When to Rebalance an Investment Portfolio

investment rebalancing

Rebalancing the investments in a portfolio helps to manage risk in the portfolio over time. A systematic approach to rebalancing also encourages disciplined investing rather than emotion-driven portfolio adjustments. This post explores the mechanics of how frequently to do it and how aggressively.

When should I rebalance my portfolio?

As humans, we are innately attracted to market-timing (attempting to “buy low and sell high” thinking that we may know where the market will head next). It is hard-wired into our psyche to try to recognize patterns to predict future events. Unfortunately, the future is unpredictable. This is coupled with the powerful emotions of greed and fear. These may have been helpful from an evolutionary basis, but they can fuel poor investor performance.

The act of rebalancing a portfolio, in practice, involves buying and selling investments. The more that we can link that to an objective trigger instead of our intuition and emotions, the better. The trigger to rebalance a portfolio is usually either time or an event.

Time as a rebalancing trigger.

A time trigger uses a set date to rebalance. That could be annually, quarterly, or monthly. No specific frequency has been definitively shown to produce better returns or volatility reduction than another. For example, Vanguard compared monthly, quarterly, annually, and never rebalancing over an 80 year period. Their results are summarized in the table below.

More frequent rebalancing resulted in a tighter tracking to the desired asset allocation. It also reduced the annualized return slightly in exchange for reducing the volatility. This is the expected result of rebalancing. Some theoretical investment return is given up to hopefully have a better real-life return. The real-life return would presumably be improved by reducing the performance gap from bad investor behavior.

The main benefit was seen from annual rebalancing. More frequent rebalancing did not reduce volatility further. In fact, frequent rebalancing comes at a cost.

With each rebalancing event, there would be buying and selling. More frequent buying and selling increases the transaction costs from brokerage fees and the bid-ask spread. If there are capital gains realized in a tax-exposed account, then there would also be more tax drag. These worsen actual after-tax performance. The effects of rebalancing frequency on transactions in the Vanguard model are shown below.

Importantly, rebalancing a portfolio also takes time. That may be a good thing for investors who feel compelled to tinker with their portfolio in some way. For most of us, less time spent tending to our investments is better.

Thresholds as a rebalancing trigger.

Since rebalancing frequently is not helpful, and may even trigger fees or taxes, using a threshold is another approach. Using a time-based strategy, you would be buying and possibly selling at the designated time even if your asset allocation is only off by 1%. Probably not worth it!

Perhaps you would be willing to accept some drifting under and over your planned allocation. The risk may fluctuate slightly more, but you would avoid unnecessary transactions. This is where a threshold trigger fits in.

A commonly suggested threshold to trigger a rebalance is if an asset allocation is off by 5% absolute (relative to the portfolio) or 25% relative to itself. The 5/25 rule. That threshold could be breached in either direction with an asset being too high or too low compared to its target.

How does the 5/25 rule work?

For example, if my Canadian equity target allocation is 10% and it currently sits at 13%. The 3% difference falls under the 5% absolute target. So, that does not trigger a rebalance. However, it is over the 25% relative-to-itself target (1.25 times 10% equals 12.5%) and that does trigger the rebalance. A rebalance is triggered if either of the 5% or 25% threshold is breached.

Which threshold of the 5/25 rule gets breached depends on how large the asset allocation is. Larger allocations tend to breach the absolute threshold first. For example, if our bonds allocation is 40% and currently sits at 46%. The 25% relative threshold would be 1.25 times 40% equals 50%. The 46% bonds does not breach the 25% relative threshold (it is under 50%). However, 6% is greater than the 5% absolute threshold – that triggers a rebalance.

I know. Confusing. Below is a theoretical portfolio below and I have highlighted where the 5 or 25 of the 5/25 rule has been triggered.

How does the 5/25 rule perform compared to other thresholds?

As outlined in the preceding post, that would depend partly on the time-frame used and whether the assets were in trending or choppy markets. Rates of return and correlation of the different assets in relation to each other also have an impact. Attempting to tackle that for a multi-asset portfolio would cause my brain to explode. Messy. Most readers would tune out or be grossed out. Let’s not go there.

However, I have a hard time accepting “rules of thumb” without testing them. So, let’s look at different absolute thresholds using a very simple portfolio of US Large Cap (Stocks) and Intermediate-Term US Treasuries (bonds). A daily closing price above the threshold triggered a rebalance.

threshold performance

Moving from a 1% to a 5% threshold cuts down on the number of transactions dramatically. The overall effects on return were minimal regardless of the threshold used. Volatility gradually increased with increasingly loose thresholds by a minuscule amount. So, it seems that the 5% part of the 5/25 rule held up well over this time period. You could probably even go a little looser up to 10 or 15%.

In the appendix of the Vanguard study already referenced, the results were similar over a shorter time period (1989-2009). They also did not find a difference in whether you monitored for threshold breeches daily, monthly, quarterly, or annually. Returns and volatility were similar, but transactions increased with more frequent monitoring.

A Pragmatic Approach To Triggering a Rebalance

The above data shows that there is no need to rebalance more frequently than once per year. Most people contribute or withdraw from their portfolios at least once a year. So, it makes sense to do a rebalance if you are making a contribution or withdrawal. This is sometimes called a “cashflow triggered” approach. You rebalance when you have the cash to invest, or you need to access it.

During the accumulation phase of our investing, that basically means rebalancing whenever you have enough excess money that you don’t need in the short-term to invest it for the long-term. Aggressively “investing” money that you will need in the near term is really gambling.

