MIDAS regressions: Further results and new directions
Citations
1,848 citations
696 citations
608 citations
369 citations
References
16,349 citations
11,896 citations
"MIDAS regressions: Further results ..." refers background or methods in this paper
...A kernel-based correction was first introduced by Zhou (1996) and further developed by Hansen and Lunde (2003), Barndorff-Nelsen, Hansen, Lunde, and Shephard (2004) among others....
[...]
...A kernel-based correction was first introduced by Zhou (1996) and further developed by Hansen and Lunde (2003), Barndorff-Nelsen, Hansen, Lunde, and Shephard (2004) among others. Corrections based on sub-sampling were introduced in Zhou (1996), Zhang, Mykland, and Äıt-Sahalia (2005b) and Zhang (2005)....
[...]
...It is of particular interest, because the notion of Granger causality, as put forth in Granger (1969), is subject to temporal aggregation error that can disguise causality or actually create spurious causality when a relevant process is omitted.7 While the MIDAS regression framework does not…...
[...]
...A kernel-based correction was first introduced by Zhou (1996) and further developed by Hansen and Lunde (2003), Barndorff-Nelsen, Hansen, Lunde, and Shephard (2004) among others. Corrections based on sub-sampling were introduced in Zhou (1996), Zhang, Mykland, and Äıt-Sahalia (2005b) and Zhang (2005). Bandi and Russell (2005b) and Bandi and Russell (2005a) studied optimal sampling in the presence of microstructure noise....
[...]
10,019 citations
"MIDAS regressions: Further results ..." refers methods in this paper
...The above specification is very much inspired by the EGARCH model of Nelson ( 1991 )....
[...]
...One parametric choice for g of interest in the context of volatility is yt+k = 0 + K∑ i=1 L∑ j=1 Bij(L1/mi ) ( r (m)t + L ∣∣r (m)t ∣∣)2 + t+1 (17) The above specification is very much inspired by the EGARCH model of Nelson (1991)....
[...]
...Using different methods, Campbell (1987) and Nelson (1991) find a significantly negative relation, whereas Glosten et al. (1993), Harvey (2001), and Turner et al. (1989) find both a positive and a negative relation depending on the method used....
[...]
7,837 citations
"MIDAS regressions: Further results ..." refers background or methods in this paper
...Considering multivariate MIDAS regressions (18) allows us to address Granger causality issues. It is of particular interest, because the notion of Granger causality, as put forth in Granger (1969), is subject to temporal aggregation error that can disguise causality or actually create spurious causality when a relevant process is omitted....
[...]
...Considering multivariate MIDAS regressions (18) allows us to address Granger causality issues. It is of particular interest, because the notion of Granger causality, as put forth in Granger (1969), is subject to temporal aggregation error that can disguise causality or actually create spurious causality when a relevant process is omitted.(7) While the MIDAS regression framework does not necessarily resolve all aggregation issues, it might provide a convenient and powerful way of testing for Granger causality. Indeed, in typical VAR models based on same-frequency regressions, Granger causality may be difficult to detect due to temporal aggregation on the right-hand side variables. The restrictions on the polynomials to test for causality are very much the same as those in the regular Granger causality tests. It is also worth noting that MIDAS regression polynomials, univariate or multivariate, can be two-sided, i.e., they can involve future realizations of x. This allows us to conduct Granger causality tests as suggested by Sims (1972). The multivariate specifications include systems of equations that can address ARCH-in-mean (7)There is a considerable literature on the subject....
[...]
...Using different methods, Campbell (1987) and Nelson (1991) find a significantly negative relation, whereas Glosten et al. (1993), Harvey (2001), and Turner et al. (1989) find both a positive and a negative relation depending on the method used....
[...]
6,294 citations
Related Papers (5)
Frequently Asked Questions (7)
Q2. How many lags would be needed to estimate yt?
For instance, if monthly observations of yt is affected by six months’ worth of lagged daily x (m) t ’s, the authors would need 132 lags (K = 132) of high-frequency lagged variables.
Q3. Why do the authors report the mean absolute deviation (MAD)?
The authors report the mean absolute deviation (MAD) as a measure of goodness of fit (fourth column), because it provides more robust results in the presence of heteroskedasticity.
Q4. What is the function B(L1/m; ) of a few parameters?
As a way of addressing parameter proliferation, in a MIDAS regression the coefficients of the polynomial in L1/m are captured by a known function B(L1/m; θ) of a few parameters summarized in a vector θ.
Q5. What is the predictor of conditional volatility?
Santa-Clara, and Valkanov (2003) show that the best overall predictor of conditional volatility is the realized power and that, not surprisingly, better forecasts are obtained at shorter (weekly) horizons.
Q6. How many parameters are there to estimate?
If the parameters of the lagged polynomial are left unrestricted (or B(k) does not depend on θ), then there would be a lot of parameters to estimate.
Q7. What is the way to model the volatility of financial markets?
It is also worth noting that for stochastic volatility models the problem is even more difficult since the volatility factors are latent and therefore need to be extracted from observed past returns.