e¤ect.=) this can also cause the cost of �rms investment and �nancing toincrease.

A similar thing may happen with equities - if the future is discounted witha higher interest rate, then equities decline with interest rate increases...but what if the policy is interpreted as signalling better growth prospects?

House prices, as a result of higher interest rates, may fall because lesspeople can a¤ord to take out large loans.


How Monetary policy a¤ects ADExchange Rates

When the interest rate increases, then the exchange rate should, ceterisparibus, appreciate:

Foreigners will get a higher return on savings in the UK thanelsewhere - money �ows into the UK;

UK residents will rather save at home rather than abroad - less money�ows out;=) higher demand for the pound means that it will increase in valuerelative to other currencies.

The exchange rate is simply the price of one country�s money in terms ofanother country�s money.


How Monetary policy a¤ects ADConsumption

The e¤ect on consumption comes through:

1 Market interest rates

depends on net position!depends on how quickly the interest rates are changed;depends on whether interest rates are �xed or �oating;depends on country of borrowing;

2 Asset prices and wealth e¤ects

Poorer people spend less.


How Monetary policy a¤ects ADInvestment and Net Trade

The e¤ect on investment will come through:

1 the e¤ect on market interest and other �nancing costs;2 the e¤ects on demand.

The e¤ect on net trade is driven:

1 the exchange rate

appreciation lowers (X � Z ) as foreign goods seem cheaper (Z ") andforeigners will sell more of the home produced good (X #).


Monetary Policy in Action

ProblemUsing the AS-AD framework developed yesterday, answer and explain eachof the following questions:

1 Can monetary policy can o¤set a reduction in consumption spending(perhaps related to uncertainty)?

2 How might monetary policy lead to �stag�ation�by wronglyresponding to an oil price increase?


Second Round E¤ects and ExpectationsE¤ects described so far are �rst round e¤ects - directly the result ofchanges in the interest rate. We must also consider the second rounde¤ects:

Price shocks (such as oil) feeding into the wage setting process;

First round e¤ects of GDP leading to other changes in demand in theeconomy - explains the widespread e¤ects of interest rates on allsectors.

While the central bank can do nothing about the �rst round e¤ects ofsupply shocks, they can deal with the second round e¤ects.How the policy decisions are interpreted can have a key e¤ect on theeconomy:

In�ation expectations are particular important

Role in wage-setting to determine in�ation;Real interest rate is key = int.rate � πe


Monetary Transmission Mechanism - Reality


Long and Variable Lags


Policy Response and Zero Nominal Bound on InterestRates I

Central Banks in the US and UK reacted to the current recession bycutting interest rates aggressively.One reason is that they are concerned about the liquidity trap.


Policy Response and Zero Nominal Bound on InterestRates I

Central banks cannot set a negative nominal interest rate (it � 0);But it is the real interest rate that matters:

rt = it � πet

Therefore, if πet > 0, real interest rate can go negative.

When demand is very weak, in�ationary pressure will fall and peoplemight expect de�ation (πet < 0)

If it = 0 and πet < 0 then the real interest rate is positive - sodemand falls more and people expect even more de�ation...

...so the real interest rate rises again and demand falls even more.

This happened in Japan�s lost decade of the 1990s.


Quantitative Easing as Alternative to Interest Rates


Monetary Policy Independence

We will shortly discuss di¤erent monetary policy frameworks, but for nowthe relationship we are interested in is Central Bank Independence (CBI).

One of the great trends in monetary policy of the last twenty years (e.g.Gordon Brown�s �nest hour).

CBI independence means that the CB runs monetary policy independentlyof the executive.

BOE, FED, and ECB all quite independent

Various reasons for CB independence, but a clear one is to counteractthe in�ation bias that is inherent in democratic politics.


How to de�ne independence


Independence and In�ation

Central Bank Independence and Inflation, 195588

0

1

2

3

4

5

6

7

8

9

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5

Central Bank Independence Index

Ave

rage

Infla

tion

Rat

e


Time Inconsistency and Rational Expectations

The key idea is:

1 When the govt is given an opportunity to cheat workers, they will;2 Knowing that they will be cheated, rational agents will build thisexpectation of being cheated into their plans;

3 This leads to a higher level of in�ation but with no gain in terms oflower unemployment.

=) Comes from the timing of the govt making its choice after the publichave to set their in�ation expectations - time inconsistency.

Time inconsistency problem - when the economic agents were rational,and where expectations in�uence the decision of government policy,discretionary policymaking made everyone worse o¤ compared with asituation in which the government could commit to a particular policypath - is actually due to Kydland and Prescott.


Monetary Policy Rules I

In order to make monetary policy more transparent (and credible)independent central banks will generally operate under explicit rules ofconduct:

Exchange rate targeting

Monetary targets

In�ation targets


Monetary Policy Rules II

Currently most fashionable rule, in Western countries, is in�ationtargeting.Exchange rate targets are still very popular.


In�ation Targeting I

De�nitionAn in�ation-targeting central bank is one that announces aonce-and-for-all policy of keeping the in�ation rate at or close to somelong-run target, e.g. 2 percent.

