...

PDF

by user

on
Category: Documents
1

views

Report

Comments

Description

Transcript

PDF
\Vool Futures and Type :Premium or Discount Risk
C. .}. Cunningham
'Vool International
Australian Agricultural Economics Scocicty
Annual Conference
14 - 16 February 1995
Perth
Address for cor.respondtmce:
C. J~ Cunningham, 'Vool Interriation;ll, GPO :Pox 4867, 'Melbourne~ Victoria
3001, Australia
Chris Cunningham
Wool International
ABSTRACT
Since the removal of the Reserve Pri.ce Schem~ in 1991, int~x:est has· rlsen in
price risk management using wool futures. The abllity to shift .pclce risk :is
affected by price relanonships bet\veen different wool ttypes' in. the .spot
market, particularly the cl1ange in type premiums and discounts ove~ time. The
incentive to hedge will only exist if type premiums. and .disc~oUilts retnairt
predictable and stable over time regardless ofwhether prices rise or fall.
A problem facing hedgers using the current 22 micron wool futures contract is
that premiun1S or discounts between the 22 micron indicator and the price of a
particular wool type may not be predictable or stable. Cointegration analysis
has been used to show that ch(Ulges in type premiums and discounts relative, to
the 22 micron indicator are asymn1etrical. That is they d~pend on· the diregtion
ln which the value of the, 22 micron indicator moves over tin1e. Th~ asymmetry
of type premiums and discounts generates type premium a.nd discount risk and
necessitates the use of ratio hedging when attempting to achieve more cerb.lh
hedge outcomes.
Key '\Vords: Wool futures
Price risk
Cointegra.tlon
Type premiums and
cH~courtts
lntr(:ujuction
A commodity futures market which functions :proper~y improves theprocess.of;prlce
discovery, facilitates the management of iiWentorle$ through time, enables som¢/Qfth~
business risk associated with price volatility to, be shifted, and provides. ,btmefits
throughout the marketing chain. Since 1960; there has been a wool fl.ltutes mfll"l<etm
Australia conducted by the Sydney Futures Exchange. De$pite some ·short :bllrsts of
activity the wool· futures market 'has mostly been little
used.
The ·existence of. thQ
Australian Wool Corporation and the operation of the Reserve :Price :S¢hemc (RP$)
between the early 1970s and mid~1991 did not providethe.conditicms conducive: to
the
development or a robust market in wool futures} Overall, in the :past, th~ yph,Irn¢ ·of
trade on the wool futures market in Australia has been too small for.fntur~s to be an
attractive risk~shifting option for most woolgrowcrs, processors .and. speGulaton;. ·An
active wool futures market will only develop and op~rat~ su~cC$sful)y. ·if ·pti9e
fluctuations in the wool market make hedging desirable~ if it prm!ides ~ppotttmities for
speculators to profit, and if the specifications of the contract make rl~k"shifting ·
possible.
The paper is organised as follows. First, the importance is discussed :ror ·the effectiv~·
hedging of Australian wool of the price relationship between. the indiciltpi· of ~me
micron category used in the futuJ;eS contract and. that of other.· micron 'cat~gotl~S;
Second, preliminary estimates of the relationships. which exist~db¢t'Neen·tl}~ .~ricesxif
different. micron categories of wool over the period:
coint~gration
1
";f.~:f•l·""''l ::1~
analysis.
