Tuesday, December 17, 2013

The Third Rail

Canada’s pension system is in trouble and we need to do something about it. This is the main message of the book The Third Rail, written by Jim Leech, CEO of the Ontario Teachers’ Pension Plan, and Jacquie McNish, senior writer with the Globe and Mail. The book is a fairly easy read and is worth a look.

The authors take a detailed look at pension crises in New Brunswick, Rhode Island, and The Netherlands, and describe how the problems were solved. A common theme is that the pension plans were changed to make benefit levels conditional on the returns on pension assets. On one hand this makes a lot of sense. We can’t expect pension backers (taxpayers or companies) to grow benefits faster than they can grow the savings set aside to pay those benefits. On the other hand, if we make cost-of-living increases conditional on pension asset returns, this automatically takes the pressure off pension administrators to manage the funds well. They can award themselves excessive fees or allow companies and governments to take pension contribution holidays, and the automatic reductions in cost-of-living increases will cover up the abuses. If pension benefits are going to be conditional, we need plan administrators to have strong financial incentives to run the plans efficiently and effectively.

Turning to the pension situation in Canada, the authors tell a parable about a farmer whose most productive cow dies. A chance encounter with a magic fish grants the farmer a wish. Does he wish to have his cow back? No, he says “I want my neighbour’s cow to die!” The authors claim that this is “the social dynamic of the pension debate” in Canada. Those who have no pension want to take away others’ pensions. The authors think we should instead try to build a strong pension system for all Canadians.

So the authors accuse Canadians of spitefulness, but I have practical concerns. If we take federal public servant pensions as a model, can we afford to extend such a plan to all Canadians? I think not. According to Statistics Canada, the average federal public servant retires at age 58, and only about 55% of the Canadian population is between 20 and 58 years old. If we eliminate the unemployed and those who cannot work, less than half of Canadians would be working, and this fraction will continue to decline over time. Do we really think we could run our economy including distributing food and other goods and keeping golf courses manicured if less than half of Canadians are working?

The only logical conclusion is that if we are going to design a pension system covering all or almost all Canadians, it cannot be as generous as federal public service pensions. We need a system that encourages Canadians to work later in life.

The authors describe a “longevity pension” being considered in Quebec where benefits begin at age 75. This would be in addition to QPP which generally begins around age 65. This kind of proposal makes sense to me. It says, we’ll take care of you after age 75, but if you want to retire earlier, you need to save some money.

Instead of expanding the existing CPP, I like the idea of creating another system that only pays benefits after age 75 and is entirely pre-funded. The idea is that you only get benefits in proportion to the amount you pay in. Young Canadians today could look forward to knowing that they’ll be taken care of after age 75. Until then, they have to work or save enough to live on until they reach 75.

In a discussion of expanding CPP, the authors suggest that “Ottawa could consider speeding up the delivery of enhanced pensions by temporarily bridging the gap to fully fund the new benefits.” Put another way, Ottawa could take some of the increased CPP payroll deductions of young people and give it to current retirees. I’m opposed.

The authors argue that the current system of voluntary pension savings in RRSPs fails in a number of ways. One of their main arguments is that too many people just don’t save. This is very true. And even if you don’t care about the welfare of non-savers, you end up paying for it anyway in the form of more GIS payments. Another strong argument is that “Canadian investors pay average annual management and trailer fees equal to 3.43 per cent of invested assets.” This may not sound like much, but over 40 years, this yearly fee compounds to consume 75% of assets!  (I've now investigated the references that led to the 3.43% figure and it appears to be incorrect. The origin of this figure seems to be from Table 1 of The Ambachtsheer Letter (April 2013), which indicates that the average MER in Canada is 1.93% and Trailer Fees are 1.50% (which sum to 3.43%).  These figures came from "CSA Discussion Paper and Request for Comment 81-407 Mutual Fund Fees," December 13, 2012.  Page 11249 says "At the end of 2011, the asset-weighted average MER of all Canadian mutual funds was 1.93%."  Footnote 10 says "Canadian no-load funds may pay trailing commissions of up to 1.50%."  While 1.50% may be a maximum for trailing commissions, it is far from typical.  Further, trailing commissions are a component of MERs, so it makes no sense to add in the trailing commissions again.  So, the 3.43% figure used by Leech and McNish makes no sense.  Further, it ignores other fees such as front loads and deferred sales charges.  In the end, the conclusion that total fees are very high still holds, but the 3.43% figure should be ignored.)

