CPP Investments changed how it measures its performance in fiscal year 2025. That same year, it changed the way it pays its executives. While the pension manager pitched the benchmark change as a necessary adjustment to its investment strategy, critics believe the efforts are a way to obfuscate poor returns while maintaining competitive compensation packages.
For nearly two decades, CPP, which manages the equivalent of around USD 793.3 billion on behalf of Canadian workers, measured itself against “market risk target,” a portfolio of public stocks and Canadian bonds the fund had used since 2006. This portfolio was designed to mimic the performance of a low-cost public-market index, or market beta. Any return above this was outperformance, or alpha.
Starting in 2025, the fund replaced that barometer with several “benchmark portfolios,” or combinations of passive public-market assets that mirror the different strategies in CPP’s portfolio. This approach captures only the alpha generated within each asset class, not the fund’s portfolio construction. Critics view this as a key driver of outperformance.
“CPP’s benchmark portfolios are just saying that CPP is only going to benchmark the alpha piece. Forget all this portfolio construction stuff,” says Alexander Beath, founder of Toronto-based consulting firm Alex Beath & Associates and a longtime advocate for better private-market benchmarking at Canadian pensions. “They just assume that they’re better [than market beta], which I think isn’t right.”
CPP has defended the shift as a natural consequence of adopting the total portfolio approach, which gives investment staff discretion to allocate capital across asset classes within board-set risk limits, rather than abiding by fixed targets for each asset class.
The benchmarking shift arrived alongside a redesign of CPP’s incentive compensation plan. The fund’s total fund performance multiplier, which helps determine staff bonuses, had previously been weighted evenly between two quantitative measures: absolute performance against a return target and relative performance against the old benchmark. In 2025, CPP added a new, qualitative third weighting, where 20% of the multiplier’s formula is tied to progress on “strategic objectives.”
The timing is drawing scrutiny. Under the old benchmark, CPP’s total portfolio missed its target by 11.9 percentage points in FY 2024. Under the new one, performance was better but still fell below the benchmark, missing by 1.6 points in 2025 and 5.4 points in 2026. Executive pay moved in the opposite direction.
In the 2026 performance multiplier, CPP’s investment staff scored 1.05 on absolute performance and just 0.81 on relative performance, the measure most directly tied to the benchmark miss. On the qualitative strategic-objectives component, however, they scored 1.8, the highest of the three. The blended result was an overall multiplier of 1.1, indicating that the strategic score more than offset the shortfall in relative performance.
“This is a giant governance failure,” says Rachel Wasserman, a Toronto-based corporate governance attorney and founder of Wasserman Business Law. She argues that the new benchmark is easier to beat, and that the qualitative overlay gives the board cover to reward staff even when relative returns lag. “They’re missing their benchmarks, and then they introduce this qualitative assessment, which they’re all scoring great on.”
CPP said in a paper that while benchmarks remain “essential tools for attribution, discipline and accountability” under TPA, they are not “singular verdicts on success or failure.” The paper argues that TPA portfolio returns are less correlated to benchmark returns than traditional strategic asset allocation returns. Under TPA, investment staff is responsible for both performance relative to the benchmark and the target asset class, and for risk exposures. This requires them to invest in asset classes such as private equity, which have low correlation with a benchmark of stocks and bonds. The paper claimed performance is thus harder to measure against a single benchmark, and such benchmarks create a “distorted incentive problem,” wherein concentration is rewarded and diversification penalized.
In an email to PitchBook, CPP chief corporate affairs officer Michel Leduc said the benchmark change came from a years-long evolution and was not a reaction to recent underperformance. He also noted that the fund’s 2025 annual report disclosed that the benchmark portfolios have a slightly higher long-run expected absolute returns than the market risk targets.
“The public record shows a multi-year evolution of the Total Portfolio Investment Framework, disclosed in annual reports from fiscal 2022 through fiscal 2025,” Leduc said. “In fiscal 2025, five-year annualized net relative performance against the Benchmark Portfolios was negative 0.75%, below the target of 0.39%, producing a value-added multiplier of 0.17 times for this one portion of the compensation framework. This five-year view directly refutes the suggestion of a self-serving compensation adjustment over the shorter period.”
Beath pushes back on CPP’s framing that TPA adoption necessitated the benchmark change. Instead, he says it’s a product of institutions moving into TPA and needing to “show that they’re good at it.”
Other large pensions moving to TPA have taken a different path. Calpers, the first major US pension fund to adopt the approach, measures its total portfolio against a reference portfolio of 75% equities and 25% bonds—a structure similar to the one CPP just abandoned.
Even Beath, who disagrees with CPP’s new methodology, doesn’t think there’s a clean fix. He says benchmarking portfolio construction is inherently noisy over any observable timeframe: “It’s more like a game of snakes and ladders. There’s a lot of luck involved … Seeing your skill through all that luck is very hard.”