22 .micron inQ.icator, Qn which the .futgr.es ·contract
cat~goty incUcators ·over,t:his:pe~rio.date. itest:edl:anc:t··f<)Unld•:.t&·~,toclucQ;:·asYrOOJi~ltY,.jtf':~f.i~
Australian ttpparel 'WOOl is,t\'vecy . h~te.tdg~g¢(:>g$ ~proq~ph.·· i?rl.or<to :§ttl~ wooi,:is::~tt¢d . ·.
into types according to
strength~ vegetuble
chtu"~c~¢rlsd.c$ ~s\l.clt . · ~·~· '~bl'Q :iditW:l~ter1 . ftbre' ,l~h$~,; .rtb.¢
matter content. and c919Pr. ·Fib~
d;.amet~r Js:
the mosfful.t?r:n:tanf
detenninant·of prlce. Most wbol.proqu¢¢(1 is \PQUght l,lncl so.lQ 4t ~uctiori l\nCl:prlc:,e~t
can flttctuate markedly. The price, risks ·may :be shi~tcci: by Jrad.ing :i.rJ ·woot. fqntt~~
contracts which, over the period. of this amllysis. wet·e: ~p(!cifi¢d;lintetrns cif. a 22 '.micrpn
(p) indicator price comprising a weighted l\vera:ge .of~le\!.en,. Z2 . micron 'Wool {JJ?e$.~J
Hedging relies on the fundamental relationship between 'fut(jte,s 'ptlces and. ,ca~n 'prices
in the underlying commodity market. Thar is the tendency that ·~s contntcl roaturl~y
approaches the
Ca$h
price of conttaot.,specific wool and the futqre~. price ·convet~e•.
\Vhen hedging contract~specific wool; converge~nce w.ill oc;otJrbyd~tinintion,hPW¥ver~
convergence to zero isless likely or may not occur when ·heqging .ncnt~~pnt,tactf~p¢otrio
woot
the optimal futures position for a risk~averse competitive producer d'e,pends :.on a,
npmberofinterrelatedvariables, e.g·. the nature.ofthe ptice~ basis tmd:pro4ilcnon·tis}<S
(Lapan illld Moschlni 1994). In this :paper, production .risk is 1gnore9. :a~r~ :fislc
occurs whethet ot not .delivery is made on the futures c;ontract. T:he :s~t«f;went:prlc¢
on the. futures contract ·js imperfectly correlated with the cash :price ~n fh¢
titne.(;)f~s~l¢,
a sale 'Which is made necessary .because it is usually undesirable ;or infeasib1¢Jfdr· :(l
producer to deliver the. product in order to setae the !utu.res.cc:mtnl.¢~~
by the ·producer :and that sp~cifieclln the futures' ¢OIIltl.'~~ca.·~eJli1t!¢t¢11t; ..~~.·m,ng.7J'.~.
th~ fu.tt1res
con.r.ract price at time t'Iot:(j¢!ive.cy at 1[]Jile, ,r,rJ,.;T,+i
. at time t+.l, .aJ\ilPr+l be tli~:qJ~~h pr'ic~'tit·. Pffi~.
.~pe~ificl!tion of·. the .fut\lre$. ·~ontrac~,
~.~'~n<~:,:se~tm~m.(m~~p[!tp~
'
.
category, there is an additional,·pot~ntittl source of~risk, ~mm~ly, th~ pos~ibili~y~of'tll~
imperfect correlation 'bt!tween the 22. micron indicator (p22) and th~'\casb 'pncQ t>fother
wool types (Jn).
Th~refore,
the total price risk facing growers of .coJegqcy I wool
comprises basis risk and what Will be referred to in Jhe rei111llnCier ofth¢;:p~j>et· as.<type
premium (pf > P-22) ortypedlscoun~ (p; <]>22) risk~ llenc<h.the oastscfor Wookcategql)'
i at contract maturity '(time t+l} can be prtrtitiort~d as:
To hedge. non-contract ~ecifip· wool eff~ti,vely usmg th~ 22 mi<::ton. futur~s:. c~qtt~gt'
the premim.'.l or discount between the 22 .micron indicator .and, .the ca.~h ;pdc~:of .th~
cype of wool that they wish to heqg~, i*e~ fn2~,t+J - Pl,t+J), myst be precii~l~ble :@d:.the
relationship between them ~table over ~he conttact period-. This t(}l~d,onship :n:u.t&t
remain prt!dictable .and stable rega,rdless o.f whethev prices n~e or fall QY~r' th~. :lif~ .of
the: contract if there is to exist an ·l'''!entiveto he<:}ge,. ·oth~rwise there1s!1anotlJ~r$P1JfC~
of :rlslc for· the grower to content: ·,,-tth; UltixnAtt!lY. it 1s the combined m9vt!medts :in
the. t,ype premium or discount and the basis over time which detemun~s ;fhe '91lt¢p1lle·Qt
the.hedge.