In contrast, “the typical defined benefit pension fund management expense ratio of 0.4 per cent” consumes only 15% of assets after 40 years. This is obviously far better, but still seems high when we see that Vanguard manages to run funds with much lower expenses. Perhaps pension plan administrators face costs that Vanguard does not.

In summary, I think the authors are right that we need to face our pension problems. They are right when they say that existing defined benefit pensions have sustainability issues. They are also right when they say there is a problem with so few younger Canadians covered by pensions. However, if we are going to have mandatory pension systems (like CPP), the focus should be on large enough benefit amounts to provide a decent standard of living but starting late in life (at least age 75). Attempts to shift money around to give us all great pensions in our late 50s or early 60s are doomed to failure.


  1. I wouldn't be surprised if pensions had a lot of other costs to deal with. AFAIK, Vanguard has to track their holdings, rebalance, track who owns what shares, and distribute dividends on a quarterly basis. The brokerages partially share in some of that burden.

    A pension fund has to do that investing stuff, and also track years of service, who dies, audit who's alive and eligible to receive pensions, process divorce settlements and split the cheques, and run projections of what their asset base is and future distribution needs... 0.4% sounds like about the right ballpark to me.

    1. @Potato: Vanguard runs some actively-managed balanced mutual funds that charge 0.25%. If CPP has higher costs in dealing with the public, then 0.4% makes sense. I don't know enough about the work that CPP has to do to price these things.

    2. Unbiased information is hard to come by, but my understanding is that the effective management expense of the CPP is something like 0.55%. The managers at the CCP Investment Board pick other managers to manage chunks of the CPP assets and those outside managers charge fees. And this is just costs at the CCPIB, I don't think it includes any of the actuarial and accounting costs mentioned by Potato. And has the CPP been beating the indexes with all that wiz bang professional management? Nope, a 10 year annualized return of 7.4% vs. 8.3% for the TSX, 7.7% for the S&P500. CPP has a 5 year return of 4.2% vs. Vanguard international index ETFs ranging from 12-16%. Even Vanguard balanced funds look better than the CPP. OK, maybe these comparisons with US funds are off due to US$ depreciation, but it looks to my like active management in the CCP is just leaching from investors as with most other actively managed investments.

      I agree with you Michael, though all these pension plans changes appear well meaning, they tend to just transfer money from younger generations, and tend to just buy current votes. I like solutions that involve education and more challenging financial lives for those who can't bring themselves to save and plan. It might take a few generations for the lessons to really sink in, but forcing everybody into a sub-optimal investment scheme and expanding it gives the wrong incentives.

    3. @greg: Those are some depressing numbers. With politicians' desire to appear to be improving pensions for both current workers and current retirees, public descriptions of any new plans are usually vague on the point of whether there will be any additional money transferred from current workers to current retirees.

    4. Greg are you comparing the returns of just the part of the CPP fund that's in the market with the market returns, or the return of the entire CPP investment fund with the market returns?

      I would hope that there are requirements to keep large amounts of the CPP fund in cash and bonds so they can sustain payments even through prolongued market crashes. I would hope that the CPP must invest more conservatively than even most balanced funds.

    5. @Bet, gone are the days when the CPP was a pay as you go system invested 100% in bonds (thankfully, because it was almost a Ponzi scheme with the baby boom only paying enough to fund the retirements of much smaller generations before them, not enough to fund there own retirements). Since reform in 1997, contribution rates have been dramatically increased so GenXers and younger get to make up the difference.

      The mandate of the CPP Investment board is pretty much what anybody should have for their retirement savings- to make inflation + 4% long term. It is unlikely to achieve this with more than 1/3 in fixed income and the CPP fixed income proportion has ranged from 25% to 33%. So comparing with a balance fund with 50% in fixed income would actually be unfair to the balance fund.

    6. Ok, I guess it depends which balanced funds you were speaking about as I've seen ones that only have 25% in bonds and fixed.

