In my previous post, I looked at the reasons why investors should occasionally rebalance their portfolios. When a portfolio contains both fixed-income and equities, rebalancing is mostly about risk management: if it also results in higher returns, that’s a secondary benefit. Having a disciplined rebalancing schedule also helps you control your behaviour and resist the pressure to chase performance.
That’s the theory behind rebalancing. But it doesn’t answer the most common question: “How often should I do it?” Like so many aspects of investing, the answer depends a lot on your situation. There are at least three strategies to consider.
1. Rebalancing by the calendar
Perhaps the most common rebalancing strategy is to make adjustments once a year. There’s nothing magical about that one-year interval. Indeed, many academic studies have tried to determine the optimal rebalancing period — monthly, quarterly, semi-annually, annually or even every two years — but the research is inconclusive. Since there’s no clear benefit to rebalancing more frequently, once a year should be fine for most investors.
If you’re using a tax-sheltered account, it doesn’t really matter which date you choose, but there are some practical reasons for rebalancing early in the year. If you make a lump-sum contribution to your RRSP just before the deadline, you can rebalance at the same time. In addition, just about all ETFs pay dividends or interest at the end of the calendar year. That means the cash balance in your account is likely to be highest in early January, and you can reinvest all that idle money when you rebalance.
If you’re investing in a taxable account, you might combine your buys and sells with a tax-loss harvesting strategy, which many people do at the end of the year.
Annual rebalancing has the benefit of being simple and convenient. The downside is that you may go many months with an off-target portfolio if a big market move occurs shortly after your rebalancing date.
2. Rebalancing by thresholds
Another common strategy is to rebalance only when an asset class drifts off target by a certain percentage. Financial author Larry Swedroe recommends the “5/25 rule,” which says you only need to rebalance when an asset class is off by an absolute 5%, or a relative 25%.
Following this rule, if your target bond allocation is 40%, you would rebalance anytime it was off by an absolute 5% — that is, above 45%, or below 35%. For asset classes with smaller targets, the “relative 25%” figure is more useful. If you’ve allocated 10% to emerging markets, you’d rebalance any time this fund dipped below 7.5% or rose above 12.5% (because 2.5% is 25% of 10%).
This method requires you to monitor your holdings more closely, but it works well for portfolios with a small number of funds. However, it’s not necessarily well suited to portfolios with small allocations to multiple asset classes, since it would result in a lot of tiny adjustments. For example, if your target is 4% emerging markets and you followed the 5/25 rule to the letter, you would rebalance any time this went up or down by a single percentage point.
Threshold rebalancing may also be harder emotionally, because you’ll be doing it after every significant market correction and rally, selling what’s hot and buying what’s not.
3. Rebalancing with cash flows
A third option is to rebalance whenever you add new money or make withdrawals. Let’s say an investor allocates equal amounts to Canadian, US, and international equity index funds, and that she contributes $1,000 a month to her account. The value of her funds are currently $15,300 Canadian, $15,500 US and $15,000 international. When she makes her next contribution, she would divide it as follows: $300 to the Canadian, $100 to the US, and $600 to the international fund. When she’s done, each fund will have exactly $15,600.
This strategy is ideal for monthly contributors who have small portfolios of index mutual funds. With larger portfolios (or smaller contributions) it may not be possible, because there won’t always be enough new money to prop up all the lagging funds.
You can also rebalance with cash outflows. Assume the above investor needs to withdraw $5,000 from her portfolio. She could redeem $1,700 of the Canadian fund, $1,900 of the US fund and $1,400 of the international fund. That would leave $13,600 in each asset class.
Advisors often use this strategy when managing large retirement portfolios. RRIF investors are required to withdraw a certain percentage of their accounts each year, which may require them to sell some of their assets. By trimming back whichever funds have risen above their target allocations, they can rebalance the portfolio at the same time.
Costs are always a factor
There’s no reason why you can’t use some combination of all three of these strategies. You might schedule an annual rebalancing date, but make an interim adjustment if an asset class moves off target by five percentage points during the year. And if you make any unplanned contributions or withdrawals to the account along the way, you can do a little ad hoc rebalancing then, too.
However, one of the most important concepts to remember is rebalancing often comes with costs, either in trading commissions or taxes. If you rebalance too often, the costs can easily overwhelm the benefits. In my next post, I’ll take a closer look at this trade-off.
Liked your discussion on when to do calendar rebalancing in the year ….
One thing I’ve been wondering about: would it be better to design portfolios with as few ETFs as possible to keep rebalancing costs down?
@Larry: I think there are a lot of good reasons to keep as few ETFs as possible in a portfolio. My Complete Couch Potato has six, and that is plenty for most investors. If you have a very large account and low brokerage fees, then you can make a good argument for including 11 or 12, but that certainly isn’t necessary.
People tie themselves up in knots trying to optimize rebalancing. As long as costs are reasonable and allocation percentages never get too far out of whack, there is little to choose among the different options. My preference is for using cash flow first and if this is not adequate, thresholds second. Unfortunately, choosing thresholds is a little tricky.