The price of each type of wool is determined at: auctic,m, f;Utq; the :r~l4tiort$llip'·P¢tw¢·~n
the 22 mipron indicator and '~he ~price of anoth~r wool tYp~ may :rtot::ne,~es$Mi1y;.ti!m~.m
:stilble over drne~ !he. demand nnci suppJyoftiiffetentwool~t1p~;:y~,c;Iwl.u~~tlle:y¢~~
and" hencei type premiunJs. 'or ~discourlts .m-e .no:t: :Uke,Iy to: ~m~rii·con$p.lilf~Jor fli~. Ut~ 9fl
a 'heqge.
Some reqent e~pirlcal evidence $i~yr~ l:} ,U14$.tlJites, J~1lt, :£he ;~clc~'
1;3QO
P M
~1192
s
J
an& thC? price ()t a wool type ch~ng~s :ov~r ;(b~:J.ife pf tt1}~. ·¢9~i•P'·~Q~j, lb~~'~~v~~1A!~ ·~~~Jl
:po!;~ti<>n
:of toe,:
h~(lge,r l~
e,itne,r
r¢I~.tigns~p;Nm~t)eq;·$.taPl~.,
~tte,1;
.Qr
wotse. :..th.an ,iti'·Wi:i~1fl:tll.:'l1l~v~t.:JX~~n~·Ih.~K!;:;!t!t~
particular Wt)Ol typt!S re1~dve 10 that of the :~rrad~sp~cified ~n/th~; 'fl.lt4r~s, tonttact., illli~'
more· strongly and positiv~h' c;o~'lnr~d. a,r,~ th~ :pric¢s. of Jh~· tlJffe!:~nt t}'p~$ ;of' wool,
3Q0
itHUUI iillii 411111111 OliHHi.IHIIi-ttiiJUJI UHH I I IJHIJIKIHI Pl>lH)
S
0
9119~
M
s
J
o
9,2~3
M
Th¥re ate essenti;tlty three. stt;p$ req\llred to test the hypothesi$. that t11e pdce ln(licator
of MY ca~~gocy of wool followed: a long-ro.nt eq~tilibrh;m rel~donship wi~h the .~2p
indJoatort i.e,. a coimegradon teladonshi,p..
In the frrst s~~p, the order of: inte~nn!on>of
each of the indic~tor series bas t.o be deteonine(l, as $erles whiqh we· nPt of tfle;. :s~ii~
order of in~~ID1ltion c~nnot ~ coiutegrijt~q.
l11
the second st~p~. and assqn1.lnJ~ th'lt.th~
lndiontor or the care.gozy of wool of jnterest un<J th.e iudJgqt<Jr oF '42p WQ¢l
·+tt~·
tm~wat~d ·of ord~r onQ~ then an ordinary least .squiU'es .(O}JS) ~gNSsiQo c~n ~
estlP:Jat¢4 \lSing the ingiQ~lQt for 4~).1 WOQl ~$ .thv :ind~pertpent VaJi.able: apq. Jh¢,
indi.cato.r of the cQqtegqcy ofimerest .~~ the d,¢p~odent vtUi~.blel :.Ih thr; ·mird $t~n,. :it bAA:
to be shown lh~ t the :reslouais frQm this ·wgre$Sion. are intctr~tc(i: ·otpr(l~r: ~e.to.~ ·ft.lr ·th~
to be vaUtl~ he~ that th~ 'tWO ;jrt(ijp~tQf: :setle$ form: «Jon~fPID; ·
~uUibrlum:reladonship~ lf i.t is qpnqlud,~<l that: th~ ·indtcatof<-:fQr ~. J?~qq1al!:¢.~t~a9t.:Y ·
PointeiD'~tion r.<:}gres$jqn
of wool ~nd th~ i.ndiQator of:~2p Wool ·Were cohft~~a.teq~. thert• Uli~ WoP!~ nfe~p th4~ .
