CPP Investments Looks Beyond Benchmarks to Evaluate Performance

Freschia Gonzales of Benefits and Pensions Monitor reports a winning portfolio can still look like a loser, CPP Investments warns:

CPP Investments argues that beating a benchmark no longer measures whether a portfolio is working.  

Its modelling shows a diversified pension fund can trail its market benchmark almost 30 percent of the time over a decade, even when its underlying strategy genuinely adds value. 

That figure comes from a July 2026 paper from the CPP Investments Insights Institute, the second in a series on the total portfolio approach (TPA).  

In a stylized illustration, the authors modelled two portfolios carrying the same total risk: a strategic asset allocation (SAA) portfolio highly correlated with its benchmark, and a TPA portfolio built to diversify away from it.  

Even with a higher expected value-add of 100 basis points, the diversified portfolio showed a 29.8 percent chance of underperforming its benchmark over 10 years, the report said, compared with 6.5 percent for the more benchmark-hugging design.  

These are false negatives, cases where a sound portfolio looks like it failed. 

The paper frames the problem as one of accountability outgrowing its measuring stick. 

Under traditional SAA, the report explained, a board sets a policy portfolio and management is judged on whether it beats that benchmark.  

Under TPA, management instead owns a broader set of interconnected choices, including how much risk to take, how to diversify, how to balance liquidity, and how to implement, and all of those decisions fall within the scope of performance assessment. 

CPP Investments set out three reasons a single benchmark falls short.  

First, the authors noted, short-term outperformance or underperformance often reflects luck rather than skill, a problem amplified when a portfolio is intentionally built to differ from its index.  

Second, benchmarks drift: as market concentration rises, capitalization-weighted indices can take on exposures managers never intended to hold.  

The report pointed to research by Sorensen, Alonso and Belanger describing this as a benchmark's "chameleon" nature, which it said can reward concentration when concentration is winning and penalize diversification held for long-term resilience.  

Third, as portfolios add private assets, illiquidity premia and long-duration cash flows, comparisons to public-market indices become, in the report's words, "apples-to-oranges." 

Diversification's recent optics feature prominently.  

Over the past few years, the upper deciles of feasible portfolio returns have been dominated by heavy exposure to US tech mega-cap listed equities, according to the paper, leaving diversified designs in the lower portions of the distribution.  

That does not mean diversification failed, the authors argued, but rather that portfolios built for many environments will lag concentrated, equity-heavy portfolios in a market that rewards listed equities far beyond expectations. 

For plan sponsors weighing whether a differentiated strategy is paying off, the report offered CPP's own sustainability record as evidence on its risk-setting decision. 

Investment performance is estimated to have cut the base CPP minimum contribution rate from 9.54 percent in the 31st actuarial report to roughly 9.19 percent, a fourth consecutive triennial decline. 

Measured from 2012, the rate has fallen from 9.84 percent despite adverse demographic surprises that would normally push it higher.  

The Spring Economic Update 2026 proposed reducing the legislated base statutory contribution rate from 9.9 percent to 9.5 percent starting in 2027, the report noted, citing the improved funding position.  

CPP Investments targets a level of market risk equal to a portfolio of 85 percent global equities and 15 percent Canadian government bonds for the base plan, and 55 percent equities to 45 percent bonds for the additional plan. 

Rather than replace benchmarks, the report positioned them as a diagnostic input within a six-part framework spanning total return outcomes, risk and capital allocation, portfolio construction and diversification, investment selection, decision quality and process, and resilience across market regimes.  

Benchmark outperformance does not always deliver institutional aims, the authors wrote, since a strategy can beat its index simply by adding concentration or exposures already held elsewhere. 

The challenge reaches beyond pension investing, CPP Investments said, as more organizations manage portfolios against multiple long-horizon goals such as inflation protection and sustainability.  

For boards, the paper's takeaways were to align evaluation with delegated decision rights, connect outcomes to objectives, and treat benchmarks as one part of an overall assessment.  

The question, the authors concluded, is no longer whether a portfolio beat its benchmark but whether every management decision improved the delivery of long-term goals. 

The paper was written by Sally Shen, manager at the Insights Institute; Derek Walker, managing director of total fund transformation initiatives; and Geoffrey Rubin, senior managing director and one fund strategist.   