      I'm still not clear about your numbers on this part though: "[CPP had] a 10-year annualized return of 7.4% vs. 8.3% for the TSX, 7.7% for the S&P500." Do you mean that CPP had a 7.4% return on only the part of the CPP fund invested in equities?

      If CPP managed 7.4% on the entire fund including a large chunk of fixed income, while the markets (with no fixed) only averaged 7.7-8.3%, then it looks like the CPP did very well indeed.

      By your own measure of "inflation + 4%" it looks like CPP is meeting its targets, which is somewhat reassuring. Personally, we've never expected CPP to be available when we need it, so we save more and hope for the best. We can always spend more money in retirement if someone gives it to us. : )

  2. With people living longer and the ratio of time spent (potentially) in retirement vs the working career increasing, delaying the average age of retirement is almost inevitable. I think it's a matter of how much and how long we'll kick and scream about it being the only question.

    Or mandatory age limits. On life. But I imagine that won't be too politically saleable as a fix.

    1. @Preet: I can just see it. I'll be retired and having to dodge gangs of young workers who are trying to reduce their CPP payroll deductions :-)

  3. I would be happy to fund my own early retirement to say age 80 if I knew that after that point, CPP would pick up the slack and make sure I had a decent income that was guaranteed.

    1. @Tara C: I feel the same way. Every time I try to plan out my retirement spending I change my mind on whether to use 90,95, 100, or 105 as the end age. It would be a lot simpler if I just had to plan to 80.

      The problem is that so many Canadians who haven't saved much, but who hate their jobs, are trying to figure out how to retire at 60. CPP rules create the illusion that you can retire at 60, but few people would be happy with that standard of living.

    2. @Richard: I'd say that the current RRSP and TFSA rules are quite a strong incentive to save, but they make little difference for many people. I think what we need is a change of perception. People need to believe that they will have to work until they are at least 75 if they save nothing. The problem I see right now is people who are getting close to 60 with little saved and trying to figure out how to get by on CPP until they can draw OAS at 65. The ones who pull the trigger on this plan then find that money is extremely tight after 65. So they complain that they need more. We need to put an end to dreams of retiring at 60 with little saved.

    3. The comment above is a reply to Richard's comment:

      Maybe people could accept a higher minimum age for government plans if there were also incentives to save independently so most people could afford to retire younger. The conservatives might be hiding a plan to throw in a quick adjustment to the minimum age if they double the TFSA room before the next election :)

      Or it could just be done as an extension of the increased benefit for people who delay taking their CPP payments. If any increases were simply made by adding on options to delay the payments longer and get more, it could work out well without taking away options people have now. Then CPP beneficiaries would face a wide range of choices similar to anyone who builds up their own portfolio. Unfortunately a lot of people would probably think that's not fair enough.

  4. The following exchange is reproduced to remove broken links:

    ----- Richard December 17, 2013 at 10:35 AM

    I used to think that the most reasonable way to expand the CPP would be a "reset" where all additional contributions were directly attributed to the employee, without the drag of past shortfalls. But it makes even more sense to target a higher age at the same time which would make it cheaper. Another easy way to simplify retirement planning would be to create a new plan that has lifetime income after a very high age such as 95 so people know they only have to save for a fixed number of years in retirement. The risks at that point become so uncertain that it's hard for one person to manage them but cheap for a collective plan to cover them.

    There could be sound economic reasons for wanting to take away others' pensions. For one thing many of them are associated with unionized government workers who already have relatively good salaries (at least in the lower ranks) and relatively few incentives to perform their jobs better. And on top of that, all of us have to pay for them so it is easy for someone to see that as a personal cost.

    In a broader sense it might be argued that replacing the best pensions with a mandatory contribution to a universal plan would at least ensure that everyone would enjoy the same benefits, even if they would have to be lower.

    Not all of those arguments are necessarily reasonable. Governments could very reasonably have criteria for hiring other than simple productivity for example, and it could be worth paying for those criteria. And it's questionable whether someone wielding a pitchfork and torch is really thinking through the economic implications.

    Personally I think there could be improvements in every area. Chances of seeing that all at once are low :)

    ----- Michael James December 17, 2013 at 11:09 AM

    @Richard: I think we could afford a modest lifetime income starting at age 75 or 80 as long as we don't try to also provide most of a lifetime income starting at 60.