Total Fund Management Part 7: Envisioning the Canadian Model 2.0

This is the last part of a seven part series on integrated Total Fund Management brought to you by Mihail Garchev, the former VP and Head of Total Fund Management at BCI and I. Please take the time to read Mihail's synopsis on envisioning the Canadian model 2.0 below followed by my brief comments and a clip where he delves deeply into today's topic (added emphasis is mine):

This is now our last week of the Introduction to Integrated Total Fund Management ("TFM") executive series. At the beginning of the series, we discussed some of the organizational and investment beliefs behind the increased interest in TFM and the corresponding organizational and investment strategies and objectives to support these beliefs.

With Episode 6 last week, we concluded the TFM implementation topic, which included a closer look at the TFM framework, process, and what is required to turn these into a capability. In the process, we also looked at 11 engaging case studies to illustrate examples of the use of TFM for real-life decision-making.

Typically, Episode 6 would have exhausted the topic of TFM as it covered all the functional aspects of TFM and how these functional aspects lead to the more efficient and effective management of the client portfolios.

However, throughout the series, we have been highlighting that TFM also has a structural role in enabling the next stage of development of the Canadian pension model. This is the focus of today's Episode 7.

More specifically, in Episode 7, we will discuss the evolution and critical success factors for the Canadian model. We will look past the usual discussion around the model, which looks at the macro factors, such as governance, talent and plan design, and we will try to look "under the hood" – understanding some misconceptions about the model, key characteristics and impacts, and the decisions that led to its success. And then we will ask the question, "Could there be a Minsky moment for it where stability breeds instability?"

While TFM could address many challenges to the current Canadian model, a critical challenge emerges, which relates to the embedded principal-agent conflict in the organizational design. This principal-agent conflict has been accelerating and manifests itself in the culture of the organizations. It is not surprising that culture has been a critical focus for all of the Canadian pension funds. And culture is directly and inversely related to size, scale, and complexity.

Unfortunately, the organization's culture is deeply embedded in the system and is therefore extremely difficult to change. An organization's culture comprises an interlocking set of goals, roles, processes, values, communications practices, attitudes and assumptions.

But if culture is so difficult to change and so embedded in the system, would it be possible to change the system and structure instead?

This discussion leads us to discuss organizational structure and design, specifically in the context of the Canadian pension model. What emerges from this discussion are four potential avenues to further advance the Canadian model to what we called its version 2.0.

In the end, the Canadian model version 2.0 endeavors to bring back the entrepreneurship and competitiveness, efficiency and effectiveness, innovation and creativity, and create a positive feedback loop back to the culture. So, this is at the. But in the beginning, it was Episode 1 and the in-depth discussion on the challenges the Canadian pension funds are currently facing.

In the beginning – A Flashback to Episode 1

We first discussed that the Canadian funds might be at a critical point of the organizational maturity because of size and scale, complexity, among other issues. The efficiency of the Canadian funds, the so touted "economies of scale," and by extension, the Canadian model as it is today, might be facing their biggest challenges.

From this perspective, TFM's goal is to restore the economies of scale by pursuing second-order efficiencies. We called these "economies of scope." These efficiencies occur because one uses an optimal top-down view to evaluate and adjust what may seem optimal on a standalone basis for the asset classes or operations. Some examples of such second-order efficiencies include reducing overlaps and associated costs and removing unintended bets and consequences, which is a hidden cost. Also, ensuring an optimal flow of capital and liquidity and that asset classes and Total Fund both function optimally and Total Fund is not subsidizing, in a way, the asset classes. Finally, help evolve other functions, such as risk, performance, reporting, and communication.

Client maturity and external environment are additional reasons for an increased Total Fund focus. In many cases, the pension plans' maturity leads to an increased sensitivity to negative outcomes due to the interaction of net cash outflows and market downturns. As such, it becomes increasingly important to manage the short and medium-term returns better and avoid adverse outcomes in a lower expected return environment going forward. Being efficient matters even more because cost becomes an increasing part of the overall return.

We have also witnessed an extraordinary environment of significant structural changes, disruption, pandemic, ESG, and various policies and regulations. Thus, selectivity would be an essential part of portfolio management. Just buying beta might not be enough anymore.

Finally, it might also be a question of survival for the pension model itself. With increased competition and the commoditization of strategies and low cost, sometimes even zero cost for some strategies, it becomes increasingly important for the pension fund managers to remain relevant. As much as clients might be captive, they could also influence organizational restructuring and investment strategy via the board of directors and even opt-out in some instances.