Sally Shen, Derek Walker, and Geoffrey Rubin of CPP Investments wrote a second report entitled Measuring What Matters: Evaluating the Total Portfolio Approach:

For decades, institutional investors operated in a relatively stable world. Inflation was subdued, globalization deepened, and public markets expanded. In this environment, traditional Strategic Asset Allocation (SAA) frameworks offered a practical way to organize and measure investment decisions: establish a “policy portfolio”, measure total-fund returns against it, and assess active managers relative to their benchmarks.

That world is changing.

Economic fragmentation, heightened geopolitical uncertainty and a growing need for resilience across a wider range of market conditions have led many investors to adopt variants of the Total Portfolio Approach (TPA). As we explored in a previous report, TPA offers greater flexibility to construct portfolios around long-term objectives by managing risk, diversification, liquidity, and capital allocation at the total-fund level rather than within asset-class silos.

Yet increasingly, the way portfolios are managed under TPA no longer aligns neatly with how performance has traditionally been measured. Under TPA, management becomes accountable for a much broader set of interconnected decisions: how much risk to take, how to diversify exposures, how to balance liquidity and illiquidity, and how to implement those choices efficiently over time.

This leaves a fundamental question: how should realized performance be evaluated when the portfolio itself is increasingly dynamic, differentiated, and intentionally designed around multiple objectives? This question is especially important for pension funds that are accountable for delivering outcomes over decades—for CPP Investments, over a 75-year actuarial horizon—while providing full transparency and appropriate scrutiny.

Benchmark comparisons remain essential to evaluating active management decisions and maintaining accountability. But under TPA they are no longer sufficient on their own. Measuring the health of a total portfolio requires a broader, multi-faceted framework capable of assessing not only returns, but also the total portfolio’s resilience to adverse market and economic circumstances, the impacts of diversification, concentration and tactical decisions, the effectiveness of implementation, and alignment with long-term objectives. This report examines how CPP Investments is developing its approach.

The Limits of Benchmarks in TPA Performance Assessment

For all the enthusiasm around TPA, its credibility ultimately rests on a simple question: has it worked? Indeed, if institutions cannot assess whether an approach has worked in the past, it is difficult to maintain conviction that it will work in the future.

Traditional SAA frameworks split the overall investment decision into two parts. First, the portfolio owner or board—often with the assistance of a third-party consultant—chooses a policy portfolio of asset classes and weights, with the returns forming the benchmark. Management is then tasked with outperforming that benchmark. This approach provides a straightforward view of management performance: value added relative to a low-cost, investable benchmark.

However, one of the most important decisions—the choice of the policy portfolio itself—typically escapes ongoing evaluation. This decision is stored in the figurative “attic” while performance assessment is focused solely on management’s deviations from the policy and active management benchmark portfolios, rather than the policy portfolio’s suitability for achieving the institution’s objectives (Figure 1).

TPA changes this accountability structure. Management becomes accountable not only for active decisions relative to a benchmark, but for the entire portfolio—from risk choices through targeting of total portfolio exposures, to implementation. All portfolio decisions fall within the scope of performance assessment and must be evaluated based on how well they advanced institutional objectives. This broader accountability requires tools that extend beyond the single benchmark framework typically employed by SAA.

There are good reasons why benchmarks are deeply embedded in institutional investing. First, they constitute simple, unambiguous instructions from principal to agent. Coupled with an active risk limit and a capital allocation, the managing agent understands that their task is to beat the benchmark. And this same benchmark can be used for ex post performance evaluation and incentive compensation, unifying both measurement and management in a single, simple device.

Unfortunately, these features are insufficient for effective performance information and management under TPA.

Issue 1: Noisy Signals

Even genuinely value-additive investment strategies can underperform a benchmark due to randomness—sometimes for extended periods. While this challenge exists under traditional SAA frameworks, it is more pronounced under TPA, where portfolios may intentionally differ from conventional benchmarks in pursuit of diversification, resilience or dynamic positioning.

Barras, Scaillet, and Wermers (2010) show that benchmark-relative outperformance and underperformance often reflect luck over shorter horizons. Barras, Beath, and Betermier (2026) further highlight how noise accumulates along the total portfolio value chain, increasing the likelihood that value-adding diversification decisions appear unsuccessful over finite horizons.