Various versions of in�ation targeting currently adopted by (amongothers): New Zealand (1990), Canada (1991), U.K (1992), Sweden(1993), Finland (1993), Australia (1994).

Very simple and easily understood rule.

Removes some uncertainty about direction of policy

It does not imply central bank will do nothing about recessions =) Ifdemand is down and prices start falling (or are forecast to fall) centralbank will react with lower rates to prevent in�ation falling belowtarget.


In�ation Targeting II

Some issues:

What rate of in�ation? =) Low, stable and positive in�ation:

In�ation indices are biased (+1.1% in US - Boskin Commission Report);Downward rigidities in wages;Zero nominal bound on interest rates - liquidity trap.

Why not price level targeting?

With price level targets, we would have to reverse the surprise in�ationWith in�ation targets, we allow bygones be bygones.


UK In�ation

0

5

10

15

20

25

30

1964 Q

1

1966 Q

1

1968 Q

1

1970 Q

1

1972 Q

1

1974 Q

1

1976 Q

1

1978 Q

1

1980 Q

1

1982 Q

1

1984 Q

1

1986 Q

1

1988 Q

1

1990 Q

1

1992 Q

1

1994 Q

1

1996 Q

1

1998 Q

1

2000 Q

1

2002 Q

1

2004 Q

1

2006 Q

1


Taylor Rule

Rt = R + 1.5.(πt � π�) + 0.5.(y � y �)

The CB increases the interest rate when in�ation is high and whenoutput is high;Variations on this rule also work relatively well as descriptions ofmonetary policy in Europe although weights di¤er

- weight on in�ation higher (lower) in more (less) conservative countries


Taylor Rule in the UK

Actual Base Rate Compared with Simple TaylorRule

05

1015202530354045

1973

Q1

1974

Q4

1976

Q3

1978

Q2

1980

Q1

1981

Q4

1983

Q3

1985

Q2

1987

Q1

1988

Q4

1990

Q3

1992

Q2

1994

Q1

1995

Q4

1997

Q3

1999

Q2

2001

Q1

2002

Q4

perc

ent i

nter

est r

ate

Base Rate Simple Taylor Rule


Monetary Policy in Practice I

As already discussed, the BOE has an in�ation target (2%� 1pp) butmust also look at GDP growth:

"...the objectives of the Bank of England shall be �(a) to maintain price stability, and(b) subject to that, to support the economic policy of Her

Majesty�s Government, including its objectives for growth andemployment." .Bank of England Act, 1998


Monetary Policy in Practice II

The decisions are made by a 9-person committee called the MonetaryPolicy Committee (MPC):

Governor

2 x Deputy Governors

Chief Economist

Director for Financial Operations

4 x External experts appointed by HM Treasury.

MPC has operational independence but not goal independence (contrastwith ECB).

Decisions are made on a monthly basis - the �rst Thursday of everymonth, and the decision is taken by way of a majority vote (Governor hascasting vote in event of a 4-4 tie).


Monetary Policy in Practice III

In reality, because of long lags from interest rate decisions to in�ationoutcomes, the Bank of England is really an in�ation-forecast targetingCentral Bank:

1 the MPC forms its best judgement about where in�ation is going (thehorizon is generally 18 months - 2 years);

2 if in�ation is too high at that horizon (above target) they will increaseinterest rates, whereas if in�ation is below target they should cutinterest rates;

3 They will also worry about the outlook for GDP (given (b) above andalso the e¤ect of demand on in�ation)

4 They aim to get a good chance of hitting the in�ation target(2%� 1pp) in the future.

The question is how they �gure out what in�ation is going to be like?


MPC TimetableTiming EventThroughout the month Circulation of brie�ng material and analysis of

data releases and market developments by sta¤.Friday before policy meeting Half-day pre-MPC brie�ng meeting

Monday/Tuesday Sta¤ undertake follow-up work requestedby the Committee.

Wednesday Policy meeting commences early afternoon.Committee identi�es the key issues anddebates their implications for in�ation prospects.

Thursday Policy meeting concludes. MPC providetheir assessment of the appropriate policy stanceand vote on the level of interest rates.Policy announcement at noon, implementedimmediately


Pre-MPC TimetableInternational environment Global oil and commodity price trends;

Developments in the US, euro area and Japan.

Monetary and �nancial UK money and credit aggregates; interest rates�nancial markets and exchange rates.

Demand and output UK GDP & demand components

Agents�overview Business contacts�perspective on latest economictrends. Special topic also presented.

Labour market Employment, hours worked, unemployment andinactivity; vacancies; skill shortages; earnings andsettlements; productivity and unit labour costs.

Prices UK energy, and other commodity prices;producer prices; retail prices; price de�ators.


MPC Forecast

Every quarter (3 months), the MPC agrees on a forecast for in�ation andproduces "fancharts" for GDP growth and in�ation.