Another aspect is the ability of TFM is the need to manage complexity: the efficient flow of capital, managing beta, avoiding adverse outcomes, ensuring liquidity, and efficient implementation Our final point in Episode 1 was that ultimately, what matters are outcomes. For a typical pension plan, key outcomes are wealth to pay benefits, contribution rate stability, inflation adjustment, and liquidity. Outcomes could be achieved by designing a preferred mix based on asset classes and their long-term assumptions and hoping that these assumptions will realize. Or one should not be concerned about the asset mix as an essential belief, but rather, establish the required outcomes and tolerance levels around them and have this as an objective and as a policy portfolio. Then, continuously evaluate the best short and medium-term returns, their probability and confidence levels, and build a portfolio based on this. At any time, whatever the current asset mix is, it will have the highest probability of keeping outcomes intact.

Given the discussion above, we can outline the following key challenges:

  • Size, scale and efficiency challenge the efficiency of the Canadian model 
  • Client maturity challenge of increased sensitivity to adverse outcomes 
  • External environment challenge of lower returns (less margin of error), disruption, headwinds and tailwinds of themes, and external competition (relevance of the Canadian model) 
  • Portfolio effectiveness challenge due to the need to manage complexity: the efficient flow of capital, managing beta, avoiding adverse outcomes, ensuring liquidity, and efficient implementation 
  • Increasing focus on outcomes instead of reliance on stable asset relationships
  • Throughout the series, we have demonstrated that adopting TFM could address many of these challenges. We called this the functional role of TFM.

But could it be that in the ever-increasing size and complexity, TFM could only do so much? Beyond a certain point, these challenges may manifest themselves again and even reinforce their negative impact? For example, CPP Investments alone is projected to grow from $400 billion to $1 trillion, and similar growth for some of the other funds. Could size & complexity still overwhelm?

In most of the discussions so far, we highlighted the role of TFM to impact specific aspects of the Canadian model, mostly related to effective investment decision-making and resulting efficiencies (functional TFM). But is it possible that TFM could impact the entire organization, not just some parts of it (i.e., the structural role of TFM? And this includes not only investment management, but operations, structure, and importantly, culture.

Understanding the Canadian Model 1.0

There are many excellent papers on the Canadian model published over the last three years.

More recently, it is the excellent paper by Clive Lipshitz and Ingo Wolter titled "Public Pension Reform and the 14th Parallel: Lessons from Canada for the U.S.", as well as the World Bank paper titled "The Evolution of the Canadian Pension Model: Practical Lessons for Building World-Class Pension Organizations," as well as the numerous and again important papers by Keith Ambachtsheer and CEM Benchmarking.

My goal is not to repeat all the conclusions in these papers but look under the hood of the general conclusions. I will summarize the general conclusions nonetheless for completeness sake below:

  • Joint sponsorship, unified legislation, arm's length, independent governance 
  • Risk sharing and mandated funding of accumulated deficits 
  • Size and scale, the certainty of assets, long horizon, in-house management, broad and global diversification, a significant focus on private assets, less regulatory constraints 
  • Top talent, competitive compensation, professional management & boards 
  • Ability to pool assets, strategic partnerships and knowledge capital 
  • Conservative determination of pension funding formulas (discount rate). Contributions and benefit payments are evaluated holistically so that adjustment of one results in automatic adjustment of the other

These success factors were undoubtedly critical to success but did not tell the whole story. What is also important is to look "under the hood" and understand what happened at the organizational level, not the macro level. This exploration's insights could then point to what we could address in version 2.0 of the Canadian model. My insights are a combination of in-house specific knowledge as well as access to detailed peer data by CEM Benchmarking, in addition to several publications by CEM on the topic (see, for example, "The Canadian Pension Fund Model: A Quantitative Portrait," 2020). First, there is this common misconception that the Canadian model is more efficient. While it is true that size, up to a point, leads to economies of scale, the Canadian funds were only marginally more efficient than non-Canadian funds. The Canadian funds were, however, more effective. Efficiency is about doing things right — i.e., completing a task cheaper or faster. Effectiveness is about doing the right things — things that yield positive results. So, what were the right things that the Canadian funds do?

You can watch the presentation for more details, but the most important thing was to internalize the asset management in a nutshell. All the cost savings were then reinvested in the capability to manage real assets, build the foundation for active management, and support paying for talent and building risk management and systems (both crucial parts of the capability). Thus, costs were not necessarily reduced, but rather, reinvested in the capability. To gain a perspective, Canadian funds invested 4-5 times more in risk management and systems than non-Canadian funds.