Figure 2 illustrates this challenge. Because TPA portfolios intentionally depart from conventional benchmarks, short-term benchmark-relative underperformance remains plausible even when portfolio decisions are sound and expected long-term outperformance is intact. In short, performance relative to a benchmark provides important evidence, but rarely a complete assessment of whether the total portfolio is succeeding.

This figure provides a stylized illustration of benchmark underperformance across evaluation horizons. The ellipses represent the one-standard-deviation confidence region of 5-year outcomes for the “SAA” and “TPA” portfolios with the same total risk target. The dot within the ellipse represents the expected outcome.

The “SAA” portfolio in the left-hand panel is assumed to have a value-added expectation of 90 bps (the vertical distance or “alpha” to the diagonal line), delivered with relatively narrow error bands due to its lower tracking error versus the benchmark. The “TPA” portfolio in the right-hand panel is assumed to have a value-added expectation of 100 bps, but larger error bands due to substantially lower correlation of its returns with the benchmark, arising from differentiated portfolio construction and diversification decisions. The shaded area below the diagonal line represents benchmark underperformance despite positive true alpha (“false negatives”).

Over a single year, both the “SAA” and “TPA” portfolios show significant incidence of underperformance to the benchmark (see the table below Figure 2). At five- and 10-year horizons, the compression of the “SAA” error bands makes false negatives increasingly unlikely. By contrast, while the “TPA” error bands also compress over time, realized benchmark underperformance remains plausible even at longer horizons despite higher alpha expectations, as false-negative probabilities remain elevated. Importantly, Figure 2 is not intended to suggest that either the SAA portfolio or the TPA portfolio is inherently superior. Rather, it illustrates a measurement challenge that arises when portfolios are intentionally diversified away from a single benchmark.

Issue 2: Benchmarks That Do Not Reflect Intended Portfolio Design

A second challenge is that benchmarks can evolve significantly over time. As market concentrations shift, capitalization-weighted indices can come to embody risk exposures that differ materially from those that portfolio managers intentionally seek to maintain. This matters because TPA practitioners typically aim to diversify away from concentrations that accumulate in widely used benchmarks.

Sorensen, Alonso, and Belanger (2023) describe this as the benchmark’s “chameleon” nature: as market concentration rises, diversification falls, and tight tracking-error constraints can force managers to hold portfolios that are more concentrated than intended or prudent.

This creates a distorted incentive problem. Benchmark-relative evaluation can reward concentration when concentration is winning and penalize diversification that is being maintained to support long-term resilience. Over time, this can pull portfolios closer to the benchmark—even when doing so undermines the broader objective of diversification and total portfolio resilience.

Issue 3: Apples-to-Oranges Comparisons

As portfolios become more differentiated, comparisons to any single benchmark become less informative. At CPP Investments, portfolio outcomes reflect exposures to illiquidity premia, long-duration cash flows, operational value creation, and other risk factors that are not fully captured by public market benchmarks. The result: simple public benchmark-relative performance is an apples-to-oranges comparison that can mislead as much as it informs.

This limitation is particularly evident in private assets such as infrastructure, where public benchmarks often fail to capture contractual cash flows, inflation linkage, and long-term active management of the portfolio assets. As Shen and Blanc-Brude (2022) demonstrate, these characteristics are difficult to replicate through conventional public-market benchmarks. As a result, private assets must ultimately earn their place in the portfolio relative to public-market alternatives. However, simple benchmark comparisons may not fully capture their contribution.

These benchmarking issues, and expanded accountability for the entire investment problem, make assessment of TPA performance difficult but not impossible. Ex post performance assessment remains vital to sustaining confidence and improving decisions over time.

The challenge, then, is assessing performance through multiple lenses rather than via a single verdict.

Moving Beyond Benchmarks

Ex post performance assessment under TPA begins with a clear statement of institutional objectives. For some organizations, the primary objective is pension plan sustainability (e.g., funded status or contribution rate stability); for others, objectives may include liquidity preservation, inflation protection, intergenerational risk-sharing, or climate and other broad policy considerations. A credible ex post assessment must evaluate success against each of those ex ante objectives while also considering whether the decisions that produced the outcomes were sound given the information and alternatives available at the time.

