Professional Documents
Culture Documents
American Statistical Association
American Statistical Association
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
American Statistical Association is collaborating with JSTOR to digitize, preserve and extend access to Journal
of the American Statistical Association.
http://www.jstor.org
This article considers forecasting a single time series when there are many predictors(N) and time series observations(T). When the
data follow an approximatefactor model, the predictorscan be summarizedby a small number of indexes, which we estimate using
principalcomponents. Feasible forecasts are shown to be asymptoticallyefficient in the sense that the difference between the feasible
forecasts and the infeasible forecasts constructedusing the actual values of the factors converges in probabilityto 0 as both N and T
grow large. The estimated factors are shown to be consistent, even in the presence of time variationin the factor model.
KEY WORDS: Factor models; Forecasting;Principalcomponents.
consistent estimation of the factors is not relevant. However, by associatingA with the orderedorthonormaleigenvectorsof
several authorshave noted that with large N, consistent esti- (NT)-1 T l AFtFt'A and {Ft}T=,with the principalcompo-
mation is possible. Connorand Korajczyk(1986, 1988, 1993) nents of {AFt}t=. The diagonal elements of EFF correspond
discussed the problemin a static model and arguethat the fac- to the limiting eigenvalues of (NT)-1 ET AFFt'A'. For con-
tors can be consistently estimatedby principlecomponentsas venience, these eigenvalues are assumed to be distinct. If they
N -> oo even if the errorsterms are weakly cross-sectionally were not distinct, then the factors could only be consistently
correlated.Fomi and Reichlin (1996, 1998) and Forni, Hallin, estimatedup to an orthonormaltransformation.
Lippi, and Reichlin (1999) discussed consistent (N, T -- oo) Assumption Fl(b) allows the factors to be serially corre-
estimationof factors in a dynamic version of the approximate lated, althoughit does rule out stochastictrendsand otherpro-
model. Finally,in a predictionproblemsimilarto the one con- cesses with nonconstantuncondititionalsecond moments. The
sidered here, Ding and Hwang (1999) analyzed the properties assumptionalso allows lags of the factors to enter the equa-
of forecasts constructedfrom principal components in a set- tions for xit and Yt+h . A leading example of this occurs in the
ting with large N and T. Their analysis is conducted under dynamic factor model
the assumption that error process {et, 8t+h} is cross-sectionally
and temporally iid, an assumption that is inappropriatefor Xit = Ai(L)ft + eit (3)
economic models and when interest focuses on multiperiod
and
forecasts.We highlightthe differencesbetween our results and
those of others later in the article. Yt+h = f (L) f, + fw, +tl+h' (4)
The article is organized as follows. Section 2 presents the where Ai(L) and 8f(L) are lag polynomials is nonnegative
model in more detail, discusses the assumptions,and presents
powers of the lag operator L. If the lag polynomials have
the main consistency results. Section 3 generalizes the model finite order q, then (3)-(4) can be rewrittenas (1)-(2) with
to allow temporal instability in the factor model. Section 4
Ft =(ftft I .
ft-q)', and Assumption Fl(b) will be satisfied
examines the finite-sampleperformanceof these methods in a if the ft process is covariancestationary.
Monte Carlo study, and Section 5 discusses an applicationto In the classical model, the errors or "uniquenesses"are
macroeconomicforecasting. assumedto be iid and normallydistributed.This assumptionis
2. THE MODELAND ESTIMATION clearly inappropriatein the macroeconomicforecasting appli-
cation, because the variablesare serially correlated,and many
2.1 Assumptions or the variables(e.g., alternativemeasures of the money sup-
As described in Section 1, we focus on a forecasting sit- ply) may be cross-correlatedeven after the aggregate factors
uation in which N and T are both large. This motivates our are controlled for. We therefore modify the classic assump-
tions to accommodatethese complications.
asymptoticresults requiringthat N, T -> oo jointly or, equiv-
alently, that N = N(T) with limT, N(T) -> oo. No restric- AssumptionMl (Momentsof the Errors et).
tions on the relative rates of N and T are required.
