where σ02=2E{θ2(1-θ)2f (zt)}{∫K2(u)du} under the null.
Step 2. Conditional asymptotic equivalence:
Suppose that both Thm/2(JT-JTU) =op(1) and Thm/2(JT-JTU) =op(1).
Then Thm/2(Jt - JT ) = op (1) . (A.6)
Step 3. Asymptotic equivalence:
Thm/2(JT-JTU) =op(1) and Thm/2(JT-JTL) =op(1). (A.7)
The combination of Steps 1-3 yields Theorem 1 (i).
Step 1: Asymptotic normality.
Since JT is a degenerate U-statistic of order 2, the result follows from Lemma 2.
□
Step 2: Conditional asymptotic equivalence.
The proof of Step 2 is motivated by the technique of Hardle and Stoker (1989) which was
used in treating trimming indicator function asymptotically. Suppose that the following two
statements hold.
Thm/2(JT-JTU)=op(1) and (A.8)
Thm/2(JT-JTL)=op(1) (A.9)
Denote CT as an upper bound consistent with the uniform convergence rate of the
nonparametric estimator of conditional quantile given in equation (13). Suppose that
sup ∣ qθ (x) - qθ (x)l≤ Ct . (A.10)
If inequality (A.3) holds, then the following statements also hold:
{ Qθ(x) > yt + CT } ⊂ { (Qθ(x) > yt } ⊂ { Qθ(x) > yt -CT }, (A.11-1)
1( Qθ (x) > yt + CT ) ≤ 1( (Qθ (x) > yt ) ≤ 1( Qθ (x) > yt - CT ), (A.11-2)
Jtu ≤ Jt ≤ Jtl , and (A.11-3)
| Thm/2(Jt - Jt)| ≤ max {∣Thm/2(Jt - Jtu)| , |Thm/2(Jt - Jtl)| } (A. 11-4)
12
More intriguing information
1. The name is absent2. The name is absent
3. The name is absent
4. Climate change, mitigation and adaptation: the case of the Murray–Darling Basin in Australia
5. Evidence of coevolution in multi-objective evolutionary algorithms
6. The name is absent
7. Problems of operationalizing the concept of a cost-of-living index
8. Second Order Filter Distribution Approximations for Financial Time Series with Extreme Outlier
9. Needing to be ‘in the know’: strategies of subordination used by 10-11 year old school boys
10. TOWARDS THE ZERO ACCIDENT GOAL: ASSISTING THE FIRST OFFICER MONITOR AND CHALLENGE CAPTAIN ERRORS