.
\
-
.
thti matk,et .con<U~on$ were mot~ ~ongg¢~v~ to succ¢s~fyl '1\~~m~ :9f\~H~t. :~.~f~~gfj;:¢f
WOOl :thttn WOQJq; have: been $,e. ·c,\Se.li.\ tl\e, ~bs,eilce, :pf ~ijch .a}ft?l~t\oti~.hi~.
The ~l91\~y.-JJqtl,.¢rl~$t·Wa$· U$¢g: l<l ide.ntif~~ me. <;>tg~r of .int¢~~tloh of ~~9hdn9l¢~l~t·
$~ri¢.$. Jilld lhtf r~s.uh~ qr~ ~\'t1o:
·m tt\bl~ 1,
It' :i~ ¢on9lu.C,\~~ lht\t:a,llh~: ihdi9~t9r· §~d¢s~
whh;t.he; ~~~~ption c;>(·tbos~for c~'~gorie&l9JI4t)d ZOp~ W~t(} ro.r}(l:) ~nd,.thert!fqre,.tijat
e~chcotild>.fomt.~·t:Oin~~gration·relt!~i9rrsllipwilh··~h~··~~HihdiGatm-.~~ti~s~
'
.,.,,,,._.,,,.. ,,.,,_,,,,
....... _..
_,"'""""'"-"
,.,
lllt~gratign; ofp;,r
t9
..Q~Ol~404
. ~. a:l
20
21
,;Q.Q0$\909
.,Qi006S20
.. }.5)9
.. J.41
?4
--0.004038
. otss
23
.,Q.0024S2
.,Q.49
24
25
,.0.27
~()
-0.001200
-0;000899
-0.0006'31
27
.,(Ml003S3
.-Q;Q9.
-0.23
..Q.l'l
28
-.0.000184·
,.o.os·
29
.,o~.ooo~9l
-0.08
30
,.Q,(>00489
-,Q.1.4
-1.0404
".8.76
~1.0400
.,9;02
-1.1309
. .Q.79
.-9,30
A!J.p;,t
1:9
40
41
22
23
24
"tQ819
,.7;6~
26
..Q.8936.
-.JiQ19$
.. e~72
~7
,.Q.9l04
,.~h87
28
.. o~8~4S
-:-7.48
;2.9
.,o~so·~o
~6,9q
30
.:Q~~59t·.
. ~~',@<
25
J(Q)
t(Q)
..
NoJet Wh~ c::rific"lY~\lef?r;·~ic~~y;F~U~r,~~J~t:4ti9'ta:t;>{S~?·· n~$0)'.$ 4~Q?~:~q
~t.·(i %, n;;SO) i~,o:4:itl~· ().;qll{~r,l~7~).
wherep;.r and P~2,.t are each ~l(l) setie,s. i. :;~1. 2a~ 24, ... ~. $0~ Th¢:rcS\.llts tfQn'l;fu~~~:
regressions are, give.n. in Table ~. From the
t~.stt\tbtics, it i$.
qon¢h1d¢<:l :tht\t. '.ther~ .i$ ~
positive and significMt relationsh!p beiWe<m e~ch 'mii~atot series and. the, i:rtdipatorJor
22p wool. In all eqtmtions the nu.ll hypothe,~js that q,1;::; o i$ rejected;
Eow~ver, tQr
these equations to be 111eaningful and to be cointegtation relationships, the r~~!c:lY~$
must be ~1(0).
