More on Canada's Dirty Pension Secret?

Last week, I wrote a comment on the dirty secret behind Canada's pensions going over a report put out by the Fraser Institute, Risk and Reward in Public Sector Pension Plans: A Taxpayer’s Perspective.

The report was authored by Philip Cross, the former chief economist of Statistics Canada, and Malcolm Hamilton, a Senior Fellow with the CD Howe Institute and a retired pension actuary who spent 33 years at Mercer where he advised large pension plans in both the public and private sectors.

That comment received a lot of attention and I didn't hold back on my skepticism on the $22 billion risk transfer the authors claim is a subsidization of Canada's public-sector plans, nor on my opinions that these plans are 'unambiguously good' for the economy and should be bolstered and expanded so everyone has access to them, not just public-sector employees.

Before proceeding with the follow-up comment, it's worth going over (again) the main findings of the report:
The most striking feature of Canada’s retirement system is arguably the large and growing gap between pensions in the public and private sectors. Eighty percent of public sector workers participate in defined benefit (DB) pension plans. Only ten percent of private sector workers can make the same claim.

With the collapse of interest rates in the early 2000s, DB plans became prohibitively expensive in the private sector, yet they flourished in the public sector. If private sector employers can no longer afford even modest DB plans, how can public sector employers afford much more expensive plans—plans with larger pensions, earlier retirement, and full inflation protection?

Canada’s public sector DB plans frequently attribute their success to the “Canadian Pension Model.” A recent World Bank study identifies superior governance, economies of scale, innovative investment practice, responsible funding, and visionary leadership as important features of this model.

Without disputing the virtues of the Canadian Pension Model, we attribute the success of Canada’s public sector DB plans to large public subsidies made possible by practices that are neither admirable nor virtuous: bad accounting, poor governance, imprudent risk taking, and inadequate financial disclosure. Responsibility for these failings lies not with the pension boards who administer the plans but with the employers who sponsor them. These employers, usually governments, fail to represent the public interest when it conflicts with the interests of their employees.

The narrative advanced by Canada’s public sector DB plans raises a perplexing question. If innovative investment strategies abetted by good governance explain their success, why don’t private sector employers adopt the Canadian Pension Model and provide comparable pensions to their employees? Our answer is that Canada’s public sector DB plans do things that private sector DB plans are prevented from doing for good reason. In particular, public sector accounting standards allow public sector employers to materially misrepresent the cost of their pension plans. Private sector employers are prevented by private sector accounting standards from doing the same thing.

Taking investment risk is a legitimate tactic provided that those who bear the risk also reap the reward. This is not what happens in Canada’s public sector DB plans. Consider the plans covering employees of the federal government. Plan members, whose interests are ably represented by powerful public sector unions, are handsomely rewarded for investment risk taken by their pension plans and borne by the public. The public, whose interests are poorly represented by the federal government, receives no reward for bearing this risk.

Public sector DB plans cite their independence from government as a key to their success. We argue that this independence is a flaw, not a virtue, of public sector pension governance. The plans take investment risk to advance the interests of plan members while the interests of taxpayers, who ultimately bear this risk, are ignored. These practices are best described as moral hazard, not good governance.

This paper questions whether Canada’s public sector pension plans have discovered a formula that makes them a model for the world to emulate. The exceptional feature of Canada’s public sector DB plans is not “world-beating” investment strategies or good governance. It is the ability to enrich public employees by shifting large, undisclosed investment risks to taxpayers without fair compensation. By our estimate, this provides an unacknowledged $22 billion annual subsidy to Canada’s public sector DB plans and, ultimately, to the members of these plans. This large public subsidy, not the virtues of the Canadian Pension Model, explains the plans’ success. Without it, public sector DB plans would be no more viable than private sector DB plans.
You can read the full report here and the executive summary here.

Now, I solicited a few experts to talk about this report and my comments.

