The board of California Public Employees’ Retirement System – with a few trepidations – officially approved the transition to its long-anticipated total portfolio approach at a meeting this week, providing staff more discretion in investment decisions.

For the past year, CalPERS staff have pitched the TPA model as offering added flexibility and an opportunity to focus on investments that can best contribute to performance, instead of those that best fit pre-set allocation targets for asset classes.

Stephen Gilmore, the system’s CIO, told the board at a November 17 investment committee meeting attended by Buyouts that the change avoids “having each asset class investment team filling its bucket without reference to other opportunities.” The TPA model was used by the NZ Super Fund during Gilmore’s previous post there as CIO.

Gilmore indicated that this model will be crucial as the system builds exposure to private assets to avoid another “lost decade,” alluding to the years immediately post-financial crisis when the pension maintained significant under-allocations to private equity.

“We do want to avoid [that], and we want to be consistent,” he said. “Having those consistent flows is important. The reality is we don’t know when we’re going to find opportunities. Sometimes, there’s more than we expect. So, there has to be a be a degree of flexibility for investing teams.”

Although CalPERS cemented its status as the first institutional allocator in the US to adopt the total portfolio approach at its November 17 meeting, the fund’s current strategic asset allocation model won’t be replaced until July 2026.

Under the prior model, the board set specific asset allocation targets every four years, with an intervening mid-cycle check-in allowing adjustments of targets. Each asset class also had its own benchmark.

With the transition to the total portfolio approach, the system is eschewing specific targets for its portfolio and consolidating its 11 benchmarks to a single benchmark. That benchmark will act as a reference portfolio, equivalent to a generic mix of 75 percent equities and 25 percent bonds.

One of the concerns expressed by the board was whether that represented more risk for the pension than was warranted in an unsteady market.

“One of the challenges of the past was being procyclical – taking risk up at the wrong time or down at the wrong time,” Gilmore said. “I know people will say things look relatively expensive right now, but they can remain expensive for a long time. Often, when things look overvalued, they stay overvalued for years.”

Board member Frank Ruffino asked that the pension’s staff attach to the TPA approval a one-time review to determine “whether to continue, modify or revert any elements of the TPA governance framework.” He argued the outcomes of this new model should be weighed against the previous structure within no longer than two years.

That motion ultimately failed, with Gilmore advising the board that the suggested review would be too soon for the plan to make any determinations about performance.

Earlier in the meeting, he pointed to a survey from research group Thinking Ahead Institute earlier this year that surveyed 26 funds utilizing TPA and found investments outperform peers under other models by 1.3 percent per year over a 10-year period.

Considering the small subset of systems operating this way, Gilmore noted to the board that those numbers can be challenged.

“So, I wouldn’t rely on that too much,” he said. “But, conceptually, if you think about optimizing the portfolio as a whole, it should be superior to optimizing asset class by asset class.

“But it requires sufficient collaboration across the team. And I think we’ve got the organization to do that.”

Read more @buyoutsinsider