Monetary Policy at the ECB and Fed

US Euro area UKWho makes Decisions? FOMC Governing Council MPC/Chancellor

Composition: 12 members predominantlyFed presidents

19 members predominantlyrepr. national interests

9 members economic experts

Stated Objective None (A) Price Stability (inflationbelow 2%)

(A) Inflation 2% +/1pp

(B) Growth & Unempl. (B) Growth and Unempl

When do they meet Every 6 weeks Monthly to make interest ratedecisions (was twice monthlyuntil 2001)

Monthly

How decisions are made Officially vote.In practice Fed president (i.e.Bernanke) makes the decision

The Governing council reachesan agreement and no vote isever taken or published.

Vote is taken and published inthe minutes which come outafter the meetings

Communication strategies Statement on decision Statement on decision Statement on decisionBernanke before congress Press conference after decision

is madeQuarterly inflation report andpress conference

Minutes of FOMC meetings (3weeks delay)

Minutes of MPC meetings(2 weeks delay)

Speeches Speeches SpeechesOther official publications Other official publications Other official publications


Fiscal Policy

Governments can boost AD by:

cutting taxes (which will boost C or I )spending more (which boosts G )

The e¤ects are the same as the e¤ects of monetary policy.

However, the use of �scal policy is subject to:

Recognition lag;Decision lag - hard with all the politics;Impact lag.

Thus, �scal policy not a good instrument for ��ne tuning�.

) Often just allow automatic stabilisers to operate.


Fiscal Policy and The E¤ects of Taxation I

Highest Marginal Income Tax RateSource: Forbes Misery Index (2005)

High marginal tax rates distortionary and to be avoided –seen significant declines in taxes over last decade

0

10

20

30

40

50

60

70

France

DenmarkSwede

n

Belgium

Netherl

ands

Finland

Austria

Poland

Sloven

iaJap

anIsr

ael

USA + N

ew Y

ork st/cit

y

Austral

ia

Canada

(Onta

rio)Chin

aSpain

NorwayIta

ly

Germany

and B

erlin

GermanyIre

land

Greece

Portugal

United K

ingdo

m

Switz

erland

(Zuri

ch)

South Afri

ca

Taiwan

South Korea

Luxem

bourg

Hungar

y

USA +

Illinois

Thailan

d

Turkey

Argentin

aMalt

a

USA + Tex

as

Indon

esiaUSA

India

Lithuani

a

Czech R

epublic

MexicoCypru

s

Malaysi

aBraz

il

Latvia

USA an

d Refo

rmEsto

nia

Singap

ore

Slovakia

Romania

Hong K

ong

UkraineRussi

a

GeorgiaUAE


Fiscal Policy and The E¤ects of Taxation II

Highest Employer Social Security Contribution RateSource: Forbes Misery Index (2005)


0

5

10

15

20

25

30

35

40

45

50

France

DenmarkSw

eden

Belgium

Netherl

ands

Finlan

d

Austria

Polan

d

SloveniaJap

anIsr

ael

USA + New

York st/

city

Austral

ia

Canada

(Onta

rio)ChinaSp

ain

NorwayIta

ly

Germany

and Berl

in

GermanyIre

land

Greece

Portu

gal

United K

ingdo

m

Switzerl

and (Z

urich)

South Afri

ca

Taiwan

South

Korea

Luxem

bourg

Hungar

y

USA + Ill

inois

Thailan

d

Turkey

Argentin

aMalt

a

USA + Tex

as

Indon

esiaUSA

India

Lithuani

a

Czech R

epublic

MexicoCypru

s

Malaysi

aBraz

il

Latvia

USA and R

eformEsto

nia

Singapore

Slovak

ia

Romania

Hong K

ong

UkraineRuss

ia

GeorgiaUAE


Fiscal Policy and The E¤ects of Taxation III

Highest Employee Social Security Contribution RateSource: Forbes Misery Index (2005)


0

5

10

15

20

25

30

35

France

DenmarkSwede

n

Belgium

Netherl

ands

Finland

Austria

Poland

SloveniaJap

anIsr

ael

USA + New

York st/

city

Austral

ia

Canada

(Ontar

io)ChinaSpa

in

NorwayIta

ly

Germany

and Berl

in

GermanyIre

land

Greece

Portugal

United

Kingdo

m

Switz

erland

(Zuri

ch)

South

Africa

Taiwan

South

Korea

Luxem

bourg

Hungar

y

USA + Ill

inois

Thailan

d

Turkey

Argenti

naMalt

a

USA +

Texas

Indon

esiaUSA

India

Lithuani

a

Czech R

epublic

MexicoCypru

s

Malaysi

aBraz

il

Latvia

USA an

d Refo

rmEsto

nia

Singapore

Slovakia

Romania

Hong K

ong

UkraineRussi

a

GeorgiaUAE


Fiscal Policy Response to the Current Crisis I

Large �scal stimulus response by Barack Obama�s Administration totry to boost US demand

The Administration expects large e¤ects (see Christina Romer paper):

Real GDP Employmentbn of 2000 $ ,000s

Without Stimulus 11, 770 133, 876With Stimulus 12, 203 137, 550E¤ect of Package +3.7% +3, 675


Fiscal Policy Response to the Current Crisis II


Fiscal Policy Response to the Current Crisis III

Source: Congressional Budget Office


Fiscal Policy Response to the Current Crisis IV

ProblemRead the summary article in the course pack. Comment on the reasonsthat some economists believe that the Obama Administration is beingoptimistic.


