According to the CFA Institute, a standards-setting organization for finance professionals that provides guidelines for benchmarking, the move did not break the rules.

But some experts, including Knut Norheim Kjær, the founding CEO of the Government Pension Fund of Norway — now one of the largest pension funds in the world — say the benchmark change is concerning because CPPIB’s managers personally benefitted as a result, which is why many national pension organizations do not set their own yardsticks. 

“You could say that they get better numbers with the new one,” Kjær told the Star. 

“They are able to by themselves change the measurement standard and to pay themselves accordingly,” he added. “That is not the normal integrity in such a business.”

CPPIB spokesperson Michel Leduc said that the pension fund has been evolving its performance-measurement framework over several years as the fund became more diversified.

“It would be misleading to imply that the benchmark portfolios were adopted to pay management more for value-added performance,” he said in an emailed statement to the Star, adding that the benchmark switch was approved by the CPPIB’s board. 

Leduc said that it would also be wrong to suggest that the new benchmark portfolios created “an easier hurdle” for the CPPIB. 

The organization expects the new benchmark portfolios to have “slightly higher long-run absolute returns” than the former reference portfolios, which were a simple mix of equities and bonds, due to diversification benefits, according to Leduc. 

The CPPIB’s 2026 annual report is expected to be released later this month.

The reference portfolios are made up of 85 per cent stocks and 15 per cent bonds, representing returns fund managers could have earned had they followed a passive or “hands-off” investment strategy. 

But, last year, the CPPIB began assessing its performance against a more complex benchmark crafted internally and made up of over 100 indices. CPPIB stated that it now measures value-added performance against the benchmark portfolios for all periods starting in 2016. 

As a result, relative performance appears to have improved.

The fund underperformed both the old and new benchmarks over the past five years, but it underperformed by less under the new benchmark.

The relative five-year annualized return under the new benchmark was -0.75 per cent, versus approximately -4.2 per cent compared to the previous one, according to the Star’s calculations.

After introducing the new benchmark, the CPPIB applied it to past years to calculate performance-based compensation in 2025, as first noted by Amir Barnea, an associate professor of finance at HEC Montréal. 

Claude Lamoureux, former CEO of the Ontario Teachers’ Pension Plan, said he is “disappointed” by CPPIB’s retroactive application of the benchmark.

“To me, that’s something that should not have been done,” he said, adding that this is not common practice among large pension funds.

“It’s human nature to want to say, ‘well, we’re just as good as the benchmark,’ when sometimes you have to say ‘no, we could have done better.’”

He also raised questions about the timing of the change, which was first announced in 2024.

“This was their worst year,” he said. “So why did they not change in 2023 when they were close to the benchmark?”

Employees at the CPPIB are awarded a base salary plus incentive compensation, which reflects, in part, how the fund fared relative to the benchmark over the last five years. 

In 2025, CEO John Graham saw his pay increase by $1.3 million, some of which was performance-based, taking home a total of $6.4 million. 

At the same time, total personnel expenses, including salaries, incentives and benefits, went up by $79 million, partly due to higher incentive compensation, according to the CPPIB’s 2025 report. 

“I can’t understand why the CPPIB gets to internally pick their own index, and why this isn’t done by completely unrelated, unbiased third parties outside the organization,” said Moshe Arye Milevsky, a professor of finance at the Schulich School of Business at York University.

“In fact, I really don’t understand why they don’t invest in the index itself, which would save (Canadians) millions a year.” 

Leduc said that while CPPIB’s managers have historically used the reference portfolios to measure value added, they have maintained for years that actually investing in a passive “85/15” reference portfolio is an unrealistic “do-nothing” alternative to the active approach taken in the current fund.  

“Doing so would leave the fund dominated by equity risk,” he said, adding that would be “reckless.” 

Given that the former reference portfolios don’t reflect how the fund is actually invested, CPPIB introduced a new benchmark to align with its long-term investment strategy without undue risk of loss, said Leduc.

The new benchmark portfolios don’t just track stocks and bonds, but also other asset classes, such as real estate and private equity. The company still uses the reference portfolios as a target for risk. 

He added that relative performance is just one component of overall compensation, accounting for 12 per cent of total incentive compensation for most employees. That percentage is higher for the senior management team, according to CPPIB’s 2025 annual report.

Employees are also evaluated at an individual and department level.  

Generally, investment managers should change their benchmark going forward only and not retroactively, according to the CFA Institute’s guidelines. 

However, retroactive changes are allowed with proper disclosure if the new benchmark better reflects the asset owners’ investment strategy, said Beth Kaiser, director of global industry standards at the CFA Institute. 

“According to the standards, the most important thing is that the performance presentation is not false or misleading, and that it’s transparent,” she said. “It seems as though, at least in the 2025 annual report, the disclosure is very clear.” 

Kaiser added that the type of benchmark CPPIB switched to, which represents different asset classes, is the most commonly-used benchmark by investment managers, including pension funds, endowments, foundations, and sovereign wealth funds.

Other pension experts interviewed by the Star agreed that the new benchmark is better suited to measure CPPIB’s relative performance. 

“If you want to assess the quality of your investment managers, you need a benchmark that reflects the investment strategy pursued,” said Eduard van Gelderen, former chief investment officer at PSP Investments.

“As long as the investment strategy has not changed over time, I don’t think there is anything wrong with applying it to historical performance data. It only provides more insights (on) how well the investment managers have done over time,” he added.

“My only concern would be why this wasn’t done before.”  

Campbell Harvey, professor of finance at Duke University, said he believes the CPPIB should be subjected to an independent external review every four years to avoid “the perception of conflicts of interest.”

“They can look at a broad range of items, including performance, including perhaps modifying the benchmark,” he told the Star. “A benchmark should be externally mandated, so part of this committee would be to establish a benchmark portfolio, and that would be ideal.”

Leduc emphasized that the CPPIB uses a board-approved “benchmark architecture” tied to the institutions’ obligations and overseen by risk, audit and public-reporting mechanisms. 

“Our governance disclosure clearly describes how management develops benchmarks that objectively measure the markets and strategies in which it invests, but the board assesses and approves the methodology and receives reporting on the composition,” he said. 

“That is not management marking its own homework. It is management proposing and an accountable oversight body approving.”

In Norway, the ministry of finance sets the benchmark. Kjær discussed the issue in a presentation on global best practices of long-term investing. 

The Norwegian fund is “being measured all the time against a yardstick that is constant and given from outside,” he said. 

“You don’t give a benchmark to yourself. Other people give you a benchmark.”