Autocorrelation, Investment Horizon and Efficient Frontier Composition

This paper analyzes how the skewness and kurtosis of securities' returns are affected by the length of the differencing interval over which returns are measured. Hawawini's previous analysis of this “intervaling effect” on log returns is shown to be incorrect, and the correct effects are derived. While Hawawini only considered log returns, we also derived the intervaling effect on the skewness and kurtosis of “simple” returns. Our results show that the length of differencing interval has very different effects on log and simple returns, but in both cases the effects on the returns' skewness and kurtosis are substantial and also quite tractable. These results also enable us to reconcile some of the contradictory results published earlier on the intervaling effect.

show abstract

“…The validity of equations (14) and (15) has been confirmed with simulations similar to the one described in Table 1.…”

Section: Intervaling Effect On Moments Of Simple Returnssupporting

confidence: 58%

The Analytics of the Intervaling Effect on Skewness and Kurtosis of Stock Returns

Lau

Wingender

1989

The Financial Review

View full text Add to dashboard Cite

show abstract

“…The reported t-statistics indicate that all the estimated slopes b i are significantly greater than one, and that 40 percent of the estimated intercepts are significantly negative at the 5 percent level of 'Comparing eqs. (7) and (9) we see:…”

Section: The Relationship Between the Intertemporal Cross Correlatmentioning

confidence: 89%

“…Also, it is evident from eqs. (7) and (9) that while b i contains an unweighted average of the q-ratios, a i contains a weighted average of the same ratios. In ai' the higher order q-ratios have higher weights and this has the effect of magnifying the value of small and probably insignificant q-ratios of higher orders.…”

Section: The Relationship Between the Intertemporal Cross Correlatmentioning

confidence: 99%

“…Last, the existence of intertemporal cross dependence among securities' returns will produce significant noncontemporaneous (lead or lag or both) systematic risks for individual securities as well as for well-diversified portfolios, if daily returns are used, as reported by Hawawini and Vora [13], [14]. 7 As securities with larger market values, and hence lower q-ratios, tend to be listed on the NYSE, on average the NYSE securities appear to lead the AMEX securities. The list of phenomena discussed in this section is not meant to be exhaustive and the intertemporal cross dependence among the daily returns of securities may be responsible for other phenomena not surveyed here.…”

Section: Some Implications For Empirical Work In Financementioning

confidence: 99%

“…See Gilster [7] for the effect of correlation on the shape of the efficient frontier, Schwartz and Whitcomb [19] for the effect on the market model residuals and coefficient of determination, and Hawawini [9] for the effect on the moments of the distributions of securities' returns and the evaluation of the investment performance of institutional investors.…”

mentioning

confidence: 99%

See 2 more Smart Citations