A Consistent Nonparametric Test for Causality in Quantile



(A.24), we have

J2(Qθ)=O(T-1h-m).                                                       (A.33)

Next, we show that the result of Step 3 in the proof of Theorem (i) holds under the
alternative hypothesis. Since
F(Qθ (Xj ) zj ) - θ ≠ Ofor j = t, s under the alternative
hypothesis, we have

J3(Qθ)-J3(Qθ-CT)

1 ɪ ɪ 1   ( Z.-Z I

=∑ ∑ mK-lγ~^{F(Qθ(x)|zt)-θ}{F(Qθ()|Zs)-θ}

T (T 1) t=1 sth L h J

1TT1(ZZA

--1∑ ∑ J- K l^I

t(T -1) ⅛ ⅛ hm L h J

× {F(Qθ(xt)-Ctz1 )-θ}{F(Qθ(xs)-Qk)-θ}

1T

= 7 ! F(Qθ(X, ) I z, ) - θi!F(Qθ (Xs ) I zs ) - θ}f (Z, )

T t=1

1T

- 7 {F ( Q ( Xt ) - Ctz, ) - θ}{ F ( Q. ( x ) - Cτz, ) - θ} fz ( z, ).      (A.34)

T t=1

By taking a Taylor expansion of Fyj, (Qθ (Xj ) - Ct | zj ) around Qθ (zj ) for j = t, s, we
have

J3(Qθ)-J3(Qθ-CT)

1T

= -∑ {F(Qe(Xt)zt)-θ}CτS(Qe(Xs))f,(z,)

T t=1

1T

+7CtS(Qθ(X, )){F(Qθ(Xs ) I z,) -θifz(z■,)

T t=1

-ɪCTS(Qθ(X. ))S(Qθ(Xs)).f (z, ).                                   (A.35)

T t=1

We further take Taylor expansion of Fyz (Qθ (Xj ) | zj) around Qθ (Zj ) for j = t, s and have

J3(Qθ)-J3(Qθ-CT)

20



More intriguing information

1. The name is absent
2. A Location Game On Disjoint Circles
3. Handling the measurement error problem by means of panel data: Moment methods applied on firm data
4. CONSUMER ACCEPTANCE OF GENETICALLY MODIFIED FOODS
5. WP 36 - Women's Preferences or Delineated Policies? The development or part-time work in the Netherlands, Germany and the United Kingdom
6. An Empirical Analysis of the Curvature Factor of the Term Structure of Interest Rates
7. ESTIMATION OF EFFICIENT REGRESSION MODELS FOR APPLIED AGRICULTURAL ECONOMICS RESEARCH
8. Education and Development: The Issues and the Evidence
9. HOW WILL PRODUCTION, MARKETING, AND CONSUMPTION BE COORDINATED? FROM A FARM ORGANIZATION VIEWPOINT
10. Regional science policy and the growth of knowledge megacentres in bioscience clusters