Given that Malcolm Hamilton is one of the authors of the report, I think it's only fair to begin with his feedback on the comment I wrote last week (added emphasis below is mine):
First, let me be clear about my objectivity. I am not employed by the Fraser Institute. I am neither a member nor a donor of the Fraser Institute. I was not paid to write the paper. I have not earned any employment or business income since retiring 6 years ago.

I have held most of the views expressed in the paper for years. Some were expressed in two papers published by the C. D. Howe Institute in April, 2014. Others were explored in presentations made in 2016 and 2017. I believe that I have fewer ulterior motives and financial entanglements than others expressing views about public sector pensions.

I do not consider myself an opponent of public sector pension plans. I advised three public sector pension plans for 10 to 20 years before my retirement. My eldest child is employed by a public sector pension plan. I admire and respect the employees, executives and board members with whom I worked over the years.

While not an opponent of public sector pension plans, I criticize public sector employers for using public sector pension plans to unjustly enrich their employees at public expense. I also criticize the public sector pension plans that abet this process. I do not believe that Canada’s public sector DB plans are nearing perfection. Finally, I believe that the Canadian Pension Model contributes to the success of Canada’s public sector DB plans without explaining this success.

Let me explain the $22 billion “subsidy”. We estimate that public sector pension plans expect to earn about $22 billion each year by taking investment risk borne by the public. There are two questions that must then be asked and answered.
  • Who deserves the $22 billion? We believe that those who bear the risk earn, and deserve, the $22 billion.
  • Who receives the $22 billion? The actuary, in setting contribution rates, anticipates the risk premium and uses it to reduce contributions. The $22 billion is therefore not available to reward those who bear the risk – it was distributed years ago as a reduction in contribution rates that benefited plan members, not the public.
This is explained more fully in the paper. The bottom line – the public earns $22 billion per annum by bearing investment risk taken by public sector DB plans but receives none of the $22 billion. This is the $22 billion subsidy to which the paper refers.

In your comments you dismiss the $22 billion and pivot to an entirely unrelated matter.
“What I do know without a doubt is the success of Canada's large DB plans is unambiguously good for the economy and government tax revenues over the long run.”
In other words, it doesn’t matter whether the public subsidizes public sector DB plans; the important thing is whether DB plans are good for the economy. With respect, making public employees pay for their DB plans will not make their DB plans less good for the economy.

You reference a 2013 study by the Boston Consulting Group (BCG) – a study commissioned by public sector pension plans to celebrate the good that public sector pension plans do for the economy - a study that would never have seen the light of day had it concluded otherwise.

The study, which was commissioned by HOOPP, OMERS, OPTrust and the OTPP, extended an earlier study that BCG conducted for the 10 largest public sector pension funds in Canada. The earlier study had concluded that Canada’s “top 10” pension funds were large, efficient, capable, successful investors – exhibiting all the qualities that world-class investors should possess, with the possible exception of humility.

The extended study was supposed to look more generally at defined benefit pension plans in Canada, not just at the top 10, and demonstrate that Canada’s DB pension plans were not just effective retirement savings vehicles, but important contributors to a healthy, growing Canadian economy. This, it turned out, was difficult to demonstrate – possibly because it was untrue. Undeterred, the commissioners found creative ways to interpret the data and spin the findings.

Why was this so difficult?

The earlier study found that, in 2011, Canada’s 10 largest public sector pension plans collected $71 billion in contributions and paid $74 billion in benefits. The plans were obviously important players, but what did this mean for the Canadian economy? Not much.
  • Contributions increase savings, but decrease spending and taxes.
  • Pensions decrease savings, but increase spending and taxes.
With $71 billion of contributions and $74 billion of pensions, the top 10 were responsible for $3 billion of net spending in a $2000 billion economy. Depending on whether you favor spending or saving, this can be seen as good or bad. It cannot, however, be seen as important.

