Comment on “Public Pension Fund Management: Best Practice and International Experience”

• Published date: 01 July 2015
• DOI: http://doi.org/10.1111/aepr.12107
• Date: 01 July 2015
• Author: Peter R. Orszag

Comment on “Public Pension Fund

Management: Best Practice and

International Experience”

Peter R. ORSZAG†

Citigroup and Bloomberg

JEL codes: F21, G11, G23, H55, J26

Rozanov (2015) makes two core points: first,that pension funds should be professionally

managed, and second that in practice the Canadian model of such management is attrac-

tive for multiple reasons. Except for those who would explicitly favor incompetence, the

first point is, I would think, relatively uncontroversial. So I would like to focus my com-

ments on the second point.

Before examining the performance of a public pension fund that has invested in

asset classes beyond bonds, it is important to revisit whether funds should be invested in

such assets in the first place. A classic examination of this issue in the context of private

pension funds is Bodie (1988); Broadbent et al. (2006) provide an updated summary of

the arguments. A key,lingering question is whether the higher expected return associated

with other asset classes justifies the additional risk (especially given the effect of interest

rate changes on the liability side of a pension fund; a bond portfolio on the asset side

provides a natural hedge). Rozanov correctly argues that this type of question needs to

be evaluated by each fund, according to its investment philosophy, and that pension

funds may have different perspectives on this question than endowments.1

A second, more problematic issue involves the comparison of returns on portfolios

with different weights assigned to various asset classes, which drives the policy recom-

mendation of following the Canadian model. Rozanov’s conclusions about the superior-

ity of the Canadian model relative to Norway rest heavily on Ambachtsheer (2012),

which is a four-page newsletter with little detail. Table 1 reproduces some of the data in

Ambachtsheer’s (2012) table.

Table 1 seems somewhat underwhelming, in and of itself, as the basis for declaring

that the Canadian model is a more compelling alternative to the Norway model. To be

sure, Rozanov updates the figures,as shown in table 3 of his paper. That update, however,

merely underscores the problems involved in the comparison.

First, it is challenging to construct an adequate benchmark portfolio for illiquid,

private investments, which have historically been more dominant in the Canadian

model. This is reflected in the much higher tracking errors for the Canadian portfolio.

The figures presented as “excess returns” are thus not directly comparable with one

another.

†Correspondence: Peter R. Orszag, Citigroup and Bloomberg, 388 Greenwich Street, New York,NY

10013. USA. Email: orszagp@gmail.com

bs_bs_banner

doi: 10.1111/aepr.12107 Asian Economic Policy Review (2015) 10, 296–298

© 2015 Japan Center for Economic Research296