But just having the capability is not enough; one needs to put this capability to work. It was critical to quickly deploy capital in the real assets in conjunction with more active management. This requires the ability to build relationships, networks, and "foot on the ground" via partnerships and platform investments and be nimble and have fast and best-in-class execution (legal, tax, etc.). This is where reinvesting in the capability was essential.

Other important decisions were to reduce the home bias (relocate to global and EM markets) and the efficient use of leverage (3-4 times more leverage than non-Canadian funds). Naturally, the above considerations led to an asset mix with fewer equities, more real assets, and fewer bonds (but leverage). Also, substantial changes were made within asset classes (e.g., equities due to the home-bias reduction and Real Estate and Infrastructure – see the presentation for more details).

When one looks at all these decisions together, the stars aligned for the Canadian funds, as all these decisions proved to be the right decisions. Contrary to the common perception, size and active management was not a significant driver of the risk-adjusted performance. What was crucial was the internalization, the reinvestment in capability, and deploying in real assets. 

But could it be that this stability may breed instability, this classical (now) Minsky moment, as illustrated in the figure below?

Translated to the Canadian pension fund context, you can think of this as the combination of accelerating organizational challenges (as discussed earlier) and the potential of market impacts, both structural and market shock-based. The more complex a system becomes, the more fragile it becomes. Part of this fragility is hidden and latent relationships that are not entirely understood (you need TFM for that!) and human behavior, which is as complex (and egocentric, no matter the parade of LinkedIn value statements) as it gets.

An excellent example of such a Minsky moment was the ABCP debacle for some of the Canadian pension funds during the financial crisis. The crisis was bad itself, but it had a severe impact because it was reinforced by human behavior.

This necessitates digging deeper into the impact of culture on the organizations, beyond just the façade of well-intended and well-crafted value statements and LinkedIn posts. After all, contrasting LinkedIn and Glassdoor is an excellent first glimpse at the difference between perception and reality. We explore the various facets of culture and its impact in detail in the presentation. We even go up to a point to talk about "ant wars." What is this? – more of it in the presentation.

Culture is too general of a term. What is important is how one decomposes the culture into underlying components. Significant in this respect is the principal-agent conflict. We discuss and illustrate this conflict in detail in the presentation. Still, the gist of it is related to the conflict between the agent's egocentric self-interest versus what is right to the principal (beneficiary).

While TFM could address many challenges to the current Canadian model, a critical challenge emerges, which relates to the embedded principal-agent conflict in the organizational design. This principal-agent conflict has been accelerating and manifests itself in the culture of the organizations. It is not surprising that culture has been a critical focus for all of the Canadian pension funds. And culture is directly and inversely related to size, scale, and complexity.

Unfortunately, the organization's culture is deeply embedded in the system and is therefore extremely difficult to change. An organization's culture comprises an interlocking set of goals, roles, processes, values, communications practices, attitudes and assumptions.

But if culture is so difficult to change and so embedded in the system, would it be possible to change the system and structure instead?

This discussion leads us to discuss organizational structure and design specifically in the context of the Canadian pension model and inspired by some early work by Deloitte on the "Adaptable Organization."

The figure below illustrates how we think organizations work versus how they work.

If that is the case, then it is essential to harness the power of cross-functionality and embed it in the pension organization design. It becomes even more apparent when we bring back one of the tenets of TFM about managing to outcomes, not necessarily asset classes. By extension, outcomes are the product of cross-asset, multi-strategy investment management and decision-making, which the core capability of TFM.

This means that the cross-functional organizational structure is the natural extension of TFM. This is how TFM now starts playing a functional role. In the presentation, we discuss many aspects and additional dynamics and structures related to this notion.

However, what is essential is that combining TFM and such a cross-functional organizational design could allow us to address many of the principal-agent conflicts as part of the overall cultural context. If organizations could become more entrepreneurial and more innovative and creative in a more competitive setting, they would become more efficient and effective. The structure itself would force out the negative sides of culture and the principal-agent conflict.

The final part of the presentation brings together all the insights on TFM, culture, and organizational development in a set of evolved Canadian pension models (the various version of the proverbial 2.0). We discuss these models in detail in the presentation, so next, we will provide just a brief illustration with the salient points.