CPP Investments’ objectives are grounded in the Canada Pension Plan Investment Board Act¹:

• To invest CPP assets with a view to achieving a maximum rate of return, without undue risk of loss, having regard to the factors that may affect the funding of the Canada Pension Plan;
• To manage the fund in the best interests of CPP contributors and beneficiaries;
• To assist the CPP in meeting its obligations to contributors and beneficiaries.

While these objectives are clear, they are multi-purpose and multi-horizon. They must be translated into a more complete set of aims that define risk appetite, outcome horizons, and the desired level of portfolio resilience through market and economic cycles. Clarifying and balancing those aims, and the constraints within which they must be pursued, is a prerequisite for assessing TPA performance.

TPA performance is best understood by examining the many decisions that collectively produce total portfolio outcomes. Decomposing performance into its component decisions does not diminish management’s accountability for overall results, but it does provide useful evidence on the performance of each major link in the chain. Traditional benchmark-based frameworks assess only parts of the investment problem—typically the active components relative to benchmarks—while disregarding the rest.

In practice, CPP Investments undertakes many investment decisions in pursuit of its objectives. For the purposes of this report, we are focusing on three that have an outsized influence on long-term outcomes—risk level, exposure targeting, and investment selection. Together, these decisions shape portfolio design, realized returns, and delivery against objectives, though in different proportions (Figure 3).² Here, we discuss each of these decisions and the approach to understanding their performance, noting that this approach can be extended to a wider range of decisions and institutional aims.

Decision 1: Risk Targeting: Did the chosen level of market risk optimize plan adjustment risk?

The first major TPA decision at CPP Investments is the level of market risk the Fund should target. CPP Investments establishes a level of market risk that optimizes the combination of positive and adverse adjustments to the CPP that might arise from our realized investment performance.

Lower-risk portfolios reduce short-term investment return volatility but may deliver insufficient long-term returns to support CPP sustainability. Higher-risk portfolios increase expected returns but also increase the likelihood of adverse outcomes in the short-term. Risk targeting therefore reflects a trade-off between short-term volatility and long-term plan sustainability.

CPP Investments manages this trade-off by targeting a level of market risk that best balances positive and adverse potential plan adjustments over time.  In this model, adverse outcomes carry more weight than positive outcomes and nearer term outcomes carry more weight than distant ones.  For the base CPP, this work establishes a target level of market risk equal to that of a portfolio comprised of 85% global equities and 15% Canadian government bonds; and for the additional CPP an equity-debt risk equivalence of 55% global equities and 45% Canadian government bonds.³ These levels of market risk generate expected returns that exceed the minimum level required for CPP sustainability, reducing the likelihood of adverse plan adjustments and improving the likelihood of positive adjustments over time.

Has this decision worked? Figure 4 illustrates how these returns have impacted base CPP sustainability and contributed to sustained reductions in the Minimum Contribution Rate (MCR).⁴ Most recently, investment performance is estimated to have reduced the MCR from 9.54% in the 31st OCA Actuarial Report (2021) to approximately 9.19%,⁵ marking the fourth consecutive triennial decline. The Spring Economic Update 2026 proposed reducing the legislated base CPP Statutory Contribution Rate from 9.9% to 9.5%, beginning in 2027, reflecting the improved sustainability position of the CPP following the latest triennial review.⁶

Decision 2: Exposure Targeting: Did portfolio design improve the investment outcome?

Portfolio design is the second major TPA decision at CPP Investments. It includes choices related to diversification, leverage, geography, currency, sector, and other systematic risks, while maintaining the total risk established through risk targeting. There are many ways to design a portfolio at the same overall risk level; CPP Investments’ objective is to identify the portfolio most likely to deliver objectives across a wide range of future states, rather than maximize performance for any single objective or in any single market environment.

Under TPA, there is no natural “default” portfolio. A benchmark, whether a simple blend of 85% global large- and mid-cap equities and 15% Canadian government bonds, an equal-weight index, or something more customized, is only one feasible design among many. Portfolio design should therefore be evaluated against a broad collection of feasible alternatives that achieve the same risk target through different portfolio construction approaches, resulting in materially different exposures and trade-offs, not a single comparator.

