Panel Data Inference in Finance: Least-Squares vs Fama-Macbeth

Skoulakis, Georgios

doi:10.2139/ssrn.1108865

Cited by 31 publications

(27 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…An alternative non-regressionbased technique is the FamaMacBeth estimator. Skoulakis (2008) presented simulation evidence that in cases like ours (a large cross section and relatively small time series), both techniques yield reliable estimates of standard errors.…”

Section: As Of December 2006 Pensions and Investments Reported Thatmentioning

confidence: 65%

Absence of Value: An Analysis of Investment Allocation Decisions by Institutional Plan Sponsors

StewartScott

NeumannJohn

KnittelChristopher

et al. 2009

Financial Analysts Journal

View full text Add to dashboard Cite

To determine whether the investment decisions of institutional plan sponsors contribute to their asset values, this study used a dataset of 80,000 yearly observations of institutional investmentnstitutional plan sponsors are responsible for pension plans, endowments, foundations, and other large pools of assets. These assets are huge (estimated to be more than $10 trillion in December 2006), yet the research on institutional investors is limited. Heisler, Knittel, Neumann, and Stewart (2007) documented the importance of historical performance measures, performance trends, and product attributes in shaping plan sponsors' decisions to allocate assets among professional money managers' investment products. Del Guercio and Tkac (2002) and Lakonishok, Shleifer, and Vishny (1992) also studied the institutional investment decision-making process. The question remains, however, whether these decisions add value for the ultimate beneficiaries or stakeholders on whose behalf the plan sponsors act. This question is important because of the size of institutional plans and their sponsors' sophistication relative to that of individual investors. Pension plans, endowments, and foundations are typically staffed with professionals who have advanced degrees and years of experience. Working on their own or with the aid of consultants, institutional sponsors devote considerable time and resources to selecting asset classes and products that are expected to perform well.The central goals of our study were to assess the success of these efforts by exploring the economic significance of allocation decisions and to attribute performance components to product selection, asset class, or style category allocation decisions. Although we did not observe the plans' performance directly, we did observe the allocation of their assets to investment products over time. The data we used are available in the PSN database, which includes 80,000 institutional product observations whose data include annual returns, annual assets, and accounts for 19842007. Heisler et al. (2007) determined that long-term total returns and a track record of consistently positive or negative benchmark-relative returns factor heavily into institutional plan sponsors' decisions to allocate assets to, or pull them from, equity products. They showed that sizable negative short-term total returns play a modest but statistically significant role in decisions to shift assets away from Literature Survey

show abstract

Section: As Of December 2006 Pensions and Investments Reported Thatmentioning

confidence: 65%

Absence of Value: An Analysis of Investment Allocation Decisions by Institutional Plan Sponsors

StewartScott

NeumannJohn

KnittelChristopher

et al. 2009

Financial Analysts Journal

View full text Add to dashboard Cite

show abstract

“…1969-1978Graph B. 1979Graph C. 1989Graph D. 1999-2008 In Figure 2, R 2 is derived from regressing, for each iteration, ITPCSR factor portfolios on the 3 factors from the one-step APC procedure (plus a constant). The sample consists of 1,290, 2,218, 2,961, and 3,225 firms with complete data available on the CRSP monthly data file for the 10-year periods, Jan. to Dec.…”

Section: A Balanced Panel Of Asset Returnsmentioning

confidence: 99%

“…1969-1978Graph B. 1979Graph C. 1989Graph D. 1999-2008 For the nonconverging subperiod, we calculate the least squares objective function, Λ, from equation (10) for the two ITPCSR estimates (starting with the sunspot "factors" and the Fama-French (1993) In Figure 5, R 2 is derived from regressing, for each iteration, ITPCSR factor portfolios on the 3 factors from the one-step APC procedure (plus a constant). The sample consists of 5,680, 8,884, 11,710, and 10,221 firms available on the CRSP monthly data file for the 10-year periods, Jan. to Dec.…”

Section: Unbalanced Panel Of Actual Asset Returnsmentioning

confidence: 99%

“…The sample consists of 5,680, 8,884, 11,710, and 10,221 firms available on the CRSP monthly data file for the 10-year periods, Jan. to Dec. 1969-1978(Graph A), 1979(Graph B), 1989(Graph C), and 1999-2008, respectively. Graph A.…”

Section: Unbalanced Panel Of Actual Asset Returnsmentioning

confidence: 99%

“…TPCSR has been applied by many researchers to the estimation of factor returns and testing of asset pricing models (some examples, in addition to those above, are Fama and French (1992), Lettau and Ludvigson (2001), and Jagannathan and Wang (2007)). Econometric analyses of the properties of the TPCSR estimator include Shanken (1982), (1983), (1992), Jagannathan and Wang (1998), Kan and Zhang (1999), Skoulakis (2008), Kleibergen (2009), Jagannathan, Skoulakis, and Wang (2010), and Lewellen, Nagel, and Shanken (2010).…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

A Synthesis of Two Factor Estimation Methods

Connor

Korajczyk²,

Uhlaner

2015

J. Financ. Quant. Anal.

View full text Add to dashboard Cite

Two-pass cross-sectional regression (TPCSR) is frequently used in estimating factor risk premia. Recent papers argue that the common practice of grouping assets into portfolios to reduce the errors-in-variables (EIV) problem leads to loss of efficiency and masks potential deviations from asset pricing models. One solution that allows the use of individual assets while overcoming the EIV problem is iterated TPCSR (ITPCSR). ITPCSR converges to a fixed point regardless of the initial factors chosen. ITPCSR is intimately linked to the asymptotic principal components (APC) method of estimating factors since the ITPCSR estimates are the APC estimates, up to a rotation.

show abstract

A statistically robust decomposition of mutual fund performance

Agnesens

2013

Journal of Banking & Finance

View full text Add to dashboard Cite

Panel Data Inference in Finance: Least-Squares vs Fama-Macbeth

Cited by 31 publications

References 24 publications

Absence of Value: An Analysis of Investment Allocation Decisions by Institutional Plan Sponsors

Absence of Value: An Analysis of Investment Allocation Decisions by Institutional Plan Sponsors

A Synthesis of Two Factor Estimation Methods

A statistically robust decomposition of mutual fund performance

Contact Info

Product

Resources

About