Friday, June 21, 2019

Short Takes: CPP Active Management, Portfolio Rebalancing, and more

Here are my posts for the past two weeks:

Should CPP Exist?

Credit Card Hopelessness

Living Debt-Free

Here are some short takes and some weekend reading:

Andrew Coyne explains the active vs. passive management issue for the Canada Pension Plan. He also digs into discrepancies in the claimed recent outperformance.

Canadian Couch Potato explains why it’s not necessary or desirable to rebalance your portfolio more than once per year. This advice is for those who rebalance based on time. I prefer threshold rebalancing, which is rebalancing whenever my portfolio is sufficiently far from its target percentage allocations. This is really only recommended if you can automate the process. It doesn’t make sense to do all the necessary calculations every day just in case you hit a threshold. I have my portfolio spreadsheet set up to send me an email when it’s time to rebalance, so I never have to look at it.

Tom Bradley at Steadyhand reminds us to focus on what really counts in investing. Hint: it’s not the short term.

The MoneySaver Podcast interviews Dan Bortolotti, the Canadian Couch Potato. They discuss passive index investing, ETFs, Robo-Advisors, and getting started investing.

Frederick Vettese shows in one example scenario that deferring CPP to age 70 is better than buying one of the new Advanced Life Deferred Annuities (ALDAs).

The Blunt Bean Counter explains the rules for the pension income tax credit.


  1. Interesting you reference CCP's rebalancing post. Andrew Hallam also posted on the subject yesterday. Here's a link (if links are accepted in comments):

    @Michael, did you do any calculations with respect to variance and returns and risk based on threshold rebalancing techniques compared to time-based ones?

    1. @Returns Reaper: There's no need for calculations. For time-based rebalancing, returns and variance are higher the longer the time between rebalancing trades. This is because stocks tend to go up faster than bonds. But none of this is terribly important because the purpose of rebalancing is to maintain the desired exposure to stocks. Using threshold methods is most efficient at maintaining desired exposure.

    2. Is this assuming strong positive correlation between stocks and bonds? Or do you believe this to be true regardless of the sequence of returns, which could be very different between asset classes in a portfolio?

      I'm sure your mathematical background is stronger than mine, but I would have thought this requires some form of back-testing to analyze. Of course, back-testing isn't perfect either, since past trends have no guarantee of continuing. And if back-testing shows there is some form of "free lunch" out there, it is sure to be exploited in the future thus removing it's effect.

      I suppose the explanation as to why it appears annual rebalancing provides higher returns AND lower variance is likely due to the past trends where when asset classes drift apart they continue to do so for a period of time before reverting to the mean. So rebalancing too quickly might cause you to sell recent winners (which will continue to "win" a bit longer) and buy recent losers (which will continue to "lose" a bit longer). But just because this may have been the case in the past doesn't mean it will continue.