29
BEE1020 — Basic Mathematical Economics Dieter Balkenborg Week 15, Lecture Tuesday 10/02/2004 Department of Economics Exponential and logarithmic functions, Elasticities University of Exeter Introduction —Exponential functions: describes growth processes with constant growth rate population growth, growth of GDP, ination etc... — logarithm: inverse function — elasticities Exponential Function power x y base x index or exponent y power function: vary x exponential function: vary y admissible values for y : positive integers, integers, rationals, reel numbers problem: for general y the power x y can only be dened for positive x

University of Exeterpeople.exeter.ac.uk/dgbalken/BME/week15lecsl.pdf · 2008. 11. 18. · Exponential decay most radioactive substances decay exponentially sample of initial size

  • Upload
    others

  • View
    0

  • Download
    0

Embed Size (px)

Citation preview

  • BEE1020—BasicMathematicalEconomicsDieterBalkenborg

    Week15,LectureTuesday10/02/2004

    DepartmentofEconomics

    Exponentialandlogarithmicfunctions,Elasticities

    UniversityofExeter

    Introduction

    —Exponentialfunctions:describesgrowthprocesseswithconstantgrowthrate

    populationgrowth,growthofGDP,inflationetc...

    —logarithm:inversefunction

    —elasticities

    ExponentialFunction

    power

    xy

    basex

    indexorexponenty

    powerfunction:vary

    xexponentialfunction:vary

    yadmissiblevaluesfory:positiveintegers,integers,rationals,reelnumbers

    problem:forgeneralythepowerxycanonlybedefinedforpositivex

  • 05101520

    -4-2

    24

    x

    y=2:

    f(x)=x2

    01234

    -4-2

    24

    x

    y=−2:f(x)=x−2=

    1 x2

    0

    0.20.40.60.811.21.41.61.822.2

    12

    34

    5x

    y=

    1 2:f(x)=x1 2=√ x

    012345

    12

    34

    5x

    y=−3 2:f(x)=x−3 2=

    1x√x

    2

  • approximateirrationalindex

    ybyfraction

    m n:

    xy:=lim m n→yxm n.

    02468

    -2-1

    12

    y

    x=3:

    g(y)=3y

    02468

    -2-1

    12

    y

    x=

    1 3:g(y)=¡ 1 3¢ y

    =3−

    y

    3

  • 0.5

    11.5

    22.5

    3x

    0.6

    0.8

    11.2

    1.4

    y

    1234

    z=xy

    x≥0

    anexponentialfunctionax:

    alwaysconvex,strictlypositivevalues.

    a>1:increasingwithlim

    x→−∞

    ax=0andlim

    x→∞ax=+∞.

    0<a<1:decreasingwithlim

    x→−∞

    ax=+∞andlim

    x→∞ax=0.

    calculationalrulesforgeneralizedpowers:

    as+

    t=as a

    tast=(a

    s )t

    (ab)

    s=as b

    s

    4

  • but

    (as )t6=a(s

    t )

    Compoundedinterestsandthenumbere.

    PutP0>0(theprincipal)insavingsaccount

    fixednominalannualinterestsrate

    r>0

    Interestspaidntimesduringtheyear

    amount

    Ptinyoursavingsaccountaftertyears:

    formulaforcompoundedinterests

    Pt=P0

    ¡ 1+r n

    ¢ ntr ninterestpaidperperiod

    nttotalnumberofinterestpayments.

    The(natural)exponentialfunction:

    balanceinaccountafteroneyearifinterestspaidcontinuously:

    exp(r)=lim

    n→∞¡ 1+

    r n

    ¢ n .5

  • Thetablebelowshowsthevalueof¡ 1+

    r n

    ¢ n forvariousnandr:

    r=5.4%

    r=5.5%

    r=100%

    n=4

    1.0551033751.0561448092.44140625

    n=12

    1.0553567521.05640786

    2.61303529

    n=364

    1.0554803751.0565362252.714557303

    n=8736

    1.0554844261.0565404322.718126265

    n=524160

    1.0554845991.0565406122.718279235

    n=314496001.0554846021.0565406132.718281796

    n→+∞

    1.0554846021.0565406152.718281828

    Thenumber

    eisdefinedas

    e=exp(1)=lim

    n→∞¡ 1+

    1 n

    ¢ n .The‘naturalexponentialfunction’isindeedtheexponentialfunctionwithbase

    e:exp(r)=er

    6

  • “proof”forrationalr:

    exp(r)=lim n→∞

    ³ 1+r n

    ´ n =lim

    m→∞,n=rm

    ³ 1+r n

    ´ n=

    limm→∞

    µ 1+1 m

    ¶ rm=limm→∞

    ·µ 1+1 m

    ¶ m¸r

    =

    · lim m→∞µ 1+

    1 m

    ¶ m¸r

    =er

    formulaforcontinuouslycompoundedinterests:

    Pt=P0er

    t .