The Exchange Rate


A Brief History of Exchange Rates

1 The Gold Standard - Up to World War II

UK adopted gold convertibility in 1717;US in 1873 (de facto), 1900 (de jure).Both World Wars interrupted it

2 Bretton Woods - 1944-1970s

Coordinated system of exchange rates with capital controls;Dollar was the anchor;Collapsed as dollar continually required devaluation (as a result ofin�ation after govt de�cits and Vietnam).

3 Floating systems4 Fixs and Dollarisation

ChinaERM


The Impossible Trinity I


The Impossible Trinity IIUncovered Interest Parity I

Invest in US Invest in EUEx. rate S$,etInvestment in $ $X $Xe amount e X

S $,etInterest rate r r f

Ex. rate (expected) S$,et+1Receive back $X .(1+ r) e

�XS $,et

�.(1+ r f )

$ amount $�

XS $,et

�.(1+ r f ).S$,et+1


The Impossible Trinity IIIUncovered Interest Parity II

X

S$,et

!.(1+ r f ).S$,et+1 = X .(1+ r)

S$,et+1S$,et

!=

(1+ r)(1+ r f )

%∆�S£ ,e

�� r � r f

r = r f if �xed


Government Intervention in Exchange Rate Markets I

If the exchange rate is �oating, the government can try to in�uence it byintervening in the market:

For example, the US could try to make the $ worth more bypurchasing it in the market:

They would increase demand for $ by buying them...in exchange for foreign currency (£ ) from their reserves.

This might drive the price of $s (ex. rate) up.

But it also removes $ from the US system.

This is an unsterilised intervention.


Government Intervention in Exchange Rate Markets II

If the government did not want to a¤ect the money supply, they would:

Do as before:

They would increase demand for $ by buying them...in exchange for foreign currency (£ ) from their reserves.

And undertake an OMO to o¤set the removal of $:

In this case they buy government securities and so release $ into theUS system again.

This is a sterilised intervention - they have in e¤ect swapped foreignreserves for US government assets.

In either case, interventions do not have a very good track record ofworking:

No CB can o¤set the large �ows in the FOREX market each day.


Fixed Exchange Rates I

To �x the UK exchange rate, the government/CB agrees to buy orsell the currency at the �xed exchange rate.

Therefore the CB must be able to increase supply as necessary (printmoney);

Or it must have foreign reserves of other currencies if other countrieswant to sell £ s at the �xed rate.

This is a bit like the foreign exchange intervention material from before.

The di¤erence is that the CB is committed to buy or sell at the �xedrate...but if they run out of foreign reserves, they must break the peg.


Fixed Exchange Rates II - UK in ERM (see extra sheet)

In ERM, the UK was �xed at £ 1:DM2.95:

George Soros (LSE graduate!) thought that this rate was too high sohe attacked the currency.

Here are the main ideas (numbers just illustrative):1 He (and others) borrowed £ 10 bn, and swap it for DM at the �xed rate- thus getting DM29.5 bn;

2 Repeat this until the UK runs out of foreign reserves3 UK forced to "break the peg" - value of £ will fall (ex. rate falls to£ 1:DM2).

4 Soros can take the DM29.5bn and convert it back into £ at the newrate - £ 14.25 billion.

5 Pay back the borrowed £ 10 billion, with interest of say £ 1bn, and stillmake pro�ts of £ 3.25bn!

If UK did not break the peg - Soros loses the £ 1bn of interest.

Can always beat attack if raise interest rate above the level in Germany.


Session 6: Labour Markets

Michael McMahon


McMahon (University of Warwick) Labour 1 / 38

To Cover

Data and de�nitions

Labor demand, labor supply, and equilibrium employment

Frictional unemployment

Labor market policies: employment protection, unemploymentbene�ts, taxes

E¢ ciency wages

Unions

Monopoly and product market regulation


Role of the Labour Market

Labour is an important input to production �earns 65%-70% - of thetotal value added;

Labour income most important source of income for most people;

Ability of an economy to get people into jobs absolutely critical (thinkproduction function)...

...yet many economies with high structural unemployment and lowlabor-force participation.

The question to answer is why?


Labour Market De�nitions I

The Population ChartThe Population Chart

Population

People inthe workingage (16)

Young (and old)

Nonparticipants(education,institutionalised,sick, outsidelabour market)Participants;

Labour Force

Employed

Unemployed

Unemployment rate= Unemployed /Participants

Participation rate= participants /people in theworking age

Labour Force =participants = thoseable, willing &looking for work


Labour Market De�nitions II

Only around 50% population (297mn) contribute to output

August 2005; Source: Bureau of Labour Statistics

Unemployment rate = 5%

(100*7.5mn/(142mn+7.5mn)

Participation rate = 78%

(=100*(142mn+7.5mn)/(142mn+7.5mn+39.8mn))