In a sense, SAA is like a prize fight pitting the chosen portfolio design against a single competitor, the benchmark. Management can study this lone competitor at length, hugging or replicating the benchmark where they possess no edge and separating and fighting where they sense advantage. In this contest, the delivery of institutional aims is subordinated to a singular focus on benchmark outperformance. TPA places the chosen portfolio design in combat with all comers, aiming to outperform all feasible alternatives in the service of institutional aims. A more challenging prospect, for sure, but all possibilities are equally relevant, and realized performance should be assessed against each of these foregone alternatives.

Figure 5 illustrates CPP Investments’ approach to understanding the performance of portfolio design. At the outset of any measurement period, we identify a vast range of portfolios that share the targeted level of market risk and conform to institutional constraints. These alternatives differ by factor exposure, public and private composition, diversification, leverage, liquidity, currency, geography and sector. Each represents a viable design choice and will have delivered a distinct realized return over the period. We consider many thousands of foregone portfolio alternatives and calculate the realized returns of each over the observation period, arraying these returns in the distribution shown in Figure 5. The realized returns of our actual chosen portfolio design are also included.

While TPA aims to find the single best portfolio from among the multitude, there will always be some portfolios that outperform our choice in any fixed observation period. In fact, the “winning” portfolio design in any short time frame is typically suited only for the environment that transpired, rather than the range of possible environments that might unfold over time. The best long-horizon portfolio design is one that maintains a position in the upper quintiles of this distribution through many market cycles and under different market conditions.

Over the past few years, the upper deciles of this realized return distribution are dominated by portfolios with heavy exposure to U.S. tech mega-cap listed equities. Given the extreme dominance of U.S. public equities’ returns, particularly in a few industries, diversified portfolio designs delivered returns in the lower portions of the distribution, with private-asset heavy portfolios tending to be further to the left. This does not necessarily imply that diversification failed. Rather, it demonstrates that portfolios designed for many environments will underperform concentrated equity-centric portfolios in environments that deliver listed equity returns far in excess of expectations.

This technique rigorously assesses the portfolio design choice without reverting to comparisons with a single benchmark. The positioning of realized returns within the distribution of all viable alternatives provides critical evidence on the success of the TPA exposure targeting choice in supporting institutional aims.

Decision 3: Investment Selection: Separating Skill-Based Value Added from Market-Driven Returns

Investment selection is the third TPA decision evaluated at CPP Investments. It relies more heavily on benchmark comparisons. CPP Investments compares the realized investment performance of each investment strategy to a benchmark comprised of public market indices with similar systematic risk exposures. Results are then aggregated to the department level, helping distinguish skill-based value added from the beta-driven returns generated by broader portfolio design decisions.

The aggregation of these benchmarks provides an overall “Benchmark Portfolio” for the base and alternative CPP, providing a view of the collective impact of investment decisions. While management remains accountable for total portfolio outcomes; these benchmark comparisons help assess the contributions of individual strategies and departments. For private assets, it is critical to distinguish between two separate sources of value added: the strategic decision to invest through private rather than public markets, and the investment skill demonstrated relative to an appropriate private-market benchmark.⁷

Toward a Multidimensional Assessment of TPA Performance

Under a traditional SAA framework, the benchmark often acts as the objective, the instruction set, and the scorecard. Under TPA, management is accountable for a broader set of decisions that require multiple lenses. Performance assessment must go beyond absolute and relative returns and consider risk-setting, diversification, resilience, risk and capital allocation, portfolio construction, implementation skill and decision quality (Figure 6). An assessment must also consider whether governance processes are functioning effectively and whether the portfolio remains aligned with long-term objectives.

CPP Investments assembles the pieces of performance assessment that provide a more complete picture of whether and how TPA is improving long-term outcomes. Different metrics answer different questions. Reference portfolio comparisons help assess broad design choices, global market portfolios provide context on opportunity cost, unlevered portfolios isolate the impact of leverage, and peer comparisons offer an external perspective.