19
20
21
91,69
0.890
(~.S4)
.76
(1S.S9)~
(-1.84)b
23
-$.20
0•.948
(0.29)
(29.75)
(..l.6S)
24
20.55
0~866
(0.84)
(l9.9l)
.84
(.-3.09)
25
26.65
0,824
(l8.S2)
(".3,.9q)
(1,06)
26
69.12
0\722
(17.47)
(2.96)
.80
(-6.74)
27
104.77
0.636
(14.59)
(·8.S7)
(4.25)
28
108.59
(4.14)
0.6~9
(.13.33)
(,-8.22)
29
o.s.so
130.26
(4.86)
(11.61)
(-9.49)
30
175.13
0.435
(7.27)
(}0.21)
(-rl3.3)
Note:
a- The t-statisdc under the null hypoth~Si$ th~t ctp. 0
b,. The t,.statistic uncler th~ nqU hypothesjs tb~t <lt =1·
10
.64
Th~ Pickey.-Fullertest was qs~clto detetm.ine ~he order ofintegration.·ofthetesi4uals
from each r~gr~sslon~ Tile results p.r~ ~Yen :iti T~p}Q
s,
....
~~~
··~
Ord~rof
integnnton of
Ut~t:
19
20
21
..0.08776
-3.24
I(l)
23
-O.l70l9
-4.12
l(Q)
24
.0.19072
4.85
1(0)
25
·0.19284
4.20
I(Q)
26
-0~ 17232
-4.23
1(0)
27
·0.12879
~3.00
I(l)
28
.. Q.ll559
--2.77
1(1)
29
-0.09501
-2.37
l{l)
30
-0.09370
-2~31
l(l)
Note: Critical Value for (5%, .n=50) is -3.671 for (l0%t n::;SQ) is -3.28, from
Engle a.nd Yoo 1987.
It is concluded that the relationship between P2/,t
and P22,t was not a coit\tegration
relationship at the 5% or 10% levels of significance. However, th~ power ofthis test is
not high tmd it may~ rea.sonal:>le to reject I:Io, as the proba.bUhy vuhH~ is just in excess
of 0.1. For the remaining grp,des, a cointe~ation relationship appenrs to have existed
for the grades 23p to 26p inclusi:ve bu~ not for grades 27p to 3Qp fnclusiv~.
The irnplioations of these results for the chances .pf hrwing ~mcc~ssfqlly hec:l~ed wool
over the period AugtJst l.991 to April 199~ Qre mix.~. Ptod1,1cers of th~ fin~st wool,
19)1
and zo~, apd proqucers of the coarser wool$, 2,7p to
11
30p jmdysive, would
not
have been able to hedge successfully using a.routine risk..-spreading strategy because· of
the lttc.k of n11y cointegrntion
relation~hip
between these categories of \vool and the
22p category. Greater volatility was observed; in the indicator series for the
finest
wools than that observed in the 22p indicator series and less volatility was obsexved in
the i1ldicator series for the coarser wools. On the other hand; producers of wool cypes
which are closer in quality to the 22p category, namely 21p, and 23p to 26p inclusive,
would have been in a better position to hedge because of the cointegraticm relationship
which existed bet\\'een each of these categories and the indicator for 22p wool. The
existence of these relationship implies reduced type premium or discount risk, when
hedging some wool types within these categories because the relationships were .stable
and the type discounts predictable.
Type Premiums or Discounts and tile Cointegration Equations
The long-run linear relationship which existed beD.veen the 22 micron indicr\tdr and
each of the non-standard micron indicamrs during the post-Rl>S period, are given in
Table 2. The coefficient, a 1 , in each of these cointegration equations,. represents the
slope of the linear relationship between the 22p indicator and the relevant micmn
indicator and represents the change in the micron indicator of wool category i per unit
change in the 22p indicator. For the type premium or discount to remain constant
through the contract period, this slope parameter must be unity. From the results
shown in Table 2, it is apparent in each regression that the null hypothesis, a 1
:::
l, is
rejected at a level ofsignificance of 0.05 for grades 24p to 30fl inclusive and at a ]evel
of significance of0.1 for all grades. In each regression, the. alternative hypothesis that
a 1 < 1 would be accepted at the 0.1 level of significance.