Model 1 - The Horizontal Scorecard

Most organizations have some variation of "vertical" scorecards by departments and functions. But the notion of a cross-functional team and outcome-oriented portfolio management allows us to create a different organizational structure. For example, one could define strategic and pension sustainability goals as "outcomes" and then assign these "outcomes" to multi-disciplinary teams. The success of these teams relies on successfully delivering the outcomes. The CIO, via the TFM function (framework, process, capability), manage the "outcomes" toward the ultimate purpose. This links together the notion of outcomes, outcome-oriented portfolio management and the role of the TFM function.

 Model 2 - The Holding Company

Another possibility is to carve out functions (e.g., separate Infrastructure, RE, Public Markets, etc.). The pension fund itself becomes a holding company comprised of TPM and some support and administrative functions.

This is taking a step further what already exists in some funds where there are separate real estate and infrastructure companies (QuadReal, Ivanhoe- Cambridge, Cadillac-Fairview, Borealis, etc.). The CIO manages all required outcomes based on the "building blocks" from the Asset Co's. Asset Co's compete for capital ("internal market") but also External Co's. Operations are brought back to proximity within the Asset Co's, but there could be some shared services (e.g., custodian). Such a structure may bring less scrutiny/regulations, at least in the beginning. Innovation needs cooperation, not competition. Therefore, it should be a separate activity to support TFM, create new Asse Co's, or support Asset Co's needs (proximity). Finally, what is essential is also the ability to monetize the business model. The Asset Co's could manage 3rd party money (external Client Co's, which could be smaller or international pension funds, or any other capital). As such, fees flow back as income to pension beneficiaries.

Model 3 - The Baby Funds

The "Baby Funds" are carbon copies of the original fund. They are self-organizing and could choose any organizational model they prefer. Baby Funds compete for capital ("internal market") where the CIO allocates capital to the Baby Funds in a similar way one would allocate capital to external managers. The CIO still manages the outcomes using the TFM function. The Baby Funds could attract 3rd party money (e.g., issue MTN linked to performance). As such, we again could have the ability to monetize the business model. Asset Co's could manage 3rd party money (external clients could be smaller or international pension funds), and fees flow back as income to pension beneficiaries.

Model 4 - The Outcome Funds

Finally, the "Outcome Funds" are still self-regulating companies. Capital is captive and allocated as per the CIO and TFM function. There is no competition for capital, but there could be competing external "Outcome" managers. This structure brings clarity of mandate, explicit objectives, priorities, the measure of success. Efficiency captured explicitly via "cost per outcome." No "hiding" and "diluting" the results into vague and unclear statements and maybe convoluted measurement and reporting/communication. There is also the proximity of operations to investment. Because of the "outcome" separation, inefficiencies of the size and scale are addressed.

Lastly, it is useful to compare these variations of the Canadian Model 2.0 against some of the organization's desired effects, as depicted in the figure below.

These are just some initial thoughts to initiate a discussion, but TFM plays a central structural role in all the potential models.

Overall Episode 7 Takeaways:

  • The Canadian model, as we know, it is only marginally efficient. Its strength is in the effectiveness – the ability to align many right decisions to achieve the right outcomes
  • But the increasing challenges might put the test this past effectiveness and challenge the efficiency 
  • The functional role of TFM (framework, process, capability) can address many of the challenges related to size, scale and efficiency, client maturity & external environment challenge, portfolio effectiveness and increasing focus on outcomes 
  • But many of the challenges are related to structure and culture. They ultimately result in less efficiency (cost) and effectiveness (right outcomes). Culture might be challenging to change and might require rethinking the structure 
  • Focusing on entrepreneurship & competitiveness and innovation & creativity ultimately leads to restoring the efficiency and effectiveness 
  • The Canadian model's current strengths and benefits could be further extended by adopting structures that could significantly increase its future potential. TFM plays a critical role in this

This concludes our series. Thank you all for following Leo's blog over the last ten weeks and for all your support, comments, and insights!


Let me begin by thanking Mihail for another great comment, I personally think he saved the best for last.

Against my advice, he literally stayed up all night to complete everything, which is why he might sound a little tired in Episode 7 below, but it's still well worth watching it.

Since this is the final post on this series, if you have not done so yet, please review all of the previous parts of this series:

Alright, let me keep my comments brief on envisioning the Canadian model 2.0.