    7

  • Propertiesoftheexponentialfunction

    1.e0=1,

    2.e1=e

    3.ex

    >0forallx

    4.d(ex)

    dx=ex

    Inparticular,exisstrictlyincreasingandconvex.

    instantaneousgrowthrateofafunctiony=

    f(x):

    dydx

    . ywhen

    xisin-

    creasedbyaexponentialfunctionhasconstantgrowthrate1.

    5.Exponentialversuspolynomialgrowth:ForanypolynomialP(x)

    limx→+∞

    ex

    P(x)=+∞

    TakingP(x)asatheconstantpolynomialP(x)≡1onehas

    limx→+∞ex=+∞

    8

  • 6.lim

    x→−∞

    ex=0

    7.ea+b=eaeb

    (ea)b=ea

    b .

    Inparticular1 ex=e−

    x(because

    e−xex=e0=1).

    051015

    -3-2

    -11

    23

    x

    Thefunctiony=ex

    Aquickerwaytocalculate

    ex:istousetheformula

    ex=1+x+x2 2!+x3 3!+...+

    xn n!+...

    9

  • Letf n(x)=¡ 1+

    x n

    ¢ n .Intuitionforproperty3:¡ 1+

    x n

    ¢ ispositivewhen

    x>0orwhen

    nlarge

    comparedto|x|

    .Then

    f n(x)>0andsohence

    ex>0.

    Intuitionforproperty4:

    dfn

    dx=n³ 1+

    x n

    ´ n−11 n=³ 1+

    x n

    ´ n−1≈³ 1+

    x n

    ´ n =f n(x)

    fornverylargecomparedto|x|since1+

    x nisthenverycloseto1.

    Intuitionforproperty5:SupposeP(x)=

    amxm−1+...hasdegreem.Ap-

    proximateexby

    f n(x)withnlargerthan

    m.Then

    limx→+∞

    ex

    P(x)≈

    limx→+∞

    ¡ 1+x n

    ¢ nP(x)=

    limx→+∞

    ¡ 1 n¢ nxn+...

    amxm+...=

    limx→+∞Cxn−m=+∞

    10

  • Thelogarithmicfunction

    naturallogarithmfunction x=ln(y)⇔

    y=ex

    -3-2-10234

    -3-2

    -11

    23

    4x

    Therulefordifferentiatinginversefunctionsyields:

    dln(y)

    dy

    =dx dy=1 dy dx=1 ex=1 y

    11

  • Propertiesofln(y):

    1.ln(y)isonlydefinedforstrictlypositivey>0.

    2.dln(y)

    dy=

    1 y.Inparticular,ln(y)isstrictlyincreasingandconcave.

    3.ln(1)=0,ln(e)=1.

    4.lim

    y→0ln(y)=−∞

    ,lim

    y→+∞ln(y)=+∞.

    5.ln(ab)=ln(a)+ln(b),ln¡ ab¢

    =bln(a).Inparticularln¡ 1 a¢ =

    −ln(a).

    Logarithmicdifferentiation:Combinedwiththechainruleoneobtains

    thefollowingusefulformulawherey=g(x)isanydifferentiablefunction:

    dln(g(x))

    dx

    =g0 (x)

    g(x)

    (1)

    12

  • Differentiatinggeneralexponentialandlogarithmicfunctions

    ln(x

    y)=yln(x)generalpowersare:

    xy=eyln(x)=eindex×ln(base).

    Partialdifferentiationyields

    ∂xy

    ∂y=

    eyln(x)ln(x)=xyln(x)

    ∂xy

    ∂x=

    eyln(x)

    µ y1 x¶ =

    yxy1 x=yxy−1.

    derivativeofanexponentialfunctiony=axis

    dax

    dx=ln(a)ax.

    Theinstantaneousgrowthrateofy=axisln(a)

    Thederivativeofapowerfunctiony=xbis

    dxb

    dx=bx

    b−1

    evenifbisirrational.

    13

  • Theinversetotheexponentialfunctiony=ax(for

    a>0,a6=1)iscalledthe

    logarithmicfunctiontothebasicaandwrittenaslog a(y)

    x=log a(y)⇔

    y=ax.