When choosing the amount of percentage deviation that triggers rebalancing, portfolio size matters. In mathematical parlance, there is a square-root relationship. For a portfolio 4 times bigger, percentage deviations drop by roughly a factor of two. Even this oversimplifies the situation somewhat. But perfect mathematical optimization gives little benefit. Just choosing threshold percentages that don’t lead to onerous costs is good enough.
@Michael: I agree completely. While rebalancing is important, there’s no optimal way to accomplish it, so any of these three ideas will work just fine. The method you choose should be the one that carries the lowest cost. More on that next week!
One additionnal factor for yearly rebalancing in January is TFSA room: I intend to rebalance at least in part when adding new money in the TFSA part of my portfolio. At the same time, I’ll be reinvesting my dividends from my ETFs (in all accounts).
Good post (since I follow these methods myself)! The various studies even show that leaving rebalancing to as seldom as every four years can improve returns. The 5/25 rule even with quite small allocations like 2% has not forced me to do any rebalancing at all the last several years. My rebalancing always happens as a result of cash inflows from distributions needing to be reinvested. Apart from the cost it is a hassle to do that, would rather just have DRIPs in place but only Claymore and BMO ETFs offer that possibility.
Most studies of the optimal rebalancing interval are useless. All they discover is that stocks have outperformed bonds over the long run. With this past data it is obvious that never rebalancing and letting your stock allocation percentage to rise gives the highest expected returns. The point of an asset allocation is, in part, risk control.
I typically do # 1, by calendar, and try not to worry about rebalancing otherwise. I enjoy buying more bonds when markets are high and buying more equities when markets are low; at least that’s what I did this year with our RRSPs.
I believe in a semi-annual rebalance, when I work with clients these costs of rebalancing are included in the annual fee.
“I enjoy buying more bonds when markets are high and buying more equities when markets are low” – this is nice in theory, but who can determine if market is high or low?! For example, now is it high or low?
I have kinda problem with rebalancing, because me and my wife have many accounts: TFSAs, RRSPs, SRRSP, LIRAs, RESPs, some plans from work and many of them have completely diferent objectives.
Another point that equities are also different, some ETFs and stocks (like blue-chip) has low beta , meaning closer to fix-income ; some with high beta – more risky…
@Gibor – I’m not perfect at this and don’t pretend to be good at it; I’m buying more bonds to rebalance my RRSP this season.
@ My Own Advisor, I just think that definition 50% equity and 50% bonds is too high level.
Equity can be Small Cap and Commodities on one hand and Preferred Stocks, Income ETF or Blue Chip on other.
@CCP, what is your opinion about Seasonal Rotation ETF (ex. HAC)?
@Gibor: HAX is an actively managed fund that is totally inappropriate for index investors — actually, that’s true of all Horizons AlphaPro products.
Although the Horizons fact sheet lists the management fee as 0.75%, the full MER last year was 2.80%, and in 2009 it was 3.05%. It also pays performance fees to its manager (like a hedge fund).
@ Canadian Couch Potato, OIC, thanks for info, just idea of “rotation” sounds kinda interesting :)
I’ve never rebalanced my different accounts, was just investing as per my feelings….today tried to do calc on high level and interestingly…if figures that exactly 40% I have in fixed income (saving accounts, bonds, GIC, security GIC, protected notes etc) and 60% in equites (ranging from Income funds to physical Palladium).
Need to dig more what % of my equites are non Canadian….
Just don’t know which category to put High Yield (Junk) bonds?
@Gibor, there are quant index funds that do what HAC does, using rule-based formulas rather than active management, e.g. GTAA. (GTAA is not 100% rule-based, however; there is an active component to select the underlying “universe” and perhaps more.) You might look into the options there, though I haven’t seen any passive or almost-passive tactical products yet that I’d personally be interested in.
I actually really like the cash flow option. I use a form of that for myself. Since I pick individual stocks and I have about 30 of them I can’t really just divide my money between them. Instead I buy one stock at a time that has fallen behind (assuming the fundamentals are good of course). That way I never really have to sell but I still end up buying low!
@Dan: I appreciate rebalancing by thresholds and setting a rule like the “5/25″ which Larry Swedroe recommends, but I am not sure if this would work for larger portfolios.
In your example of 40% bonds being sold at 45 or 35, for lets say a 500K portfolio, would be mean letting that asset drift by 25k, which seems excessive when you can rebalance at such a low cost these days.
Does Larry Swedroe, or are you aware of anyone else that recommends setting a dollar figure as well?
@Que: The guideline works just fine for large portfolios. But remember, it’s just a guideline. If you are adding or withdrawing money from the portfolio you can use those opportunities to rebalance before any asset class crosses the 5/25 threshold. And there’s nothing wrong with making a couple of additional rebalancing trades in a large portfolio if you are able to do so without significant cost or triggering taxable gains.
@Dan: I guess then, you can determine what a reasonable non-significant cost would be and then produce a dollar figure out of it, for example, choosing 0.1% to be a negligible cost, and having trades costing 10$, then the dollar rule would be 10k$ instead of the 25k (above)? Thanks, Que
@Que: You are welcome to choose whatever threshold works for you.