The assumptionsabout the model are groupedinto assump- a. E(e'et+u/N)= YN,t(u), and
tions about the factors and factor loading, assumptionsabout limN-, suptE _-o[7N,t(u) < o.
the errorsin the (1), and assumptionsabout the regressorsand Let eit denote the ith element of et; then
errorsin (2). b. E(eitejt) = rij t,lim supt N-' E E=l I it < ,
and
AssumptionFl (Factors and Factor Loading). c. limNo supt s N- ENi E=I Icov(eiseit, eejt)I < o.
a. (A'A/N) --Ir
Assumption Ml(a) allows for serial correlation in the ei,
b. E(FtFt) = EFF, where rFF is a diagonal matrix with
processes. As in the approximatefactor model of Chamber-
elements aii > rjj> 0 for i < j. lain and Rothschild (1983) and Connor and Korajczyk(1986,
c. Ai,ml< A < oo.
1993), Assumption Ml(b) allows {eit} to be weakly corre-
d. T-' Et FtFt' FF. lated across series. Forni et al. (1999) also allowed for serial
correlationand cross-correlationwith assumptions similar to
Assumption Fl serves to identify the factors. The nonsin-
gular limiting values of (A'A/N) and FF imply that each of Ml(a)-(b). Normality is not assumed, but Ml(c) limits the
the factors provides a nonnegligible contributionto the aver- size of fourth moments.
It is assumed that the forecasting equation (2) is well
age variance of xit, where xit is the ith element of Xt and
the average is taken over both i and t. Moreover, because behaved in the sense that if {Ft} were observed, then ordi-
AFt = ARR-'Ft for any nonsingularmatrix R, a normaliza- nary least squares(OLS) would provide a consistentestimator
tion is required to uniquely define the factors. Said differ- of the regression coefficients. The specific assumption is as
follows.
ently, the model with factor loadings AR and factors R-'F,
is observationallyequivalent to the model with factor load-
Assumption Y1 (Forecasting Equation). Let Zt = (Ft wt)'
ings A and factors Ft. Assumption Fl(a) restricts R to be and /3 = (3 /3)'. Then the following hold:
orthonormal,and this togetherwith AssumptionFl(b) restricts
R to be a diagonal matrixwith diagonal elements of ? 1. This FF EFw a
= ''FF
a. E(ztz) = - EF positive definite matrix.
identifies the factors up to a change of sign. Equivalently, b.t zzPY_wF ww_
Assumption Fl provides this normalization(asymptotically) b. T-' E, z'z _; .
The theoremfollows directlyfrom Theorem 1 togetherwith b. The initial values of the values loadings satisfy
Assumption Y1. The details of the proof are given in the N-' Ei AoAio = AoAo/N -l Ir and sup,j/Ai, I < A,
Appendix. where Aij0 is the jth element of Aio.
3. TIME-VARYING
FACTORLOADINGS As discussed earlier, Assumption F2(a) makes the amount
of time variationsmall. AssumptionF2(b) means that the ini-
In practice,when macroeconomicforecasts are constructed
tial value of the factor loadings satisfy the same assumptions
using many variablesover a long period, some degree of tem- as the time-invariantfactor
loading of the preceding section.
poral instability is inevitable. In this section we model this The next limits the dependence in ,t. This
assumption
instability as stochastic drift in the factor loadings, and show
that if this drift is not too large and not too dependentacross assumption is writtenin fairly generalform, allowing for some
series (in a sense made precise later), then the results of Theo- dependence in the random variablesin the model.
rems 1 and 2 continue to hold. Thus the principalcomponents
Assumption M2. Let it, m denote the mth element of i,t.
estimator and forecast are robust to small and idiosynchratic Then the
following hold:
shifts in the factor loadings.
Specifically,replace the time-invariantfactormodel (1) with a. limTo sup Es --s supt q, .mIE(FitFmqis,l is+v, m) < ?.
b. limTo suPi, s E --s SPm q E(eisFmq~is+v,m) < o.
Xit= AitFt+ eit (7)
c. limN,oo N-'l i ,j SUPI,t,m,q,s, v E(FltFmqis, l~jv,m)<too
and d. limN -o N-1 Ei EjSU,su , q,mIE(eitFms,jq,m)l < oo.