The further from unity is a 1 , the larger the anticipated, change in the type ~premiurn or
discount for a .given change in. the 22 micron indicator..
Accomln~.tP
these equatipns,
the type discount will widen (narrow) for a given increase (dccren:se)"in.th¢, 22 Piicmn
indicator. Hence, the characteristics of the type: di~counts were asymtrtemcal.'m that
12
they depended on the direction in ·which the 22 micron indicator chang~d, ·with the
degree to which changes in the type discounts occurred dependent: on the size of a. 1 in
relation to unity. For example, the andcipated. change in·the type discounts woulp
~
larger for 26 micron wool than for 23 micron wool, because, from Table 2, a 1is 0~7:t.,L
for the fanner but 0.948 for the latter, implying a, smaller correlation between 26p and
22p wool tha.n between 23p and 22p wooL
Before accepting the asymmetry in the type discoUtlt, it is necessary to establish that
the slopes of the cointegration relations hi os shown in Table 2 are the same for ·rising as·
for falling values of the 22
·nllC..'TOn
indicator,
otherwi~e,
the asymmetry may be
spurious. An F-test was used to detem'line simultaneottsly whether the intercept and.
the slope of the relationship between changes in the 22 .and 24 micron indicators were
the same when price changes were positive as when they were negative. Three
regression equations were estimated by OLS using the stationary series 6.p22 as the
explanatory variable and 6.p24 as the dependent variable. The unrestricted model
comprised two equations: one for those observations for which AP22 was positive;
and the second for those periods when liP22 was negative. The restricted model
imposed the equality of intercepts and the equality of slopes on both partitions of the
data. These restrictions represented the null hypothesis. The resultin:g F~s~tistic was
0.62 while the critical value,
F*(2,71)
was approximately 3.14.
restrictions were acceptable, implying that the slope parameter was the
therefore, the
s~e
for price
increases and price decreases and that there is an the asymmetry in the type di~count.
Hedging with an Asymmctrical1:;pe J)iscount
The size of the slope parameter in equation (1) relative to unity has impllcationsfor the
design of hedging strategies.. When shon :he~ging 23, 24, .25 or 26 micron wool, dle
realisation of an .expected hedge, outcome d~pends :rtot only ,()n the behaViour of tlie
basis. and the asymmetrical type discount butalso on tile. ai11Qunt. of wool ,being hedg~d
with each contract.
13
Consider a shon h¢4ge in whi9h cpntrilcts :are sold
such thAt Ate .afuount .ofWool
representoo by the contracts. is equlvalentto the .amount of: wool.(23p' to 2~p): :c(q 'be
hedged. That is the ratio of the ~otal :amC>unt. of Wool ~p~sentecl PY the con tracts to
the amount of physical Wool bein!the4ged is ltl. As the slope ·partrmeter
tt 1
rot all
these categories is other than unity (a5ymmetrical.cype discount), the outcpme .of.such
a hedge is uncertain .and dependent on the. direction in which prices move. thls is
because, if the 22 micron indicator falls Md the type discount narrows, a :more
favourable outcome occurs than if the slope :PaTI\meter was unity (constant type
discount).. That is the f\1tures profit would more .than offset the loss· in the spot
market. However, if the 22 micron indicator rises .and,. the type discount widens.i then a
less favourable outcon1e would occur as the gain in the spot market would JlOt cover
the futures loss. In both situations, i.e. when the 22 micron indicator faUs and when it
rises, the non-constancy of the type discount increases the futures gain and loss relative
to the change in value of the physical, respectively and repuces the certainty of the
hedge
outcome~
For risk-averse producers, such outcomes are to be avoided and,
hence, a different hedging strategy is suggested.