@Que: You’re neglecting spread costs in trading. These spread costs become significant for large portfolios. See the following post for a detailed analysis:
After understanding all the factors involved in threshold rebalancing, you may prefer to go back to periodic rebalancing. In my case, I just worked out the math once and coded it up in a spreadsheet.
@Michael James: How do you get the bid-ask spreads into your spread sheet? Thanks Que
@Que: It would be possible to use the Google spreadsheet importXML function to get them automatically from a web site, but I just take a look at a quote once to get the number of cents between bid and ask and enter that into my spreadsheet. On those rare occasions when my spreadsheet tells me to rebalance, I confirm that the current spreads match the ones entered into the spreadsheet before doing any trading.
@Michael James: I noticed I can get all the data from yahoo into excel, but it seems I can only get the bid size and ask size from the stocks in the US exchange. Do you know how I can get the size of the bid and ask for the Canadian exchange?
This is how I am getting the data into excel, if you know of an easier way please do let me know:
@Que: The only way I know of to get such data into a spreadsheet is with the importXML function in Google Docs from http://www.tmx.com
I don’t have any experience with collecting data in Excel.
@CCP: In a largish Non-Registered Portfolio, I just rebalanced after a year, so far so good, and did some minor tweaking. I noticed that I had accumulated some substantial dividends every quarter, which had been just lying there as cash until now. I have rejected the idea of applying to get them DRIPed, due to complications in calculating share cost base.
Therefore, in future, rather than letting cash sit there for the year, would it be reasonable to take this obligatory reinvestment need as an opportunity to do the necessary calculations at that time and make investments into the assets that had deviated furthest below my allocation plan targets? In other words, I have to spend the money anyway, and the account is large enough that the trading costs would be proportionately small enough to be economical — is there any detriment (as far as risk-benefit efficiency through diversification) to balancing as frequently as quarterly?
@CPP: With regard to the 5/25 rebalancing rule, with taxable accounts where capital gains taxes would be triggered on rebalancing, due you recommend a higher threshold?
Thanks again for all your efforts to help so many get a fair shake from Wall St. and Bay St.
@Tristan: Thanks for the comment. While one should always be sensitive to taxes, risk management should be a bigger concern. If a portfolio is way overweight in equities, it makes sense to rebalance even if it would realize a capital gain. One way to keep this to a minimum is to rebalance by adding new money to underweight asset classes to keep selling to a minimum.
quick question: i am not really sure if i understood correctly. Let’s say I own a couple of etfs, and a couple of mutual funds… these funds have portfolio managers that specifically take care of the rebalancing for me right? why would i need to rebalance my own portfolio again by myself if there are people who take care of that already?
@thenewbie: If you are using a balanced fund that holds stocks and bonds with long-term target allocations, then you’re correct: the fund manager handles the rebalancing for you. But if your portfolio includes, for example, a Canadian equity fund, a US equity fund, an international equity fund, and a bond fund, then you would need to rebalance that portfolio yourself to maintain your target asset mix.
I have invested approximately $65,000 in these three ETFs:
I would like my portfolio to follow a little closer to the Canadian Couch Potato Assertive portfolio, so I need to increase my VSB and VXC, and decrease my VCN. It’s coming up to rebalancing time and I am struggling to determine the best way to rebalance.
I intend on rebalancing via adding addition money to my brokerage account, not via selling. My dilemma is, VXC right now is doing incredibly well, and VCN is doing very poorly historically. I feel compelled to buy additional VCN shares because they are such a good deal ($26.25 right now, with a 52-day max of $31.29), and rebalance at a later date so I don’t miss out on cheap VCN stocks.
Should I resist my temptation to buy more VCN, and instead buy VXC (which is near it’s 52-week highest), or should I stock up on cheap VCN and purchase VXC later to fix my asset allocation?
So helpful, thank you for the advice.
I have a work DC plan which doesn’t charge for moving my assets between funds. I also do monthly contributions but those are preset at 10/30/30/30 so it wouldn’t really re-balance. Would it still make sense to use the 5/25 rule, or would it be beneficial to re-balance monthly? right now my bonds are at the 25% relative threshold but the other funds are between 1% and 3% absolute. So not quite there, but on the other hand there’s no cost associated with moving. Is there a point when adjusting too often has negligible return impact?
@IndexFan: From what you have described, an annual rebalance is almost certainly enough. There is no benefit to doing it monthly.
NOOB question. I set up my couch potato portfolio as a registered account last year and I am about to rebalance. I set up my account based on this 80/20 https://cdn.canadianportfoliomanagerblog.com/wp-content/uploads/2018/01/CPM-Model-ETF-Portfolios-Taxable-2017-12-31.pdf
I need to rebalance and I am looking at going with the 80/20 again based on his updated data feb 2019 https://cdn.canadianportfoliomanagerblog.com/wp-content/uploads/2019/03/CPM-3-or-5-ETF-Model-Portfolios-2019-02-28.pdf
When i sell and re-buy am I liable for any capital gains or are those only paid as tax when I pull cash out of my investment account and transfer it back to my bank account.
@Anon: If you are holding the ETFs in a registered account (RRSP or TFSA), there will be no taxable gains to worry about.