Ait = ,Ait- +giTrit (8)
e. 1. limN,, N-1i E,j sup{lk}k2={tk}6
Icov(eit,ei2,
for i = 1,..., N and t = 1,..., T, where giT is a scalar and FlI t3 jt4, l Fl2t5 jt6, 12)
< 00.
evolved through time. The design used here incorporatesall the Monte Carlo experiment are N, T, r, q, k, TT,a, b, c, 81,
of these features, and the data are generatedaccordingto and 82.
q
The results are summarizedby two statistics.The firststatis-
tic is a trace R2 of the multivariateregression of F onto F,
Xit= E Aijtft-j+ eit, (10)
j=o
= EPF
R,F, F /El= llP EI Etr(F'PFF)/Etr(F'F), (17)
At = Ait_l + (c/T)rj't, (11)
ft = ?aft-, + ut, (12) where E denotes the expectation estimated by averaging the
relevant statistic over the Monte Carlo repetitions and PF =
(1 - aL)eit = (1 + b2)vit + bi+l, t + bVi-l, t, ((13)
1) --Fp1.
F(F'F)-1F'. According to Theorem 1, R2 F,F
Vit = O'it7it (14) The second statistic measureshow close the forecast based
on the estimated factors is to the infeasible forecast based on
and thi-
LIIC tril-
L1LU fCnitnrc
IldLVL 3,
2it = 80 + it-l +- 1Vit-l (15)
where i= 1,..., N, t = 1,...T, ft and Alit are rx 1, and y = 1-E(YT+I/T -YT+1/T)/E(YT/T+1)- (18)
the variables {}ijt}, {ujt}, and {17it} are mutually independent
Because -
0 when k = r from Theorem 1,
iid N(0, 1) randomvariables.Equation(10) is dynamic factor 5YT+1/T- YT+I/T
model with q lags of r factors that, as shown in Section 2, can S2 y should be close to 1 for appropriatelylarge N and T. S2
be representedas the static factor model (1) with r = r(l + q) is computed for several choices of r. First, as a benchmark,
factors. From (12), the factors evolve as a vector autoregres- results are shown for r = r. Second, r is formed using three
sive [VAR(1)]model with common scalar autoregressive(AR) of the informationcriteria suggested by Bai and Ng (2001).
These criteria have the form ICp(k) = ln(Vk) + kg(T, N),
parametera. From (13), the errorterms in the factor equation
are serially correlated, with an AR(1) coefficient of a, and where Vk is the minimized value of the objective function (5)
cross-correlated,[with spatial moving average [MA(1)] coeffi- for a model with k factors and gj(T, N) is a penalty function.
cient b]. The innovations vit are conditionallyheteroscedastic Three of the penalty functions suggested by Bai and Ng are
and follow a GARCH(1, 1) process with parameters80, 81, used:
and 82 [see (14) and (15)]. Finally, from (11), the factor load-
ings evolve as random walk, with innovation standarddevia-
tion proportionalto c.
The scalar variableto be forecast is generatedas (N, T) (N+T In C2
NT NT
q
Table1. MonteCarloResults
S2y
NOTE: Based on 2,000 Monte Carlo draws using the design described in Section 4 of the text. The statistic R2? is the trace of a regression of the estimated factors on the true factor, and
F,F
S2 is a measure of the closeness of the forecast based on the estimated factors to the forecast based on the true factor; see the text for definitions. The values are reported using the correct
<
number of factors and for ? estimated using an information criterion, where 0 < r k.
k = 15). The values of -2 generally exceed .9, and this is section we describe a forecasting experimentfor the Federal
true for all methods used to estimate the number of fac- Reserve Board's Index of IndustrialProduction,an important
tors. The only importantexception is when k = r = 10 and monthly indicator of macroeconomic activity. The variables
N = T = 100; in this case, ICpl and ICp2 performpoorly. The making up Xt are 149 monthly macroeconomicvariablesrep-
penalty factors for ICpl and ICp2 are largerthan for ICp3 [e.g., resenting several different facets of the macroeconomy (e.g.,
when T = N, g2(N, T) = 2g3(N, T)], and apparentlythese production, consumption, employment, price inflation, inter-
large penalties lead to serious underfitting. est rates). We have described the variablesin detail in earlier
Performance deteriorates somewhat is small samples, as work (Stock and Watson 2002). The sample period is January
shown in panel B. With only two factors, S2 y is near .9 for 1959-December 1998. Principalcomponents of Xt were used
k = r, so that the forecastsperformnearly as well as the infea- to construct forecasts of Yt+12 = ln(IPt+12/IPt), where IPt is
sible forecasts. When k is much larger than r (k = 10 and the index of industrialproductionfor date t. These 12-month-
r = 2), ICp3 performs poorly because of overfitting,particu- ahead forecasts were constructed in each month starting in
larly when T is very small (T = 25). All of the methods dete- 1970:1 and extendingthrough1997:12, using previouslyavail-
rioratewhen there are five factors;for example, when T = 25, able data to estimate unknownparametersand factors.