Benchmark outperformance is an important part of this balanced view. At the investment strategy level, benchmarks help distinguish alpha from beta returns and assess implementation skill, and at the total portfolio level they help explain aggregate implementation outcomes. Under TPA, benchmarks are a diagnostic input into performance evaluation rather than the sole measure of success. This distinction matters because benchmark outperformance does not always deliver institutional aims. A single strategy can outperform its benchmark by increasing concentration or adding exposures already held elsewhere in the portfolio. Conversely, the total portfolio may intentionally accept benchmark-relative headwinds if doing so improves diversification, resilience or alignment with long-term goals.

Conclusion

Performance assessment is difficult under a Total Portfolio Approach—not because performance is less measurable, but because the scope of accountability is broader.

A credible ex post assessment framework should be linked to institutional aims and employ multiple lenses, including risk targeting, diversification, portfolio construction, investment selection and long-term sustainability outcomes.  Benchmarks remain essential tools for attribution, discipline and accountability, but under TPA they become diagnostic inputs rather than singular verdicts on success or failure.

This challenge extends beyond pension investing. As organizations increasingly manage portfolios against multiple long-horizon objectives—including environmental sustainability, development, inflation protection, or social objectives—performance assessment cannot be meaningfully reduced to a single benchmark-relative outcome.

For boards and portfolio owners, the key takeaways are to:

• align evaluation with the delegated decision rights;
• connect outcomes to objectives; and
• use benchmarks as part of an overall assessment rather than as the sole measure of success.

The question is no longer simply whether a portfolio outperformed a benchmark, but whether all management decisions, including portfolio design and execution, improved the delivery of long-term institutional goals.

This is another excellent report on CPP Investments' total portfolio approach (TPA) that will surely stir conversations at the Maple 8 and beyond.

Make sure you read their first report, Investing in Uncertain Times: Achieving Disciplined Flexibility in the Total Portfolio Approach, which is available here. I covered it here.

Sally Shen, Derek Walker, and Geoffrey Rubin are very smart, no doubt about it, and they really do a great job distinguishing between the strategic asset allocation (SAA) framework and the total portfolio approach (TPA) and how the latter broadens diversification in a more meaningful way, and better aligns outcomes with objectives.

The big problem with TPA -- and they allude to this --  is that it's much harder to measure its success relative to the more straightforward benchmark comparison in SAA, which is used to evaluate active management decisions: 

Benchmark comparisons remain essential to evaluating active management decisions and maintaining accountability. But under TPA they are no longer sufficient on their own. Measuring the health of a total portfolio requires a broader, multi-faceted framework capable of assessing not only returns, but also the total portfolio’s resilience to adverse market and economic circumstances, the impacts of diversification, concentration and tactical decisions, the effectiveness of implementation, and alignment with long-term objectives. This report examines how CPP Investments is developing its approach.

Now, let's stop to ponder this for a second.

Critics will claim this is all rubbish. If CPP Investments cannot beat a low-cost index portfolio over a one, three, or five-year period, then why are we paying their employees big bucks to actively manage our pension assets?

For them, these discussions on TPA are nothing more than a theoretical exercise in obfuscation: "Benchmarks don't work any longer because concentration risk in equities is too high; trust our TPA approach is much more sophisticated and resilient than the traditional SAA approach."

For critics, it's simple: if you can't beat a low-cost ETF over a 5-year period, then there is no point in actively managing pension assets, even if you're delivering decent returns above the actuarial target.

And to be sure, the State Street SPDR S&P 500 ETF Trust (SPY) is up 73% over the last 5 years, led by the VanEck Semiconductor ETF (SMH), which is up 373% over that period.

In other words, five years ago, we could have closed down CPP Investments and all the other Maple 8 funds and shoved all the pension assets into the SPY, generating eye-watering returns, and saving hundreds of millions in fees to external managers and hefty compensation to the employees of these large Canadian pension funds (especially their senior managers).

Even if you put 85% or 70% of the assets in SPY or MSCI ACWI and the rest in a Canadian, US or a global aggregate bond ETF, you'd come out way ahead in terms of returns.

This is all true, but it's missing an important consideration: the risk of having such a high allocation to US or global equities when the concentration risk remains at a very high level.

Still, critics will snap back: "So what, concentration risk is fine as long as the bubble keeps inflating, and nobody knows when it will pop. It could last years"

Here is where we need to pause and ask: "Will US equities over the next 5 years produce anything close to what they produced in the last 5 years?"

And I say US because US equities dominate global equity indexes.