142.1

7.539.8

36.8

107.7

Employed Unemployed Not participating <65 >65 <16, military, prison


Pre-Crisis Participation Rates

0

10

20

30

40

50

60

70

80

90

Wor

ld

Low

inco

me

Mid

dle

inco

me

Hig

h in

com

e

Fran

ce

Ger

man

y

Uni

ted

Kin

gdom

Uni

ted

Stat

es

Italy

Japa

n


Pre-Crisis Unemployment Across the OECD

0

2

4

6

8

10

12

14

16

18

20

Poland

Spain

France

Turkey

German

y

EURO 12 co

untrie

s

Finlan

d

Czech

Rep

ublic

Belgium Ita

ly

Canad

a

Portug

al

OECD Total

Hunga

ry

Austria

Australi

a

Denmark

United

States

Sweden

Netherl

ands

United

Kingdo

mJa

pan

Norway

Irelan

d

Luxembo

urg

New Zea

land

Switzerl

and

Repub

lic of

Kore

a

Icelan

d


Pre-Crisis Unemployment Across the non-OECD

0

5

10

15

20

25

Aze

rbai

jan

Gua

tem

ala

Cypr

us

Hon

dura

s

Mal

aysia

Chin

a

Sing

apor

e

Hun

gary

Slov

enia

El S

alva

dor

Costa

Ric

a

Mal

ta

Chile

Rom

ania

Russ

ian

Fede

ratio

n

Peru

Sri L

anka

Indo

nesia

Ecua

dor

Ukr

aine

Barb

ados

Isra

el

Mor

occo

Syria

n A

rab

Repu

blic

Geo

rgia

Iran

, Isl

amic

Rep

.

Puer

to R

ico

Croa

tia

Alb

ania

Ven

ezue

la, R

B

Bulg

aria

Slov

ak R

epub

lic

Uru

guay

Arg

entin

a


Other Factors Correlated with Unemployment Rates

The rates of unemployment do not just vary across country.

Within countries, many factors are important:

AgeEducationGender


Overnight Problem



Labour Demand ITo hire or not to hire

Depends on impact on pro�ts:

Hire additional worker to increase revenue:

Output increases = MPL (marginal product of labour)Output is sold at the price p) E¤ect on revenue is p �MPL

Increases costs:

wage (w) is paid

Decision:

If MPL > wp : then hire workers

If MPL = wp : do nothing

If MPL < wp �re workers


Labour Demand II

How does marginal product oflabour behave?

We assume it declines….

Increasing employment

MarginalProductLabour

RealWage

Profitmaximisinglabour


Labour Demand III

The Labour Demand Curve

RealWages

Increasing employment

From profit maximisingbehaviour get a negativerelationship betweenlabour demand and realwage

Labour Demand curve


Labour Supply I

Rising wages have two e¤ects on the amount that people are willing towork:

Income e¤ect:

Higher wage means people have more income and want to buy more ofall goods, including leisure.This e¤ect leads to lower employment

Substitution e¤ect:

Higher wage means price of leisure (lost income) is higher and soconsume less leisure.Tends to increase employment

End result depends on which of these dominates.


Labour Supply II

RealWage

Labour Supply

Substitution effectdominates andhigher wages meanincreased laboursupply...

Labour Supply

RealWage

Income effectdominates and laboursupply falls withhigher wages

RealWage

Labour SupplyIncome and substitution effect cancelout and labour supply constant

Three Examples...


Equilibrium

Employment

Labour Demand

Labour Supply

wage

rate

W*

E* Employment

Labour Demand

Labour Supply

wage

rate

W*

E*

So no unemployment?


Overnight Problem



Frictional Unemployment IOne answer is to note that the model actually does not say that thereis no unemploymentIt says the number of jobs available is equal to the number of willingworkers, but it does not say that jobs and workers have found eachotherThe labor market works by searching and matching and both take lotsof timeThere will always be some unemployment due to search and matchingfrictions

The original people associated with this idea are Mortensen andPissarides.We need to think about job �ows:

the rates at which workers lose their jobs (�ows into unemployment)the rates at which they �nd jobs (�ows out of unemployment)


Frictional Unemployment IIThe Flows

Flows from employment to unemployment (layo¤s):

Happens when companies scale down, close down, or when a particularemployee under-performs.To make things simple, we assume that it is a �xed fraction �d�of theemployed.d is known as the separation rate

)The �ow from employment to unemployment is d � E .


Frictional Unemployment IIIThe Flows

Flows from unemployment to employment (new hires):

A �xed fraction �p�of the unemployed receives a job o¤er every period.p is the job �nding rate.A fraction H of those o¤ered a job �nd the o¤er acceptable.H is the acceptance rate

)The �ow from unemployment to employment is H � p � U


Frictional Unemployment IV

UNEMPLOYED = U

EMPLOYED = E

Employedworkers thatloose theirjobs

Unemployedthat find newacceptablejobs

= d*E = H*p*U


Frictional Unemployment VSolving the Model

1 Rewrite the �ows into unemployment:

d � E = d � (F � U)

where F is the labor force.2 So now we have both �ows (in and out) in terms of U:

1 d � (F � U)2 H � p � U is �ow out . . .

