The three-pass regression filter: A new approach to forecasting using many predictors

Kelly, Bryan T.; Pruitt, Seth

doi:10.1016/j.jeconom.2015.02.011

Cited by 278 publications

(144 citation statements)

References 37 publications

Supporting

Mentioning

143

Contrasting

Order By: Relevance

“…Because no assumption is made about the number of targetrelevant factors, we employ an automatic proxy-selection algorithm proposed by Kelly and Pruitt (2015). This approach uses the forecasting residuals from Equation 5 as the second target-relevant proxy and conduct the three-pass regression filter again.…”

Section: Target-relevant Factor Estimatesmentioning

confidence: 99%

See 1 more Smart Citation

Does a lot help a lot? Forecasting stock returns with pooling strategies in a data‐rich environment

Baetje

2017

Journal of Forecasting

View full text Add to dashboard Cite

A variety of recent studies provide a skeptical view on the predictability of stock returns. Empirical evidence shows that most prediction models suffer from a loss of information, model uncertainty, and structural instability by relying on lowdimensional information sets. In this study, we evaluate the predictive ability of various lately refined forecasting strategies, which handle these issues by incorporating information from many potential predictor variables simultaneously. We investigate whether forecasting strategies that (i) combine information and (ii) combine individual forecasts are useful to predict US stock returns, that is, the market excess return, size, value, and the momentum premium. Our results show that methods combining information have remarkable in-sample predictive ability. However, the out-of-sample performance suffers from highly volatile forecast errors. Forecast combinations face a better bias-efficiency trade-off, yielding a consistently superior forecast performance for the market excess return and the size premium even after the 1970s.

show abstract

Section: Target-relevant Factor Estimatesmentioning

confidence: 99%

“…This table shows in-sample predictability results obtained by target-relevant factor models. We make use of the automatic proxy selection algorithm proposed by Kelly and Pruitt (2015). The optimal number of factors is determined by BIC.…”

Section: Oos Forecast Performancementioning

confidence: 99%

Does a lot help a lot? Forecasting stock returns with pooling strategies in a data‐rich environment

Baetje

2017

Journal of Forecasting

View full text Add to dashboard Cite

show abstract

“…Kelly and Pruitt (2015) [55] essentially study the PLS estimator under a factor structure and forecast model (2.5). Their factorial structure allows for both relevant and irrelevant factors and the latter can distort the performance of PCR.…”

Section: Comparison Of Assumptions In the Two Approachesmentioning

confidence: 99%

“…PLS is an increasingly popular method of dimension reduction that has recently resurfaced in econometrics, within macro-forecasting applications, with Kelly and Pruitt (2015) [55] and Groen and Kapetanios (2014) [44], mostly because PLS handles regressions where p > T . 18 PLS solves the maximization problem [see Stone and Brooks (1990) [77]]:…”

Section: Partial Least Squares (Pls)mentioning

confidence: 99%

“…In attempts to boost the forecasting performance of DFMs, selected tools from the plethora of reduction methods in statistical and machine learning were explored in the econometric literature. Notable early examples are Bai and Ng (2008) [5], DeMol, Giannone and Reichlin (2008) [33] and, more recently, Stock and Watson (2012) [76], Kelly and Pruitt (2015) [55] and Groen and Kapetanios (2014) [44]. The survey by Ng (2013) [62] provides additional references, as well as a discussion on targeting, a central issue in this paper.…”

Section: Introductionmentioning

confidence: 98%

See 1 more Smart Citation

A Unified Framework for Dimension Reduction in Forecasting

Barbarino

Bura²

2017

FEDS

View full text Add to dashboard Cite

Factor models are widely used in summarizing large datasets with few underlying latent factors and in building time series forecasting models for economic variables. In these models, the reduction of the predictors and the modeling and forecasting of the response y are carried out in two separate and independent phases. We introduce a potentially more attractive alternative, Sufficient Dimension Reduction (SDR), that summarizes x as it relates to y, so that all the information in the conditional distribution of y|x is preserved. We study the relationship between SDR and popular estimation methods, such as ordinary least squares (OLS), dynamic factor models (DFM), partial least squares (PLS) and RIDGE regression, and establish the connection and fundamental differences between the DFM and SDR frameworks. We show that SDR significantly reduces the dimension of widely used macroeconomic series data with one or two sufficient reductions delivering similar forecasting performance to that of competing methods in macro-forecasting. JEL Classification Number: C32, C53, C55, E17

show abstract

Forecasting local currency bond risk premia of emerging markets: The role of cross‐country macrofinancial linkages

Çepni

Gupta

Güney

et al. 2020

Journal of Forecasting

View full text Add to dashboard Cite

show abstract

The three-pass regression filter: A new approach to forecasting using many predictors

Cited by 278 publications

References 37 publications

Does a lot help a lot? Forecasting stock returns with pooling strategies in a data‐rich environment

Does a lot help a lot? Forecasting stock returns with pooling strategies in a data‐rich environment

A Unified Framework for Dimension Reduction in Forecasting

Forecasting local currency bond risk premia of emerging markets: The role of cross‐country macrofinancial linkages

Contact Info

Product

Resources

About