My answer is that it's highly unlikely that US equities come anywhere close to producing the same return over the next 5 years, and I also fear bonds will severely underperform if stagflation develops.

So, if public equities will not provide the requisite return, and neither do bonds, pension funds are left with alternative assets like infrastructure, real estate, private equity and private debt. 

This is where a total portfolio approach (TPA) differs from strategic asset allocation (SAA), as it directly links an outcome (attaining a return above the actuarial rate-of-return over the long run) to the objective of the pension fund, ie. maximizing returns without undue risk of loss, with regard to the funded status of the plan.

TPA also does a better job linking pension assets with their long-dated liabilities in a more comprehensive way.

CPP Investments is managing over $800 billion in assets, you definitely don't want them to be all in equities and they are large enough to take on illiquidity risk in assets that offer more stable, inflation-protected returns.

Moreover, the diversity of the portfolio, whether it's geographic, sector or by asset class and strategy, makes the Fund more resilient to downside risks over the long run.

That is the point CEO John Graham was driving at when he recently penned a comment on why protecting the CPP means taking the long view on investment returns.

But critics will keep harping that the bloated CPP Board is trounced by its benchmarks again. 

If these highly-paid, smart folks at CPP Investments can't beat their own benchmarks, why are we paying them for?

In fact, the same argument can be expanded to all the Maple 8 funds since they all underperformed their benchmark over the last three years as concentration risk in US equities has risen to historic highs.

Part of the answer is you're paying them to deliver consistent returns over their actuarial target (for most it's around 6%) over the long run, and you absolutely want them to minimize their drawdown to avoid another 2008 fiasco, where most pension funds were down 20% to 40%. 

Critics will still come at you: "Who cares if stocks decline 40%, they always bounce back big." 

Well, not always. Sometimes stocks keep going down for years, and it can take decades to come back to a previous peak. 

This is where the real value of adopting a total portfolio approach is critical for pension funds, because when the music stops and things start heading south, you want your pension fund to sustain that hit as best as possible (they will lose money but nowhere near as much as a pure equity fund). 

Moreover, sophisticated pension funds like CPP Investments manage their liquidity well, so they will be able to pounce on opportunities when market dislocations occur. 

"Ok, Leo, that's all fine and dandy, but how are we supposed to measure the success of TPA properly? It's not as straightforward as SAA."

Correct, it isn't easy, and nor am I suggesting it's easy to measure the success of TPA, but the easiest way is to link it to outcomes. 

Are you getting the outcome you're desiring over the long run with a diverse TPA approach? 

If so, then that really is all that counts.

Is TPA perfect? Hell no! There are still a lot of issues to contend and lots of uncertainty. 

For example, what if there is a structural change in private equity?  How does TPA account for this structural risk properly? The same applies to other illiquid asset classes and private debt.

How does TPA account for all risks? For example, CPP Investments is underweight tech shares in global stock markets but overweight data centres in illiquid asset classes.  

Is that a disconnect or an appropriate risk to take?

And benchmarks remain important, but they are far from perfect and should be comprehensively reviewed every three or five years, and changed like BCI did last fiscal year? (to introduce more sector-neutral indexes or equal-weight indexes)

Any changes to benchmarks must be clearly stated in the annual report with a full discussion and analysis. 

Equally important: What is the opportunity cost of adopting a TPA framework over an SAA framework?

In the last five years, it's been huge, but was this always the case? 

CPP Investments and other large pension funds espousing a TPA approach need to provide numbers and hard facts to back up their case for TPA, not just conjecture and "trust us, we know best."

If not, the critics will always lurk in the background, claiming they are bloated pension funds that keep underperforming their benchmarks while they dole out huge compensation to senior managers.

This is an important discussion and I am only scratching the surface. 

I'd need to interview Sally Shen, Derek Walker, and Geoffrey Rubin to really get into it properly with them but they have done a fantastic job highlighting the important issues and why TPA is a better framework than SAA for pension funds.

Below, CPP Investments' CEO John Graham joins Nicolai Tangen, CEO of Norges Bank Investment Management, for a great discussion in his In Good Company podcast. 

Take the time to listen to this podcast, John goes over many important points on their total portfolio approach, and Nicolai does another masterful job of asking all the right questions.