    Wehave

    y=ax⇔

    y=exln(a)⇔ln(y)=xln(a)⇔

    x=ln(y)

    ln(a)

    solog ay=

    ln(y)

    ln(a)

    log ayhashencethederivative

    d(log

    a(y))

    dy

    =1

    ln(a)y

    14

  • CompoundedInterests

    Example:SupposeBankAofferstheannualnominalinterestrater A=5.5%

    andpaysinterestsmonthly.BankBofferstheannualnominalinterestrate

    r B=5.4%

    andpaysinterestsdaily.Whichbankoffersthebetterdeal?

    Solution:

    r eff,A=³ 1+

    r A 12

    ´ 12 −1=

    µ 1+0.055

    12

    ¶ 12 −1=5.64%

    r eff,B=³ 1+

    r B 364

    ´ 364 −1=

    µ 1+0.054

    364

    ¶ 364−1=5.55%

    soBankAoffersbetterdeal.

    15

  • Exponentialdecay

    mostradioactivesubstancesdecayexponentially

    sampleofinitialsize

    Q0weightsQ(t)=Q0e−

    ktattimet.

    kmeasuresrateofdecay

    half-lifeoftheradioactivesubstance:

    Example:Showthataradioactivesubstancethatdecaysaccordingtothe

    formulaQ(t)=Q0e−

    kthasahalf-lifeoft̄=

    ln2k.

    Solution:findvaluet̄forwhichQ(t̄)=

    1 2Q0,thatis

    1 2Q0=Q0e−

    kt̄ .

    Divideby

    Q0andtakenaturallogarithm:

    ln1 2=−k

    t̄.

    Thusthehalf-lifeis

    t̄=−ln1 2

    k=ln2

    k

    16

  • Derivatives

    Example:Findthederivativeofg(x)=xx.

    Solution:Usinglogarithmicdifferentiationweobtain

    g0 (x)

    g(x)=

    d(lnxx)

    dx

    =d¡ lnex

    ln(x)¢

    dx

    =d(xln(x))

    dx

    =1×ln( x)+x×1 x=ln( x)+1

    g0 (x)=(ln(x)+1)xx. 17

  • Thelogisticcurve

    Thegraphofthefunctionoftheform

    Q(t)=

    B

    1+Ae−

    Bkt

    whereA,B,karepositiveconstants,iscalledalogisticcurve.

    describesgrowthprocesseswhenenvironmentalfactorsimposea“braking”effect

    ontherateofgrowth.

    Example:ShowthatthegrowthrateofthelogisticcurveQ(t)=

    11+

    e−tis

    1−Q(t).

    Solution:Wehave

    Q0 (t)=−e−t(−1)

    (1+e−

    t )2=

    e−t

    (1+e−

    t )2

    Q0 (t)

    Q(t)=

    e−t

    1+e−

    t

    1−Q(t)=1+e−

    t−1

    1+e−

    t=

    e−t

    1+e−

    t

    18

  • Example:Publichealthrecordsindicatethat

    tweeksaftertheoutbreakofa

    certainformofinfluenza,approximatelyQ(t)=

    201+19e−1.2tthousandpeoplehad

    caughtthedisease.

    a)Howmanypeoplehadthediseasewhenitbrokeout?Howmanyhadit

    twoweekslater?

    b)Atwhattimedoesthespreadoftheinfectionbegintodecline?

    c)Ifthetrendcontinues,approximatelyhow

    manypeoplewilleventually

    contractthedisease?

    Solution:a)Since

    Q(0)=

    201+19=1itfollowsthat1000peopleinitiallyhad

    thedisease.When

    t=2

    Q(2)=

    20

    1+19e−2.4≈7.343

    soabout7.343thousandpeoplehadcontractedthediseasebythesecondweek.

    b)inflectionpointat

    t̄≈2.5.For

    t<

    t̄convexandsothenumberofnewly

    infectedincreasing.For

    t>

    t̄concaveandsothenumberofnewlyinfectedis

    19

  • decreasing.

    051015

    -4-2

    24

    68

    t

    c)lim

    t→+∞Q(t)=20,soroughly20000peoplecatchthediseaseontotal.

    20

  • NOSUPPLEMENTARYCLASSTHISAFTERNOON

    NEXTSUPPLEMENTARYCLASS:

    FRIDAY20.02AT2P.M.IN

    ROOMSCA!!!

    Elasticities

    Theown-priceelasticity

    demandfunction

    Qd=1000−P3.

    Have

    dQ

    d

    dP=−3

    P2

    currentpriceis$5,priceraisedbyapound:

    quantitydemanddecreasesapproximatelyby

    dQd

    dP|P=5=3×52=75tons.

    one-percentincreaseintheprice:increaseby5p=

    1 20×$1

    reducequantitydemandedbyapproximately3.75=

    75 20tons.

    demandat$5is1000−53=875

    percentagedecreaseinquantitydemandedis3.75875≈0.0043=0.43%.

