Is Japan’s Government Pension Fund Too Big?

Japan’s GPIF is big. How big? Weighing in at $1.4 trillion, it’s ten times the CPPIB, six times CALPERS, and two times Norway’s GPF-G. Is that too big?


Japan’s Government Pension Investment Fund (GPIF) is big. How big? Well, weighing in at $1.4 trillion, it’s ten times the size of the Canada Pension Plan Investment Board (CPPIB), six times the size of the California Public Employees Retirement System (CalPERS), and two times the size of Norway’s Government Pension Fund – Global (GPFG). So, yeah, the GPIF is big. In fact, it’s the world’s largest institutional investor.

The question, then, is what this size means for the management of the fund. After all, you might be thinking to yourself, “Gosh, a fund this big should be able access the best PE managers in the world! It could grind away on fees and better align interests between itself and its external managers.”

You’d only be partially right.

It’s true that the GPIF can use its size to extract concessions in terms of fees; the average management fees the fund pays are in the low single basis point range. But the fund doesn’t play in the alternatives space at all. Nearly the entire portfolio is managed passively in public markets. This is due, in part, to its mandate, governance and staffing, but it’s also a function of its size.

As Nobusuke Tamaki argues in an interesting new paper on the GPIF, there is a belief that the fund would swamp private markets if it allocated assets there. Consider that a 13% allocation by GPIF to alternatives (which is the average value among funds over $5 billion) would represent an allocation of $180 billion; that’s larger than the entire pool of capital managed by the CPPIB. So, the GPIF has a point here.

(And while you might say that the fund could allocate a smaller portion, is it worth building up the internal capabilities when you’re not allocating enough to even move the dial on the fund’s total return? The governance nerd in me says no.)

Anyway, here’s a blurb from Tamaki’s paper:

“In an illiquid market such as private equity, however, the marginal rate of return is likely to fall as investment increases, since one must then invest in less promising deals. In other words, the law of diminishing rate of returns works more strongly against a large investor in a more illiquid market, where fund size does matter.”

You may be more right than you know...

New research by the Pension Consulting Alliance (prepared for CalSTRS) shows clearly that there are diseconomies to scale in the private equity business. So if the size of your fund restricts you to only the biggest PE managers in the world, as it would for GPIF, you may be better off going the route of public equities, especially when you net fees out.

In short, GPIF’s size is a big constraint on its ability to execute an “ideal” investment strategy for a pension reserve fund. And I find that interesting given the debate that Sweden is currently having over the segregation of its national pension capital among a handful of different agencies. In the Swedish case, there is a push to bring the AP Funds, which have roughly $20 billion each in AUM, into a single organization that can take advantage of the economies of scale.

So if $1.4 billion is too big, and $20 billion is too small... what’s the ideal size of an institutional investor? In other words, at what size can a fund build a competent internal team and execute at the highest levels?

I’d say the range is $50 billion to $400 billion. Any smaller and you don’t have the AUM to warrant a big ramp up in internal capabilities. Any larger and it isn’t worth the effort to build up alternatives teams because of the limited capacity in those markets. So given the current size of the GPIF, the management is probably right to focus on public markets. But perhaps the pensioners in Japan would be better off if the GPIF was broken up into a series of smaller funds...