However, it was important to the commissioners - so they decided to accentuate the positive. They discussed the positive impact that pensions have on taxes and spending. They discussed the positive impact that contributions have on savings. Little was said about the negative impact that pensions have on savings or about the negative impact that contributions have on spending and taxes. Nor was much said about contributions and pensions netting out to almost nothing. These clever manipulations supported the conclusion sought by the commissioners, but not without frightening BCG, the firm ostensibly standing behind the analysis and conclusions. So BCG asked for a strong disclaimer to accompany the presentation wherever it went.
“The materials excerpted by commissioners from the Study referenced are provided for discussion purposes only and may not be relied on as a stand‐alone document. Additional analysis has been done to the data and analysis contained within the Study by third parties other than BCG. BCG has not independently verified this additional analysis and assumes no responsibility or liability for it.”
Those asking for the Study referenced in the disclaimer were told that the commissioners had decided not to release it. Consequently, users were forced to rely on material known to be unreliable. I vividly remember the impression that the presentation made on the audience… just imagine how well our economy would perform if we doubled or tripled public sector pensions! Think of the potential!

The commissioners’ self-serving interpretation of the BCG study is a slender reed upon which to build confidence in the role that public sector pension plans play in our economy. Canada’s public sector DB plans do a great job for public employees. The pension funds are well managed by capable professionals. However, the plans are not huge contributors to the Canadian economy and they do a poor job for the public at large, as explained in our paper.

You question our assertion that inadequate financial disclosure plays a role in the success of public sector DB plans.
“Really? Have they read the annual reports of these large public-sector DB pensions in detail? There are sections on first-rate governance, accounting, disclosure up to wazoo on everything from investments to compensation, to you name it.”
Let me assure you that I have read the annual reports of the OTPP, OMERS, HOOPP, OPTrust, the OPB and the PSPIB for years. Having a self-congratulatory section on first-rate governance is not proof of good governance. Our paper concedes that the plans are well governed in many ways. However, the most important duty of boards is managing the pension fund and here we find problems.

Consider the PSPIB. It has a statutory duty to plan members. Plan members receive all of the reward for investment risk taken by the plan but bear none of the risk. The public bears all of the risk and receives none of the reward. The PSPIB controls investment policy and serves only plan members. Where, precisely, do you see any semblance of good governance here? This is textbook moral hazard. It is a recipe for reckless, irresponsible risk taking with predictable consequence.

As for disclosure, the amount of disclosure, measured in pages, is impressive but this does not compensate for repeated failures to disclose important things.

Consider disclosures about risk. I looked at the 2017 annual reports for the OTPP, OMERS, HOOPP, OPTrust and the OPB. Here is what I found.
  • 490 pages
  • 1088 occurrences of the word “risk”
  • 129 references to managing risk
  • 19 references to taking risk
  • No references to bearing risk
  • No references to the size of the risks that the plans were taking.
  • 29 occurrences of “prudent”
  • 31 occurrences of “secure”
  • 33 occurrences of “stable”
  • 84 occurrences of “sustainable”
  • No stress tests
  •  No summaries of the results of the asset/liability simulations that plans periodically perform to assess long term risk.
Readers learn that the plans take risk to earn the returns they need. These risks are managed by professionals. Apparently, no one needs to bear any risk, presumably because professionals manage it out of existence.

Then there are the traditional empty reassurances. The plans are sustainable – but sustainable isn’t defined. Pensions are secure – whatever that means. The plans are committed to stable contribution rates – but does that mean that they have reason to believe that contribution rates will be stable?

With secure benefits, stable contributions and sustainable pension plans, why would anyone worry about risk taking? Readers are encouraged to believe that risk can be taken without consequence. Risk is just short-term noise that pension plans can safely ignore. Risk is something with which we pose as we grow rich. No one needs to think about how big the risks are or who bears them in the long run because the long run never disappoints.