To hedge effectively, net losses (gains) .in the future& rrmrket must be· offset l?Y ·net
gains (losses) in physical wool. To improve..hedge effectiveness and the certainty of
outcome, a ratio hedging strategy which acco"Qnts for the asymmetrY oftype·:premi,ums
and discounts is more appropriate. By altering the amount ot wool1:>ei.ng hed~ed by
each contract~ .more effective. and certain hedge outcomes :ti:lllY be achieved. The slope
parameter a 1 .of the co integration relationship benveen the indicator
of the type of
wool being hedged and the 22 micron indicator, can be used to<dctern:iine the· .ap.)ount
ofeach micron category of wool which can be more effec;tiveiy :he(lged.p~t:: contr~¢.t~
For example When hcdgir1g 23: to 36 micron ·wool, ~ mo~ c~rtain h~d~y·Outpofl)¢;Il14Y'
be achieved· by .holding (l/f1. 1 )*
2.SOO·JdJograms of' wQ()l . fot
~v~~
qp¢rt ¢.ontr~ct"
1Jhat is the ratio of wool represented by each contract tq ~e atnQUIU.ro(.J?~y~l6W,\"woP;l
being hecigeclsho\lld:l)e 1 .: (l/<X.1 ).
14
Assumil1g the change. in the oasis is :zero, i.e Xjj .. P221t) :::: lfl+l;, P.?~.t+l),\ the·n ,1f;fhe
cointegration relationship holds~ regardless of \Vbether pdcestise :t>tfallth~; 'ne,t·chafi~~
ill the futures position will be more closely offset by the net chat)ge in the value ofthti
basket of physical wool.
Exatrml¢ .of .ratio hedgjng
Futures
Mass of
contract
wool
Chru-agein
22 micron
indicator
(cents/kg)
(kg)
2,500
ln this
X:
~"<ample
(cents)
2,500x
Netchange. irt
value of
Clmng~in
Net change in Amount of
wool being
f\ltttres
position
hec.iged
value. of
hcdgedw(){)l
(cents/kg)
(kg)
(1/q l )*2,500
q J
:Physical W()Ql
(cents)
*x
2~.500x
the change in the futures price and the 22 micron .inciicator durlngf the
life of the hedge was x cents/kg.
Consistent with the .e>;istjng
coint~~tion
reladonship, the chMge in the value of the physical wool was a 1 *x c~nts/k~. As t.he
(;bange in the price of the physical wool
llt *x
was
~pread
over (1/ ~ 1)*2t500 kg of
wool, the change in the total value of the physic~, wool was i,SOO~~ This is offset bY
the opposite change! in the futures posiuon resulting in. a cet:tairt he,d,ge oUtcome
regardless
or whether ·prices rise or fall.
lit practise? the certainty of hedge outcomes
\Vill also be reduced by changes in the basis during> me. life of .the. h~$e ~ cas~ alld~
fUtures prices convet:ge. Despite this, more certain outcotm~s. ®ty ~~ acllievf!d. whe·n
hedging
non-contract~specific
wool by using ratio .he4gjng sttategi¢$.
relationships like those estimated in this. pa.per.
based~
on price
The outct>rne, of·~. hedge. using the Sy(f~e,y :Ftitlll'e s· .:t;!x:chang¢ 22. tn.ieron w9ol fut\1tes ·
conttaqt bn$ bt!en shown to dep~ng. in ;pntt on th~ chtU"acteris.tics .gfth~ l>'Pe ;p~rmutn
<>rdiscoQnUiuring the ]Jfe ofthe hedge.
It WtlS established tha( type pteminnl$ a1ld' ·aiscoums are .$ymmetricru .:in that :th~y
dep~od: On the direction in which the.value.ofthe· ~2 micron, ·lndicatorchanges~
Cointegradon rt1lationships existed between the 22 micron incllcator and the 2..3 t9 2.6
micron indicators during the period of this anal~si$. These rellitionsni,ps were sta,ble
and the type diseounts predictable.