3 We can plot these to �nd equilibrium (where �ows in = �ows out).


Frictional Unemployment VI

U

HpU (flow out)dFdU (flow in)dF

U=FU*U

HpU (flow out)dFdU (flow in)dF

U=FU*


Frictional Unemployment VII

U �F is sometimes called The Natural Rate of Unemployment:

it is the value of U�F at which the �ows in and out of employment

balance out in the frictional unemployment model

Frictional unemployment is:

increasing in the separation rate ddecreasing in the job-�nding rate pdecreasing in the acceptance rate H


Labour Market Regulation IEmployment Protection Legislation

This search and matching model is useful for thinking about employmentprotection legislation:

Two key determinants of unemployment: p and d .

Employment protection tries to reduce unemployment by lowering d -reducing separation rate

This will work so long as p does not change

But employment protection may also reduce p

�rms have to take account of expected �ring costs when they try andhire individualsthe larger the severance package, the less likely that �rms will hire newworkers

Therefore EPL should lower hire rate and the �ring rate but e¤ect onunemployment is unambiguous


Labour Market Regulation IIEmployment Protection Legislation

High EPL does reduceinflows intounemployment..

But also reducesoutflows fromunemployment


Labour Market Regulation IIIEmployment Protection Legislation

Net effect seems to be toslightly increaseunemployment rate

Perhaps a biggerproblem is that durationof unemploymentincreases with EPL


Labour Market Regulation IVEmployment Protection Legislation

The long unemployment spellsimply that many discouragedworkers abandon the labourforce altogether. This probablyexplain why we find strongerevidence that high EPL leads tolower employment rates (=employment / workingagepopulation)


Labour Market Regulation VEmployment Protection Legislation

Average job tenure increaseswith EPL


Labour Market Regulation VIEmployment Protection Legislation

Evidence not conclusive as we may wish. However, most economists tendto think:

EPL lowers both the separation rate d and the job-�nding rate p

As a consequence, it has little e¤ect on the unemployment rate

But it causes longer average unemployment spells

Leading to lower employment and labor force participation rates

BUT IMPORTANT TO REMEMBER CORRELATION IS NOT ALWAYSCAUSATION SO WE DO NOT HAVE �SCIENTIFIC PROOF�


E¢ ciency Wages IWe assumed workers are paid their marginal product

can only be done if employers can perfectly observe the marginalproduct of their employees.this is often not the case.

Furthermore, there are types of jobs in which �even if the marginalproduct was fully observable �employers might not wish to paycertain type of employees by their marginal products:

1 �Gift exchange� (the carrot):Employees that feel that their employers treat them well by payingthem a high salary will in exchange deliver high e¤ort and therefore beproductive.

2 �Shirking� (the stick):If e¤ort is hard to measure directly, paying a high salary is a way tomotivate the employees to put in high e¤ort: If you�re caught shirking,you will have to �nd another worse paid job.

) In such cases, it could be in employers best interest to pay employeesabove market clearing salaries.McMahon (University of Warwick) Labour 31 / 38

E¢ ciency Wages II

“Efficiency” Wages

Real salary

Effort

Effort Curve

Efficiency wage

S*Effort curve shows workers’ effort,initially increasing steeply with realwages; above a certain point, themotivational factor levels out.

The effort per dollar salary paid bythe firm is the slope of a line fromthe origin to any point on thecurve. At S*, the maximum effortper dollar paid is reached


E¢ ciency Wages IIIE¢ ciency wages can explain:

Why workers in jobs where e¤ort cannot continuously be monitoredstill supply e¤ort.

Why employers might be hesitant to lower wages even when demandfor their output diminishes

paying lower wages lowers e¤ort.it explains real wage rigidity.

Why there is involuntary unemployment:

Firms pay above market clearing wages.

What is important is that

1 salary is related to a motivational factor, an aspect that is clearlyrelevant in many types of jobs

2 pay is not directly performance related.


E¢ ciency Wages IV

Efficiency Wages and Involuntary Unemployment

Employment

Labour Demand

Labour Supply

Wage rate

E*

Wefficiencyunemployment

Efficiency wages lead toinvoluntary unemployment aslabour supply exceeds labourdemand at the efficiency wage


Labour Reform in GCC I

We already discussed the issue of labour markets as a motivation fordiversi�cation e¤ort.

The labour markets in the GCC still face a major challenge:

eliminate the segmentation of the labor force along skills, wages, andsector of employment lines, without impairing the economy�scompetitiveness, in particular of the private sector, which is expectedto create most of the new jobs in the coming years;

generate employment for nationals

increase the share of local workers in the private sector.


Labour Reform in GCC II

Current measures include a mix of administrative (quotas), market-based(fees and taxes for the use of expatriate labour) and supply-side(modernization of educational system) policies:

Improvement in vocational and training systems

Modernization of educational system

Creation of a national labor database to facilitate job search

Enforcement of quotas for nationals working in the private sector

Taxation of foreign workers (training tax)


Labour Reform in GCC III

UAE National Human Resource Development and EmploymentAuthority and a national labor market database.

Qatar No longer guarantees employment for Qatari graduates.Improved search and training available.

Saudi Arabia Human Resources Development Fund (HRDF) to train locallabour force (private sector partly funds it). Developed adatabase for matching and placement of local workers withprivate �rms.