You may view this as adequate disclosure. I view it as borderline delusional. Here we have the selective disclosure of things that plans are comfortable disclosing combined with the deliberate non-disclosure of things that the plans would rather not disclose.

Finally, let me address your comments about the “real crime” going on in financial services. I agree with your criticism of the high cost of financial services. This is a longstanding and serious problem in Canada. It is not unlike the serious problem we have with public sector pensions. In each case, the party bearing the risk receives little or nothing in return for doing so. In public sector pension plans, the public bears most of the risk while plan members enjoy the rewards. In high fee mutual funds, investors bear most of the risk while providers/advisors enjoy the rewards. The difference between you and I? I criticize both practices while you criticize only the latter.
Let me first thank Malcolm for providing me with such a lengthy response to my comment. His first draft was even blunter and even though I wanted to publish it, he asked me to revise it for public consumption (no problem).

I like Malcolm, I think he's a very smart actuary with great experience, so when smart people talk or write to me, I first listen, soak it all up and then reflect.

I apologize for jumping to the conclusion that the Fraser Insititute paid him to write this report. Notice he speaks for himself, not Philip Cross, the other author of the report, but I will assume both didn't get a dime for writing this report and published it out of civic duty.

That still begs the question: Why did Malcolm and Philip write a report like this for the Fraser Institute which has over the years looked to undermine the success of Canada's large public-sector pensions? (both these authors aren't stupid, far from it, they both know the Fraser Institute's angle is to criticize anything public and promote a private sector solution in its place)

Also, while Malcolm states he's not an opponent of public-sector pension plans and lauds their investment success, he does criticize them for aiding and abetting the unjust enrichment of public-sector employees at the public's expense to the tune of $22 billion a year.

He also took this opportunity to criticize a 2013 study by the Boston Consulting Group (BCG) which I referenced in my comment.

Here, I will admit, the findings of the study were interpreted in a way to emphasize the benefits of Canada's public-sector pensions not just for members but for the overall economy. Maybe there was some exaggeration but not as much as Malcolm claims.

And let's not forget there are major reforms going on to enhance the Canada Pension Plan over the next few years for the next generation of Canadians, something which Malcolm conveniently omits to point out in his criticism of the economic benefits of Canada's large DB plans.

There is something else that bugs me in Malcolm's criticism. Let's say our large public-sector plans were in much worse shape. Let's say they were severely and chronically underfunded like many large US public pensions are, and there was no shared risk model, then maybe I'd be a lot more receptive to Malcom's criticism.

But that's not the case. Canada's large public-sector DB plans use a much lower discount rate than their US counterparts, they've adopted much better governance to separate governments from their investment activities and to hire highly qualified people to manage assets and liabilities, and most of them have adopted risk-sharing (typically in the form of conditional inflation protection) so if they run into a deficit, the contribution rate is increased, benefits are cut or both until the plans are fully solvent again.

Nowhere in the report are these facts discussed, its focus is squarely on the $22 billion a year subsidization these public-sector plans reportedly receive from Canadian taxpayers.

Malcolm criticizes the sensationalism of the BCG report, overexaggerating the benefits of Canada's large public DB plans, but we are supposed to take the figures he and Philip Cross state in their report at face value as if it's the gospel truth.

Well, it turns out it's not and there are very smart people who don't agree with Malcolm.

Bernard Dussault, Canada's former Chief Actuary, shared this with me after reading Malcolm's comments:
My main relevant comment thereon is that "risk premium" is just a definition (as shown below) and that it is inappropriate to conclude that the reward ( i.e. higher investment returns) obtained by investing net PSSA contributions in risky vehicles, belongs to the Canadian taxpayers. The employer's (government's) share of PSSA contributions is a deferred salary that belongs to PSSA members, not to the public. The "risk premium" definition exists only to portray that investment in equities provides higher investment returns. It is very appropriate that the PSPIB invests a material portion of the net PSSA contributions in equities, as the mandate of the PSPIB is to maximize the rate of return on investments while sticking to the prudent man rule.
Page 88 of the PSSA actuarial report on the PSSA as at 31 March 2017 (