The exbtence of these relntionships would have jrnproved tl}e potential to effectively
hedge 23 .to 26mieron wool using the 22 micron·.contract? while,hec::lgtxJg()tfiner·and
coarser wool would not have been as successful.
lhe asymmetry of the anticipated. ·type premiums and discountS: :genetat.es .typ~
premium ll1ld di~ccmnt risk and necessit{ltes the use.ofratio hedgif\g. When atte01pting·
to achieve more certrun hedge outcomes,
OicJ.:eY~ D. A. and Flillt!t\ ·W. A~. (l~Sl)i ,'~IDhe
Llk¢Hhoo.n ·E.attC>~ ·:Si~U$ti¢s' !or·:
Autot¢~~sive ~ime s·encs··WiiD···.~·Uilit,.rt{Q<>l',,,,£a~n~~~h·itq~ 49.~ ·•td57->l07.2~···.
Engle, R. F. and \Yoo,
a. s. (l987),
•f1o¢.¢~.$tirtg .and :I:¢~tiilg ln ·~o·.integratc;d;'
Systems'~ Jotloaqt off!cottomctric~~. S$; ~l/4$';t59,
Onmg~r~
c.
\V~
J. 0986), 'l)evelopnlenLc; in tnt} Sttidy of :cointegtAJ~Q
Van~bles:',
OxfordBullctfnoJEconolnrcs antlStati~·ttc~·, 48: 2J.a ..4.ZS~
Lap~n,.
B.. and Moschinit
G~ (1294), ~FuturQs J:led.~lng
.under .:l?Qte~
Jlttsi~.tmd
PtoductionRisk', American. Jollrnal ofA:gric;z4lturaiEcon()mics,?tn46S'i!4'17~
Maddala, G. S. (199~), Introduction to Bconant<Urics, Macxmlltm 'P4bli$hing,(tQpjp~ny,.
NewYotk.
Phillips. J. (1966), 'The Theocy anci Pmcd(;e of:Fututes Tr:acUnrf/, Bevicw o/N{(:ltke.tihQ; ·
andAgticultzltal Ect;mamtcs, 34: 43.. 63.
How~vet~ even ·befor¢: th~ ;RJ?s* .ft:l.t4NS <we~;.e not ·coroniori.t~ vs¢d .~y woolgto»'er~,..
i
For instapc~, in u,\. early stu4>t' Phil tip~. (l:966) :fopnq'th~t
Sl1t\~eyed;
#
~YJ ·of ~60 ·:Wboi&row~rs ·
only 6 had use.d 'futu~.s.
Ouripg the wool selling. season1 wool Js· :SQld:
~t ·o,pen ~lU,Cdon.
at ;sellin~ centt¢s
around th.e countr:}~. The gcnctal level of m~k~t 'P~~c(}S ~fOt' ;,wpol' ·lS· ipdJgat~tf ~·w '~ ·
weighted
av~r~ge of ·the 165 reptf!senunive wool ~~s solU1 ¢~il.q~l~t~d :eaoti $ru~.·~~~,
using national Closing quotes ft>r cleart wool. ~o ·provide ,illro.nn~fioq .,a.l:>Qtn i:Pn9~$:~f9r
wool of
~.J?ecmc di~meters, these typ~ ttre grouped, accorcUn.g to fib't~ draW;\~'tt9
pro<!.uc~ lhirteen .micron combing indic~nors. and, two cttrdin~: hldic~totS~· ~~ch Pi~~~ ~:p:
of elt:.ven tYp~$. For example, the .22, :mlGron indicator is ¢t!lc~dtt.ted 'y$JJ1~· th~· \~~~¢$
of el¢wen .22. ·micron
wool typesJ seven .of ·which; are, :Q~ece wool~ wtth.:Ui~ :t¢m{l'inio~:
Fly UP