Earlier this week, in a post featuring another guest comment from the team at the McGill International Portfolio Challenge (MPIC) on impact investing, I stated this:

[Sebastien Betermier] has written a great paper on the Canadian pension fund model with Alexander Beath, Chris Flynn and Quentin Spehner of CEM Benchmarking,  

Together, the authors of this paper understand why Canadian pension funds perform better than most other pensions and registered savings plans but they too wonder:

The years ahead will put the Canadian model to the test, as the severe impact of COVID-19 on commercial real estate, equities, and corporate bonds will undeniably hurt the funds’ assets in the short-run,” said the paper. “How resilient is the Canadian model to a pandemic? Will two-pronged strategies that increase asset performance and hedge against liability risks change in the post-pandemic world? We leave these questions for future research.”

Part of the answer to this question can be found in the Total Fund Management series Mihail Garchev and I have been discussing over the last few weeks.

In order for Canada's large pensions to thrive in a record low rate environment/ post-pandemic world, they really need to adopt a TFM approach to unleash the second order effects Mihail talks about above.

Importantly, and I believe this with every fiber of my being, the Canadian Model 1.0 is stale and obsolete, stick a fork in it, it's done!

What did you say Leo? Mr. Pension Pulse, have you lost your marbles?

Nope, I know exactly what I'm talking about. Yes, Canadian model 1.0 is a great success story but it can only take Canada's large pensions so far, and when everyone is doing the same thing, and complacency sets in, that's when the "Minsky moment" can creep in as stability breeds instability.

Mihail gave the example of ABCP. Funny enough, my first blog post years ago was on the ABCPs of pension governance, discussing among other things, bogus benchmarks.

I can sum up the ABCP scandal/ crisis at the Caisse in one sentence: too much power concentrated in the hands of too few doing extremely risky and dumb things using pensioners' money.

And it wasn't just ABCP. There was another senior portfolio manager at the Caisse back then (he's still working in the industry so I'll be kind and not mention his name), who was using the Caisse's balance sheet to do exotic long-term volatility swap trades. And he had full backing from the President who thought he was a "genius".

Genius, my ass! I arranged for a lunch with this person and a couple of former portfolio managers here in Montreal who allocate funds to sophisticated hedge funds. He was trying to raise money from them and when they heard his pitch and what he does, one turned to him and flatly stated: "Stick with the Caisse, nobody in their right mind would ever give you an allocation to do such a risky strategy."

Well those two Jewish allocators were smart as hell and they knew exactly what they were talking about because a few years later, this person's "sophisticated strategy" cost the Caisse and its depositors $1.2 billion, if not more.

Amazingly, and quite worryingly, nobody at the Caisse was allowed to question this person or his strategy and he hid everything from everyone, going as far as telling his back and middle office analysts bits and pieces so nobody understood what the hell he was doing.

It took a very experienced and senior portfolio manager and another senior portfolio analyst to deconstruct the "genius's" portfolio to figure out what a mess the Caisse was in.

Over at PSP, right after I warned them of the looming US housing market collapse and the insanity of CDO-squared and CDO-cubed issuance back in the fall of 2006, it didn't take long for another credit portfolio to blow up. That portfolio manager and his team were not evil geniuses, but they had no clue of the risks they were taking and it could have literally annihilated PSP in 2008 (it was that bad!).

Culture. Sometimes you have arrogant jerks with too much power, sometimes you have nice guys or gals doing silly things, taking excessive risks, trying to look "sophisticated" but in reality they expose a pension to huge losses.

Mihail was a bit too kind and politically correct in his comment above.

Let me be blunt: without the right people and culture at all levels, especially senior levels, the best TFM approach is totally useless. All it takes is one or two powerful jerks (or nice PMs with too much power) to destroy a pension.

If you're a CEO of a major Canadian pension and you have one or more bad senior managers, get rid of them, the cost of keeping these people in senior roles far outweighs the settlement packages to oust them. 

Toxic culture is the death knell of innovation and collaboration and it renders TFM obsolete.

I'll leave it at that because I can go on and on but you get the picture. 

Let me end once again by thanking Mihail Garchev for a truly phenomenal series, it was a real pleasure working with him again and I hope he comes back with other material, like benchmarks and compensation at Canadian pensions (he implemented great stuff at BCI). 

Below, Episode 7 of the seven-episode series "Introduction to Integrated Total Fund Management" presented to you by Mihail Garchev, former VP and Head of Total Fund Management of BCI.

Great stuff, veteran, intermediate and novice pension managers can learn a lot from this series and we all owe a debt of gratitude to Mihail for delivering it to us.