the values of S2 ~ are closer to .6. To simulatereal-timeforecasting,we used data dated T and
Panel C suggests that including irrelevantseries has little earlier in all calculationsfor constructingforecasts at time T.
effect on the estimatorsand forecasts when N and T are large. Thus for example, to compute the forecast in T = 1970:1, the
Results are shown for 7r= .25 (so that 25% of the series are variables making up Xt were standardizedusing data from
unrelated to the factors) and N = 333. The results for the t = 1959:1-1970:1, and principalcomponentswere computed.
models with five factors are nearly identical to the results in These estimatedvalues of Ft were used togetherwith Yt+12for
panel A with the same number of relevant series (N = 250). t = 1959:1-1969:1 to estimate /3 in (2). Model selection with
The results for 10 factors are also similar to those in panel k = 10 based on IC,,, ICp2, ICp3, AIC, and BIC were used
A, althoughpanel C shows some deteriorationof the forecasts to determine the number of factors to include in the regres-
using ICpl and ICp2 sion. Finally, the forecasts constructed in T = 1970:1 were
From panel D, moderate amounts of serial or spatial cor- formedas Yt+2/t = Ft,.This process was repeatedfor 1970:2-
relation in the errors have little effect on the estimators and 1997:12.
forecasts. For example, on the one hand, when moderateserial We also computed forecasts using four other methods: a
correlationis introducedby setting a = .5, the results in the univariateautogressionin which Yt+12was regressed on lags
table are very similarto the results with a = 0; similarly,there of ln(IPt/IPt_l), a vector autoregression that included the
is little change when spatial correlationis introducedby sett- rate of price inflation and short-terminterest rates in addi-
ting b = 1.0. On the other hand, some deteriorationin perfor- tion to the rate of growth of the industrialproductionindex,
mance occurs when the degree of serial correlation is large a leading-indicatormodel in which Yt+12 was regressed on
(comparethe entries with a = .9 to those with a = .5). Condi- 11 leading indicators chosen by Stock and Watson (1989)
tional heteroscedasticityhas no apparenteffect on the perfor- as good predictorsof aggregate macroeconomicactivity; and
mance of the estimatorand forecasts. an autoregressive-augmentedprincipal components model in
Frompanel E, introducinglags of the factorshas little effect which Yt+12 was regressed on the estimated factors and lags
on the quality of the estimatorsand forecasts: the results with of ln(IPt/IPt_l). We gave details of the specification,includ-
r = 5 and q = 1 (so that r = 10) are essentially identicalto the ing lag length choice and exact descriptionof the variables,in
static factor model with 10 factors. However,a high degree of earlier work (Stock and Watson 2002).
serial correlationin the factor process (a = .9) does result in Table 2 shows the MSE of the resulting forecasts, where
some deteriorationof performance.For example, when T = we have shown each MSE relativeto the MSE for the univari-
100, N = 250, and r = 5, R2 F, F
= .97 in the static factor model ate autoregression.The first three rows show results from the
(a = 0), and this falls to .89 when a = .9. benchmarkAR, VAR, and leading indicatormodels. The next
Finally, panel F shows the effect of time variation on the row shows the results for the principalcomponents forecasts,
factor loadings in isolation and together with other compli- with the number of factors determinedby ICp3. (The results
cations. There appears to be only moderate deteriorationof for the other selection proceduresare similar and thus are not
the forecast performanceeven for reasonablylarge amountof reported.)This is followed by principalcomponentsforecasts
temporalinstability(e.g., S2 y remainshigh even as the param- using a fixed numberof factors (k = 1-k = 4). Finally,the last
eter governing time-variationincreases from 0 to 10). How- row shows the principalcomponents forecasting model (with
ever, when all of the complications are present (i.e., serially r estimatedby ICp3)augmentedwith BIC-selected lags of the
correlateddynamic factors, k > r, serial and cross-correlated growth rate of industrialproduction.(Again, results for other
heteroscedasticerrors,time-varyingfactorloading, and a large selection proceduresare very similar and are not reported.)