A risk premium is the difference between the expected return on a risky asset (e.g. equities) and the expected return on a risk-free asset, such as the Government of Canada long-term bond mentioned above (footnote: Long-term federal bonds are considered risk-free since they have no risk of default. However, their market value is volatile and therefore long-term federal bonds do exhibit market and funding risk over the course of their life).
Another very astute actuary who read Malcolm and Philip's paper and thought there were some good sections, shared this with me:
While I find the contents of Mr. Hamilton and Mr. Cross’ paper academically interesting, we would note that assuming the cost of a pension can be predetermined with no investment risk is not practically feasible. The investment vehicles that would be needed to fund the defined pension obligations, that span 100 years, in the manner proposed do not exist. Furthermore, the authors do not take into account other risks besides investment risk such as retirement date risk and longevity risk, which defined benefit plans efficiently pool and significantly reduce the risk for members and employers.

As noted in the article “taking investment risk is a legitimate tactic” but furthermore I would argue that it is a necessity given the time horizon and the risk pooling features of defined pension plans which are not discussed in the paper.

Furthermore, I find the tone of the paper to be aggressive and unhelpful to a serious discussion. The calculations on the pricing of risk are also significantly overstated.
I pressed this individual, who prefers to stay anonymous, on why he feels the value of the risk transfer ($22 billion a year) is significantly overstated and he shared this with me:
The paper demonstrates an interesting method in assigning a value on the amount of risk being transferred to future taxpayers. The method, while not intuitive, is arguably valid. That said, even under the chosen method, the value of the risk transferred is significantly overstated. The authors equate a pension promise made by a provincial government to a federal real return bond. They are not the same. How can a pension promise by a provincial government be more secure than a provincial bond? First and foremost, provincial bonds are priced at a discount to federal bonds. This can be seen from the difference in yield on provincial bonds versus federal bonds. Second, the authors did not make any adjustment for the illiquidy of pension benefits. They cannot be traded easily and a member cannot just take their money out any time they wish. In fact, for anyone over age 55 it is not required that they be permitted to take the value of the pension out of the Plans. Finally, the relative security of a pension promise should be weighed against that of a provincial bond. If the province were to default on its bond issuance, it would likely suffer severe repercussions regarding its creditworthiness and its ability to borrow and/or roll over existing debt. Does the province face similar severe repercussions if it renegotiates on the pension promise? I would argue it would not, especially since this did not occur for the two provinces that did just that. For these reasons, the value of the risk transfer is significantly overstated and is dependant on the level of risk each plan is taking. It cannot be generalized in the way the authors of the paper did.

With regards to governance. The authors’ analysis of the governance problem is very interesting yet impractical. A trustee on the board of a pension plan has a fiduciary duty to the members and beneficiaries of the pension plan regardless of who appoints them. A fiduciary has no conflict on who they are meant to serve. The paper argues that since the government is the backstop to the pension promise, a fiduciary is incented to take as much risk as they can in order to better look after the member. The idea being that if the risk pays off, the member benefits from the risk-taking, but if it does not, then the sponsor would simply be required to remit more contributions. I would argue that the vast majority of trustees do not view their fiduciary duty from this shortsighted lens. Instead, they view it from the perspective that a pension needs to remain sustainable, otherwise it will cease to exist because of the burden it places on those who fund it. Trustees, however, need a clear definition of what sustainable means.
I thank this individual and Bernard Dussault for sharing their feedback, it's greatly appreciated.