END


Session 7: The Economics of Oil

Michael McMahon


McMahon (University of Warwick) Oil 1 / 10

To Cover

1 The basics of the economics of oil2 The impact of oil prices on oil-importing countries - AS-AD model3 Oil and the need for diversi�cation


The Basic Economics of Oil

ProblemOil, like any commodity, good or service, can be thought of in terms ofsimple supply and demand. Using a supply and demand framework,consider the following issues:

1 Is there anything di¤erent the oil supply curve? What happens to thesupply curve over time?

2 What happens to the price of oil when:

1 supplies are disrupted2 a newly industrializing country (like China) emerges demand changes


An AS-AD Analysis Of The E¤ects of Oil Shocks I

Y

P

SRAS

ADequilibrium

level ofincome

LRAS

Y*


An AS-AD Analysis Of The E¤ects of Oil Shocks I

Y

P


LRAS

Y*

AD

Y1

SRAS’’


An AS-AD Analysis Of The E¤ects of Oil Shocks II

Higher oil prices reduce output but increase prices

Monetary policy is ine¤ective

Fiscal policy is ine¤ective

If mismanaged:

1 deep recession with controlled prices; or2 stag�ation


Hamilton on �Causes and Consequences of the Oil Shockof 2007-08�

Eventually, the declines in income and house prices set mortgagedelinquency rates beyond a threshold at which the overallsolvency of the �nancial system itself came to be questioned, andthe modest recession of 2007:Q4-2008:Q3 turned into a ferociousdownturn in 2008:Q4. Whether we would have avoided thoseevents had the economy not gone into recession, or insteadwould have merely postponed them, is a matter of conjecture.Regardless of how we answer that question, the evidence to me ispersuasive that, had there been no oil shock, we would havedescribed the U.S. economy in 2007:Q4-2008:Q3 as growingslowly, but not in a recession.


Oil Taxation in Western Economies


Oil Taxation in Western Economies II

The "Pigou club" want to see higher US oil taxes

ProblemDiscuss the issues raised in the manifesto of the Pigou club:http: // gregmankiw. blogspot. com/ 2006/ 10/pigou-club-manifesto. html


http://gregmankiw.blogspot.com/2006/10/pigou-club-manifesto.html

http://gregmankiw.blogspot.com/2006/10/pigou-club-manifesto.html

Link from oil markets to the need for diversi�cation in theGCC

1 Volatile export commodity2 Finite natural resource3 Kyoto implementation


Impact of Kyoto Protocol Implementation on Oil-Producers

Costs depend on the exact way that proposals are implemented

OWEM model of OPEC Secretariat predicts cumulative losses over2002-2010:

No trade, weak price = �$56bnNo trade, reference price = �$19.1bnGlobal trade, reference price = �$9.4bn

The global movement toward cleaner fuels is likely to gather momentumand therefore oil-producing countries need to diversify while they havesome revenue still coming in.


Session 8: Customs and Currency Unions

Michael McMahon


McMahon (University of Warwick) Unions 1 / 22

Economic Unions I

There are a large number of economic unions around the world including:

European Union (EU)

South African Development Community (SADC)

South African Customs Union (SACU)

Arab-African Union (AAU)

Arab Cooperation Council (ACC)

Gulf Cooperation Council (GCC)North American Free Trade Area (NAFTA)

Organisation for Petroleum Exporting Countries (OPEC)


Economic Unions II

These unions di¤er in their level of integration.

Even the EU has only gradually become more integrated:

Level of integration Main features PeriodFree trade area Free trade among members 1958 to the early 1960sCustoms union Free trade with a common external tari¤ 1958 until 1993Common market Free mobility of factors across members 1993-1999Economic union Harmonization of economic policy 1993, 1999Monetary Union Common currency 1999; Notes in 2002Economic integration Completely uni�ed economic policy Not yet achieved


The Bene�ts of Economic Unions

Stop wars!

Better use of existing resources (allocation e¤ects)

Specialisation (comparative advantage)Price e¤ects (intensi�ed competition)Quantity e¤ects �IRTS

Better spatial distribution of existing resources (location e¤ects)

Accumulate more resources (growth e¤ects)

Strategic e¤ects

Greater negotiating power (WTO)Better Terms of Trade

Monetary Unions later.


Preferential Trade Agreements(PTA) Welfare Analysis

Assumptions

Small country in a small customs union (i.e. �at world supply curve);

P0 is the price in home (H) when no trade takes place.

Pp is the supply price of the partner country (P).

Pw is the supply price of the world.

H and P form a customs union with common external tari¤ t.

P0 � Pp � Pwt � (P0� Pw) () Pp � Pw and t cannot exceed P0� Pw).


Preferential Trade Agreements(PTA) Welfare Analysis

1 Before the customs union, the price that the good is sold at in thehome market is always:

P0 �!for goods produced at HomePp + t �!for goods produced in the Partner countryPw + t �!for goods produced in the rest of the world

2 After the customs union, the equivalent prices that the good is sold atin the home market is:

P0Pp (since the tari¤ no longer applies to goods from the partnercountry)Pw + t (the tari¤ still applies to these goods).