number of unrelated x's), forecast performancedeteriorates Both the leading indicatorand VAR models performslightly
significantly,as shown in the last few entries in panel F. better than the univariateAR in this simulated out-of-sample
5. AN EMPIRICALEXAMPLE experiment.However,the gains are not large. The factor mod-
els offer substantial improvement. The results suggest that
In related empirical work we have applied factor mod- nearly all of the forecasting gain comes from the first two
els and principal components to forecast several macroeco- or three factors and that once these factors are included, no
nomic variables (see Stock and Watson 1999, 2002). In this additionalgain is realized from including lagged values of IP
Table2. SimulatedOut-of-SampleForecastingResults Industrial The term N-1 Eiii, is 0(1) from Assumption Ml(b) (because
Production,12-MonthHorizon N-' i Tii,t < N-1l i Ej 7it tl 0(1)). So it suffices that the sec-
ond term converges to 0 in probability.Now
Forecastmethod RelativeMSE
2
Univariateautoregression 1.00 E N-1 (e2 - Tii = N-2 - -
Vectorautogression .97 t)] E(e2t
Ttii, E ,)(e Tij, t)
L i i j
Leadingindicators .86
Principalcomponents .58 = N-2 E cov(ei2t, e2)
Principalcomponents, k = 1 .94 i J
Principalcomponents, k =2 .62
Principalcomponents, k =3 .55 < N-2EE 1cov(e2, e2t)1
Principalcomponents, k = 4 .56 i j
Principalcomponents, AR .69
N N
Root MSE,AR model .049 sup N-2 E Icov(eisei,,ejej) -> 0,
t, i= j=l
NOTE: For each forecast method, this table shows the ratio of the MSE of the forecast made
by the method for that row to the MSE of a univariate autoregressive forecast with lag length
selected by the BIC. The final line presents the root MSE for the autoregressive model in by AssumptionMl(c). Thus N-1 Yi(ei2 - rii, ) - .
native (decimal growth rate) units at an annual rate.
Proof.
growth.We have alreadyreportedsimilarresults for other real
macroeconomicvariables (Stock and Watson 2002).
(N2T)-ly'e'ey = (N2T)-1 EE yiyjei,ejt
t i j
6. DISCUSSION
This article has shown that forecasts of a single series i j t
based on principal components of a large number of predic-
tors are first-order asymptotically efficient as N, T -+ oo for '
-2EE 2
general relationshipsbetween N and T in the context of an
<(N
-
' /J Tj
sup IT-1 qt(N-' yieit)lP O. where 6-' is the largest eigenvalue of (A'A/N)1/2 (F'F/T)(A'
7 ,t i
A/N)1/2. But (A'A/N)1/2 -- I by Fl(i) r'r/
and T F by Fl(d),
Proof. supt IT-' Et q,(N-' Ei ieit)lI (T-1 tq2)/2x so that (A'A/N)1/2 (F'F/T)(A'A/N)1/2--FF and (by continuityof
(supyEr T-1' E(N-1 Ei y,iei)2)1/2. eigenvalues) 'l-^ii.
The first term is Op(1) by assumption,and so the result follows from (R9) sup(er R(y) ol.
Proof. This follows from (R7) and (R8).
sup T-1 = sup(TN2)- E E E yijeitej,
(N-1 E iei,)
Er t i ytr i j (R10) Let A1=argsupEr R(y); then R*(Ail)Caol.
= sup(N2T)-'y'e'ey-P 0, Proof. This follows from (R6) and (R9).
yEr
(Rll) Let A denote the first column of A and let S1 =
where the limit follows from (R2). sign(A1A1), meaning S1 = 1 if A1A1 > 0 and S1 = -1 if
A1l < 0.
(R4) supYErIT-1 t Fjt(N-1 Ei yiei)O forj = 1, 2....r.
Then (S,1AA/N)-A;1, where e, = (100... 0)'.
