• No results found

Commodity Price Changes are Concentrated at the End of the Cycle

N/A
N/A
Protected

Academic year: 2021

Share "Commodity Price Changes are Concentrated at the End of the Cycle"

Copied!
35
0
0

Loading.... (view fulltext now)

Full text

(1)

Commodity Price Changes are Concentrated at the End

of the Cycle

Stephen R. Ingram

Business School

The University of Western Australia

Determinants and Impact of Commodity Price Dynamics M¨unster, Germany

(2)

Introduction

Introduces a new approach to the cyclical analysis of world commodity prices.

Begin by determining that natural cycles are asymmetric. Use this result to show that large price changes can be used as leading indicators for approaching turning points.

(3)

Introduction

Natural cycles reflect long run movements in prices.

Definition

Booms/slumps are defined as periods of generally rising/falling prices.

Examples

Caused by technological progress, long run demand factors, as well as the build up of excess reserves.

(4)

Introduction

Growth cycles reflect short term movements in prices.

Definition

Here periods of rapidly rising/falling prices are referred to as growth spurts/plunges.

Examples

Caused by supply shocks (weather and geopolitical events) and rapid demand increases.

(5)

Data

28 commodities - 19 renewables and 8 non-renewables. Deflated by the 1990 base weighted MUV index.

(6)

Commodity Price Theory - Goods or Assets?

Homogeneous - traded on competitive markets. Low weight to value ratios - easy for trade

Traded on large liquid futures markets - therefore have (more or less) a single world price.

(7)
(8)

Why do Commodity Prices Cycle?

Food, beverages, and raw materials:

– Trended downwards owing to efficiency gains.

– Interrupted by strong demand for food and fuel.

– Given inelastic short-term supply schedules, sudden shifts in demand cause considerable price movements.

(9)

Why do Commodity Prices Cycle?

Energy, metals and minerals::

– Modest downward trend owing to the constant discovery of low cost deposits and efficiency gains by the FSU.

– Trend reversed by high demand, exacerbated by low investment during the 1990s, bottle necks and input cost inflation.

– New supply has considerable effects due to inelastic demand curves.

(10)

Natural Cycles

Identified by the non-parametric BB algorithm as opposed to parametric STS models.

Each phase is dissected to show their length (duration), net growth (amplitude), and trajectory (cumulative growth). This is achieved by conceptualising each phase as a right-angled triangle.

(11)

Natural Cycles

(12)

Natural Cycles

(13)

Natural Cycles

Stylised slump phases:

(14)

Natural Cycles

Stylised boom phases:

(15)

Natural Cycles

The excess index:

Eji = F i j −AREAij Ti j i = 1, . . . ,nj j = 0,1. Where: AREAij = T i j ×Aij 2

(16)
(17)
(18)

Growth Cycles

Identified by using a Hamiltonian Markov-Switching state space model.

Commodity prices are generated by:

yt =zt+nt

where

zt=φ1zt−1+φ2zt−2+wt Var(wt) =σ2w

(19)

Growth Cycles

The first component of the state space model employs a first order vector autoregression: xt= Φxt−1+wt Var(wt) =Q     zt zt−1 α0 α1     =     φ1 φ2 0 0 1 0 0 0 0 0 1 0 0 0 0 1         zt−1 zt−0 α0 α1     +     w1t 0 0 0    

(20)

Growth Cycles

The second component, the observation equation, estimates the first component:

(21)

Growth Cycles

Based on which conditional probability is greatest, At will take the

form of eitherM0 or M1: At M1= 1 −1 1 1 yt = M1xt+vt Growth spurt Growth plunge

(22)
(23)

Do Large Price Shocks Help Predict Future Turning

Points?

Growth phases range in size (amplitude), we expect that extreme growth spurts end close to the peak and that extreme growth plunges end close to the trough.

(24)

Do Large Price Shocks Help Predict Future Turning

Points?

Each commodity will experiencen1 peaks. An extreme growth spurt

is defined as the nth1 largest growth spurt.

Oil experiences 25 growth spurts and 10 peaks. The amplitude of the 10th largest growth spurt is 0.28 (28%).

(25)

Do Large Price Shocks Help Predict Future Turning

Points?

(26)

Do Large Price Shocks Help Predict Future Turning

Points?

Mean = 25.7% Min = 12% Max = 39%

(27)

Do Large Price Shocks Help Predict Future Turning

Points?

Each commodity will experiencen0 troughs. An extreme growth

plunge is defined as the amplitude of thenth0 largest growth plunge (absolute value).

Oil experiences 26 growth plunges and 10 troughs. The absolute value of oil’s 10th largest growth plunge amplitude is 0.2 (20%).

(28)

Do Large Price Shocks Help Predict Future Turning

Points?

(29)

Do Large Price Shocks Help Predict Future Turning

Points?

Mean = 26.3% Min = 9% Max = 77%

(30)

Do Large Price Shocks Help Predict Future Turning

Points?

Booms and slumps can persist for up to 10 years. They normally last 3 to 4 years.

If we observe a large price increase (decrease) of these magnitudes, then how far away is the peak (trough)?

(31)

Do Large Price Shocks Help Predict Future Turning

Points?

(32)

Do Large Price Shocks Help Predict Future Turning

Points?

(33)

Future Research

Are there common factors driving the boom/bust cycle?

Is this information useful for developing a model which explains price behaviour?

(34)

Conclusion

Stylised facts help us understand price dynamics.

Warn governments when to cut spending and inform producers about likely future movements.

The Western Australian government derives approximately $8000 per capita from mining royalties.

(35)

Questions

References

Related documents

The aim of the study was to examine the relationship between child rearing practices and adolescents’ attitudes towards sexual debut and to see which among

For high NG retail prices, micro-CHP is not a competitive technology and the system prefers conventional heating units for supplying heat requirements (although if

Meaning that Tourism sector (number of tourism arrival (NTA), Expend on tourism sector (ETS), agriculture (AGR), Transport on tourism sector (TPT), Exchange rate (EXR),

Brandwein JM, Atenafu EG, Schuh AC, Yee KWL, Schimmer AD, Gupta V, Minden MD: Predictors of outcome in adults with BCR-ABL negative acute lymphoblastic leukemia treated with

Use the national framework for school counseling: ASCA National Model for School Counseling which includes: School Counseling Program Assessment, School Counselor

The size of the droplets observed in the ESEM suggest that mushroom spores act as giant cloud condensation nuclei, aiding the coalescence of smaller droplets to form