Preferential Trade Agreements(PTA) Welfare Analysis III


Preferential Trade Agreements(PTA) Welfare Analysis IVt < Pp � Pw )tari¤ does not close the e¢ ciency gap between Partner(P) and rest of world (W).


Preferential Trade Agreements(PTA) Welfare Analysis IVt > Pp � Pw )trade creation and trade diversion.


Preferential Trade Agreements(PTA) Welfare Analysis V

Preferential trade arrangements tend to lower domestic prices andraise imports

Discrimination produces supply switching

Welfare impact ambiguous for members of PTA if removing rentcreating barriers (the EU until mid 1970s)

Welfare impact positive for members when removing frictional barriers(recent EU)

Impact on excluded nations?

Very important to understand enlargement dynamic


E¤ects on Competition I

Trade reduces monopoly power:

International trade in goods

)increased competition in local marketsSmall domestic market supports few �rms in IRTS sectors

Small number of �rms un-competitive and charge higher prices

Competition from imports restrains monopoly power


E¤ects on Competition II

Integration intensi�es competition:

Reduces �rms monopoly power

They cut prices (just seen this)

Some �rms exit so that lower prices are o¤set by larger scale ofproduction

A pro-competitive e¤ect lowers prices

An economy of scale e¤ect reduces resource costs

Bottom line: Increased competition and restructuring could lead to bigwelfare gains, but two big policy issues arise:

1 How to ensure big �rms compete instead of collude2 Do states prevent the necessary adjustments?


E¤ects of State Aid and Collusion I

Pro�t losing �rms should leave the industry:

Bought outMergedBankrupt

Involves adjustments of �rms and workers

Government may give subsidies to loss making �rms to preventadjustment

EU calls these subsidies state aids

To protect the gains to the EU project, 1957 treaty of Rome bans StateAids that give unfair advantage.European Commission is in sole charge of enforcing prohibition (so nopolitics).


E¤ects of State Aid and Collusion II

Other justi�cations for state aid:

Protect mature industries (why?)

Promote sectors strategic for growth (e.g. high tech)

Special characteristics of disadvantaged sectors (e.g. agriculture)

Protect sectors important for national security and defence

To protect employment

Retaliation against other countries policies

Strategic pro�t-maximizing policies


E¤ects of State Aid and Collusion III

A similar argument applies to collusion:

If �rms don�t compete, then prices don�t come down

gains are lost.

European Commission polices competition between �rms too.

Both have an important supranational element


Location E¤ects

Economic geography concerns the geographic concentration ofeconomic activity.

In EU, people located in the core enjoy higher incomes and lowerunemployment rates

Industry less geographically concentrated in the EU than in US:

EU: 4 members have 50% man emp, 21% surface and 45% populationUS: 15 states have 50% man emp, 11% surface and 21% population

Income disparities across EU15 member states are much wider thanacross US states:

EU15: 25% pop in objective one regions (GRPpc<75% EU average)US: 2% pop in states with (GSPpc<75% US average)


Location E¤ects II

Economic geography concerns the geographic concentration ofeconomic activity.

In EU, people located in the core enjoy higher incomes and lowerunemployment rates

Industry less geographically concentrated in the EU than in US:

EU: 4 members have 50% man emp, 21% surface and 45% populationUS: 15 states have 50% man emp, 11% surface and 21% population

Income disparities across EU15 member states are much wider thanacross US states:

EU15: 25% pop in objective one regions (GRPpc<75% EU average)US: 2% pop in states with (GSPpc<75% US average)


Location E¤ects III

Traditional explanations

Specialisation patterns can be explained by countries having . . .

Di¤erent technologiesDi¤erent endowmentsDi¤erent geographies

. . . and thinking through the allocation e¤ects we have already talkedabout [These characteristics were implicit reason why costs and pricedi¤er across countries for allocation e¤ects]


Location E¤ects IIIAgglomeration models allow for factor mobility and can generateclustering of particular sectors or overall activities;New Economic Geography, one of the reasons Paul Krugman won theNobel Prize this year, focus on the role of �rm level scale economiesand transaction costs:

With IRS, prefer to build one plant;Bene�ts of locating in large market:

Cost linkagesDemand linkagese.g. from selling to consumers [Krugman] or input-output linkages[Venables]

Costs of locating in large marketProduct market competitionFactor market competition

Changing transport costs can change the balance of agglomeration anddispersion forces

product market competition from the other market increases astransport costs fall and so dispersion force less strong


Bene�ts Of A Common Currency

The main bene�ts of joining the single currency are:

1 the reduction in transactions cost of changing currency;2 the reduction of exchange risk leading to greater trade and foreigninvestment with the rest of Europe;

3 lower risk premium embodied in the cost of raising capital;4 increased transparency in price comparison;5 Monetary policy credibility gains for some countries.


Costs of A Common Currency

The main costs of joining the single currency are:

1 the di¢ culty of dealing with shocks without:

the use of independent interest rate;exchange rate movements;

2 Constrained �scal policy.


END


session 1: introduction - university of warwick in economics from trinity college dublin msc, mres...

Documents