Proof. Because T-1 Fjt -ojj [from Assumption Fl(d)], the
result follows from (R3). Proof. For particularvalues of 8 and V, we can write A1 =
A(A'A/N)-/2 + V, where V'A= 0 and 8'8 < 1. (Note that 8 is
(R5) sup,ErF(N2T)-ly'AF'ey-1 0. r x 1.) Let CNT = (A'A/N)1/2'(F'F/T)(A'A/N)1/2 and note that
R*(Al) = 5'CNT8. Thus
Proof. (N2T)-1y'AF'ey=_j(y'Aj/N)T-1 tFjt(N-' Eiyeit),
so that -
R*(A1i)- '11 = (CNT .F)8+' -.FF- 11l
r
(N2T)-ly'AF'eyl < E (y'Aj/N) T-l
1
Fjt(N- ieit) = 8'(CNT - )FFA + (6 - l) + E i iiI
i i i=2
Thus, Because CNTp-FF and 5 is bounded,the first term on the right side
of this expression is op(l). This result together with (R10) implies
supI(N2T)- y'AF'eyl that (82 - 1)o1 + Zi=2ioii 0 Because aii > O, i = 1,... r
yEF
y,r
[Assumption Fl(b)], this implies that 8- 1 and 82->0 for i>
1. Note that this result, together with A'A1/N = 1, implies that
(maxsuply'Aj/NI) Esup T-' tFjt(N E ieit)
--w P
V'V/N--O.
The result then follows from the assumption that A'A/N -> I
< /2
sup(y'y/N)l/2)
F
max(Ak./N) [AssumptionFl(a)].
yEr jl
Proof. T-1 E, Ft = R(Aj) by the definitionsof F and R(.). The Because (SA'A/N)PI by (R12) and T-1 E, Ft,q,Fq by assump-
convergence result follows from (R9) for j = 1 and from the steps tion,
outlined in (R12) for j = 2, .., r. T-1 (SA'A/N)Ftqt Fq.
t
(R14) For i > r, T-' Et/ -0.
Now the jth element of T- N-' Et SA'etqt satisfies
Proof. Let y denote the ith ordered eigenvector of X'X, i > r,
normalizedso that ' ^/N = 1. Then T-' Et -t = R(y) follows from
T-'N-1 S Aeqt = T-'1E t N-1 E jeit
the definitionof R(.) and Fit. t T
Now, for particularvalues of & and V, we can write
< sup T /-1 q ,) o
=A(A'A/N)-'/2 + V, YEf' t ~~~~~i Pl
where V'A = 0, &a' < 1, and V'V/N < 1. Now, by construction, where the final inequality follows because Ai E F and the limit fol-
y'A= 0 for j = 1,. , r. Using the representationfor Aj given lows from (R3).
in (R12), we can write N- y'Ai = &'6, + V'VJ/N = 0. Because
"~l
'~'
At
A p ^ ^ t p^ (R17) T-'1, SFtFt FF.
from (R12), Vj'V-O and V'V/N < 1, V'V->O. Thus &'^j-0, so
t~t[Sl...^
But P[l. . . ]Pr, SO&^^.O. Proof. This follows from (R16) with qt = Ft, = ..... r.
&'[i ... 5,r]-0. -r]r, so
Assumption Fl(d) shows that this choice of qt satisfies the restriction
Thus R*(() = &'(A'A/N)l/2 (F'F/T)(A'A/N)' /2&P0. The
in (R16.)
result then follows from (R6).
(R18) T-' EtFtF,'- FF
(R15) SjFj, -Fjt,-O for j = 1,..., r.
Proof. This follows from (R16) and (R17). Set q, = SjFjt. (R13)
Proof. SjFj, - Fj = SjAjX,/N - F = (SjAjA/N - 1)Fjt + shows that qt satisfies the conditions of (R16).
EiAj(Sj_i,/N)Fit + SjA/et/N.
Because S(A'A/N)P I from (R12), (Sj jI/N - 1)--0 and (R19) For i = 1, 2, ..., r, T-' E(SiFi -Fi)
N-' IL(SjIij
N-1^k- - Aij)eit < (N- ^Ay-V-
^T E(Sjij ii)2) ) (N-
(^ E ^it
ei2t) Proof. This follows from (R16) and Yl(c) with qt equal to E,+h.