Lastly, a blog reader who doesn't want attribution sent me his thoughts shortly after I published my comment last week, insights which are favorable to the authors' findings:
The large pension plans remind me of GE, once a leader in offering great but boring manufactured products, with a very clear and narrowly bounded strategy. Then a clever management discovered their AAA balance sheet could make them a bank like financial player, and down that track they went, using capital markets, leverage and derivatives in pioneering ways.‎ All went well, and the management became celebrity like, and poster children for high executive compensation. I think you can see the parallels of where this is going.

The authors are simply ahead of their time in pointing out where moral hazard leads. The pension industry would be well served by a dose of humility, and more explicitly recognize the advantages it has enjoyed.

DB Plans are a great idea. But risk subsidizing has nothing to do with those merits. I think the authors real point is that the investment risks being taken are too high (for the system in total), and structurally encouraged. They are correct.
I thank him for sharing his thoughts and agree with him that the pension industry is well served by a dose of humility.

As I stated in my comment on Canada's pension overlords, the senior managers at Canada's large pensions enjoy huge compensation and unlike their private sector counterparts, they have captive clients, huge scale and a long investment horizon, all of which are big advantages.

In this environment, big compensation can easily get to your head but it's critically important to remain humble and always question where you can improve your operations and processes without taking undue risks.

The purpose of this comment and the previous one was to further the dialogue and hear some opposing views. I firmly believe people have a right to express their views and we should debate them in an open, transparent and meaningful way.

I think I'll wrap it up there, if you have anything to add, feel free to email me at

Below, the last time stocks were this cheap they rallied nearly 20% in 12 months. Todd Gordon of and Strategic Wealth Partners’ Mark Tepper discuss whether this is a repeat performance. (Read my October comment on whether global stocks are cheap before jumping in!).

And extreme volatility could stick around into the New Year, and Credit Suisse’s Mandy Xu says to get used to these kinds of wild swings. We shall see what happens after the Fed meets next week.

Update: After reading this comment Malcolm Hamilton shared some more feedback with me:
  • Our paper is about public sector DB pension plans, i.e. DB plans covering public sector workers. We are not writing about the CPP or QPP (social security or social insurance plans) as is clear from references to pension boards and participation rates (if the C/QPP is considered a DB plan, then DB participation rates in the private sector are close to 100%, not 10%). I have no problem with the CPP or QPP. They are not supported by public subsidies. The contributions are made by employees (both directly and indirectly through their employers). The investment risks, which are small given the poor funding levels, are borne entirely by members, whether they know it or not.
  •  I accept your view of Canada's public sector plans as less bad than US public sector plans. I don't think that this invalidates my criticism, which is not about funding levels and investment management, but about poor governance and large, unacknowledged public subsidies.
Here is the thing that interests me. So far no one has objected to the following points.
  • The public bears most of the investment risk taken by public sector pension plans.
  • The public should therefore enjoy most of the returns generated by risk taking.
  •  In fact, the public receives none of the returns generated by risk taking.
Critics want to quibble about the $22 billion estimate. The paper anticipated this.
"There are many ways to improve our estimate. Some make it larger. Others smaller. All make it more complicated and contentious. For our purposes it suffices to round the number to $22 billion and make one simple point. The subsidy is large."
After conceding the three bullets, public sector pension enthusiasts cannot win a debate about whether the subsidy is small or large. If it is large, the plans are materially exploiting the public. If it is small, the plans are taking a lot of risk for no good reason.

The plans say that while interest rates are low, they need to take a lot of risk to make contribution rates affordable. I believe them. They say that they are well rewarded for taking the risks that they need to take. Again, I believe them. Put the two together with $1.4 trillion under management... and the subsidy is big!
I thank Malcolm for sharing his wise insights and all I say on risk-sharing is if these public plans are well managed and never chronically underfunded, isn't that enough of a public good? What more should we give the public for bearing any risk?

Similarly, if we figure out a way to lower healthcare costs, should we immediately lower taxes? Is this prudent public policy? I'm just playing devil's advocate and asking some tough questions if we are to believe the public is subsidizing our large public-sector pension plans to such an extent.