    Thus1%

    increaseinpricereducesquantitydemandedby0.43%.

    Demandisinelasticatthisprice.

    21

  • Moregenerally:

    dQd

    dPapproximatechangeindemandwhenthepriceincreases

    byapound.

    initialpriceisP,

    increasebyonepoundisanincreaseby

    100

    Ppercent.

    anincreaseofthepriceby1%

    changesthequantitydemanded

    byapproximatelyby

    P 100×

    dQd

    dPtons.

    percentagechangeinquantitydemandedisapproximately:

    ped(P)=100

    Qd×

    P 100×dQ

    d

    dP=dQ

    d

    dP×

    P Qd

    Thisistheown-priceelasticityofdemand.

    rewritethisformulaas

    ped(P)=dQ

    d

    Qd÷dP P

    where100dP Pisthepercentageincreaseinpriceand100dQd

    Qdis(approximately)

    theinducedpercentagechangeinquantity.

    22

  • Inourexample

    ped(P)=¡ −3P

    2¢ ×

    P Qd=−3

    P3

    1000−P3

    Whenisdemandinelastic?

    3P3

    1000−P3<1

    or

    3P3<1000−P3

    4P3<1000

    P3<250

    P<

    3√ 250≈6.3

    Exactlywhen

    P=

    3√ 250thereisunitelasticityandabovedemandiselastic.

    23

  • Totalrevenue

    Withthisdemand,totalrevenueofthemarketis

    TR=PQ

    d=P¡ 1000

    −P3¢

    Totalrevenueismaximizedwhen

    dTR

    dP=1000−4P

    3=0

    orP=

    3√ 250,i.e.,exactlywhenthereisunitelasticity.

    Since

    d2TR

    dP2=−12P

    2<0,totalrevenuedecreasestotheleftandincreasesto

    therightofthisprice.

    24

  • Inversedemand

    Theconsumerswilldemandaquantity

    Qwhentheprice

    Pissuchthat

    Q=

    Qd(P)=1000−P3

    P3=1000−Q

    Pd=

    3p 1000−Q

    inversedemandfunction.

    25

  • Marginalrevenue

    useinversedemandfunctiontoexpresstotalrevenueasafunctionofquan-

    tity:

    TR=PQ=

    3p 1000−Q×Q=(1000−Q)1 3×Q

    quantitydemandedisdecreasinginprice.

    totalrevenueisincreasinginpricewhenitdecreasinginquantityand

    viceversa.

    Marginalrevenueisthechangeinrevenueifasmallunitmoreofthecom-

    modityissoldonthemarket.

    MR=dTR

    dQ=−1 3

    (1000−Q)−

    2 3×Q+(1000−Q)1 3

    marginalrevenueiszerowhen

    (1000−Q)−

    2 3×Q=3(1000−Q)1 3

    Q=3( 1000−Q)=3000−3Q

    4Q=3000

    Q=750

    atP=

    3√ 250.Forlowerquantitiesitispositiveandforhigheronesnegative.

    26

  • Ingeneral,marginalrevenueandown-priceelasticityarerelatedby

    MR=P

    µ 1+1

    ped(P)¶

    Thisissobecause

    dTR

    dP=d¡ PQ

    dP

    =Q

    d+PdQ

    d

    dP

    whereasbythechainrule

    dTR

    dP=dTR

    dQ

    dQ

    d

    dP

    andso

    MR=dTR

    dQ=

    µ Qd+PdQ

    d

    dP

    ¶ÁdQ

    d

    dP=

    Qd

    dQd

    dP

    +P=

    PdQd

    dP

    P Qd

    +P

    27

  • Elasticitiesandlogarithms

    data(x,y)=(lnP,lnQ).

    dy

    dQ=

    1 Q,P=ex

    dP dx=ex=P.Thechainruleappliedtwiceyields

    dy

    dx=

    dy

    dQ

    dQ

    dP

    dP dx=1 Q

    dQ

    dPP=ped(P)

    28

  • Otherelasticities.

    demandforacommodityfunctionofownprice,incomeandotherprices.

    Qd=100−p+2p∗ −

    3y

    pisthepriceofthecommodity,

    p∗thepriceofanothercommodity

    yisincome.

    ownpriceelasticity:

    ∂Q

    d

    ∂p

    p Qd=−

    p Qd

    crosspriceelasticity:

    ∂Q

    d

    ∂p∗

    p∗ Qd=2p∗ Qd

    incomeelasticity

    ∂Q

    d

    ∂y

    y Qd=−3

    y Qd

    29