Yl(d) shows that this definition of qt satisfies the conditions of
(R16).
and
(R22) With /3 partitioned as /3= (3 )', w- 3-?? and
N-1 E(SjAij - j)2 = (1.j/N) + (A,A1/N)- 2(Sj^.Aj/N)-)0 SiiZ
-
P-Ofor il
Proof Write
by (R12). Finally,N-1 E ei2 ~ Op(l) by (RI). Thus IN-' Ei(Sjij -
Aj)ei,tlt0 by Slutsky's theorem. This means that Sjie/N =
N- Ei Aieit + op(1). Now (,, )
(\o
[. Ow
2
r
E[N-1 ' ijeit = N-2E 3AijAkj E(eitekt) (-'E, F;S T-'S E FtwA (T-'S Et Ftet+h)
i i k
T-l It
_kT-Ewt FtFt'S
wtFt' T-'1 E, ,w; T-1t' W,E,+h
= < Af2N- 2 t > 0,
N-2L AijAkjTika, 71rik
i k i k
Because T-' tFtF'-t F F (R18), T-'SE,F,tw PEF, (R20),
where the final inequality uses the bound on Aij given in Fl(c) and T-'E,wtwtE2w [Yl(b)], T-'SEtFtt+h0 (R21), and T-'x
the limit uses Ml(b). Thus SjAiet/No0. EYt WtIt+h--O [Yl(c)]; and because 2zz is nonsingular[Yl(a)], the
result follows by Slutsky's theorem.
(R16) Let qt denote a sequence of random variables with
(R23) Let Z, = =.. FrtW)'
(F,,F2t... .,t ZtZt=and /3 =
T-1 Et qt q2 and T-1 EFtqt- Fq. Then T-1 tS (Et^ihz^)
then /'
F,q,t-4 Fq
(Etl Ztyt+h); ZT-p 'ZT0-.
Proof. Proof. Let R = [ ,? ], where n, denotes the number of ele-
ments in w,.
T-' >SFtq, = N-T-' SA'Xq,
t t
- /3'ZT = -
/3'Z (RB)'RT 3'ZT
= T-1 Y,(SA'A/N)Ftqt + 'N-1 E SA'etqt.
= (R/ - + (R3)'(RT
ZT)'Z - ZT).
Supi,jEjJTJjTl <K2,
sup u?E-t E[N1 lJiTeitFmt+u6it+ u m
supi,j,k EIJlTJjTJkTI<K3,
and =t s r T i N' ~~UJe,Fm,?uf,sm1
SUPi, <K4.
jk,lEIJiTJjTJkTJfTI
T-t N t+u
t Li= I-t i=1 s= 1
This follows from F2(a). s sup T-i
K1 L NiL IE(eiTFmtt+us,m)I
Si.
T-t N T-p
The error term a satisfies Assumption Ml(a). That is, <KlsupT1 L N-iL I suppFt,)m)
IE(eipFms I
ET-t_l-t JE(N-1 Ci aitait+,) 0 t U=I-t i=i v=ipsPm
T-t T-p
Proof atait+u = ei,ei+u + T(Ft + r ,?+
JiTeitFt/+U
K,supT-' C sup C sup IE(eipFms4p?v,m)I
JiTeit+uFtlSit. t u=i-t i V=1-pS,P,m
= E Em JiTJjTFltit,
S3(a). limN,,, supt,, N-' Ei= Ej= I cov(eiseit, ejsejt)l < ?o.
Proof. JiTJjT(F;it) (F;,Jj) Fmtjt, m, and
we show the requiredresult for each term in the sum: Proof. This is AssumptionMl(c).
N N
E : I|E[JiTJjTFlt,it, Fmtjt, m]I S3(b). limN,, sup,, N-' i= Ej= Icov(eiseit, JJT(Ftj,t)x
supN-'
t i=l j=1 (F,'js))lI < .
N N N N t s
< K2N-'E E l,t,s,v,q,m
sup IE(FtVis, IFmvjq, m)|
=supT-2N-'1 E Ecov(eiseit, JJTFlt.jq,IFmsjh, m)
i=l j=l t,s i=lj=l q=lh=l
< 00,
Ft'i,, and so cov(aisait, aisajt) is made of 16 terms, which have =sup T-4N E E EE E cov(JFi,tiq,ll
1, 1 Fl2s'iq2,12,
6 possible forms: t, s i=l j=l q1 q2 3 q4
N N t s t s N N s s
K4supT-4N -1L
EL <K2sup T-2 N-1 E lcov(eitFsiqll,ejtF l
2,m)
t s i=l j=l q1 q2 q3 q4 t, s i=l1=l 1q q2
Fl3t ,3Fjql4,14F ) < 00, where the last line follows from M2(e)(5).