How much accumulated cash is “enough” to trigger a contribution is variable. A question of fees, cash flow, and convenience.

For someone making small contributions and paying transaction fees.

There is a balance between waiting and missing time in the market against having a larger proportion of the contribution go to fees. An investor using $100 to buy an ETF with a $5 brokerage fee loses 5% of their money right off the top. For investors making small contributions, an online broker with free ETF purchases can help that problem.

Many high-income professionals make large contributions, making the fees a negligible issue. Cash flow and convenience are the main determinant.

An employee collecting a regular paycheck will often use automated contributions for both convenience and dollar-cost-averaging. Automation also minimizes the potential for bad behavior. However, investing upfront whenever you have the money has a slight advantage over monthly dollar cost averaging in the long-run.

For many professionals and business owners, money comes in fits and spurts. That is a result of variable revenue, but also large movements of money to pay personal and corporate tax installments.

What do we do personally? We use cash flow for timing and a threshold as our trigger for action.

We stockpile enough money and have enough certainty that we won’t imminently need it about four times per year:

  • January: Front-loading our RESP, TFSAs, RRSP, and spousal RRSP
  • April: We invest the excess, after settling our personal income taxes.
  • September/October: We have accrued some money and know where we stand after our corporate taxes are settled. Our corporate fiscal year ends July 31st.
  • December: A combination of cleaning up our budget, paying our final quarterly personal tax installment, tax loss selling (if the opportunity), and planning for the next year.

In addition to cash flow limiting us to about four rebalances per year, I don’t really want to spend time doing it more frequently than that. I have better things to do.

Incorporating a threshold trigger component.

Even though we are rebalancing quarterly when we have cash, that does not necessarily mean that we rebalance to perfection. We accept up to a 5% drift above the assigned allocation for our assets. I am too lazy to use a relative trigger and it does not add much for us. The 25% relative trigger can be useful for those with multiple holding/assets with small allocations of under 10-15% each.

In practice, this means that most of our rebalancing is simply adding our new contributions to an area that is lagging. However, as the portfolio becomes large relative to the size of the contributions, some selling may be needed to come back into the target range. Last year, we had to sell some of our US equity during our early fall rebalance. Fortunately, we were able to offset that with some tax-loss selling a couple of months later.

Rebalancing is important. However, it does not need to be frequent and it must be simple enough to execute.

Rebalancing investments can be very simple for those using a single account type or holding the same asset allocation in each account rather than trying to optimize asset location. Rebalancing within an account during the “buy-only” accumulation phase can even be semi-automated with a service like Get Passiv. An asset allocation ETF can also make it easy for those who want a simple pre-made solution at the slight cost of fees and some tax inefficiency in tax-exposed accounts.

Those facing a larger tax burden, using multiple account types, or with a large portfolio may need to consider rebalancing across account types. That can be a bit more complex. Still, it needs to be simple enough to easily execute. We’ll look into how to do that further in the next post i this series.

12 comments

  1. Great post! Right on the money;) similar to you I don’t get too concerned with some allocation drift. New contributions tend to offset this. It was helpful to see the effect of different rebalancing thresholds as I just accepted the 5/25 as optimal but it seems even without new contributions a drift over 5% is still reasonable. Kinda like wearing a stretchy belt.

    I like your cash flow discription for when to add$. I do this as well but each month I am willing to add as long as there is at lest 10k to invest. It all depends on the size of your portfolio. When i started out I used td e-series and avoided commissions all together.

    1. Hey Phil,

      I have been a bit loosy-goosy until the past year because my contributions always brought me reasonably close. The sustained US run did finally out-pace me this fall. Moving forward, I am going to use the 5/25 more deliberately because my portfolio is now getting large relative to my contributions and harder for me to just buy my way out of an underweight position. I will need the discipline it should instill.

      Discipline is hard for me sometimes. Probably contributes to why I am finding stretchy waistline pants more appealing with each passing year.
      -LD

  2. Another excellent post!

    I am currently struggling with my bad investor behaviour – my goal for my Canadian equities is 20%, but it’s been sitting at 10% for awhile. I am overweight US and EM at the moment but my brain is struggling to sell the winners and buy more of the loser. Seeing the Vanguard data which suggests never rebalancing had the best return (with the worst volatility) doesn’t help 🙂

    Do you have any cognitive tricks you use to overcome the fear of missing out on gains in your winners that you had sold?

    1. Hi David,

      The thing with the Vanguard data and the longer data-set that I used in this post is that it was balancing stocks vs bonds. So, as shown in last week’s post that is expected to decrease return slightly by trimming stocks to add bonds due to the different long-term returns. Two equity markets expected to have similar long term returns don’t have that drag-down effect of one upon the other over the long-term. Rebalancing may actually increase performance if they aren’t well correlated by forcing some buy low and sell high behaviour. So, rebalancing between US, Canada, EMM, and non-NA developed probably has better returns long-term than never rebalancing. Things do move in cycles even though not in sync, but they are often long (years).

      That said, I still struggle with behaviour too. It took me two years to get my Canada up to my 15% goal (just hit it last week with a deposit) and last year I slightly trimmed my US. What works for me is using a threshold. That way, I don’t need to sell often unless a winner is a sustained one. I am letting winners run a bit (going with their momentum) and not trimming them unless they are getting long-in-the-tooth which usually takes a fairly long sustained out-performance. Fear of loss is usually kicking in by then to counter-act my greed anyway. The other trick is to not look at your portfolio much other than to add or take out money. Easier said than done. Maybe I need a shock-collar or something if I cross the threshold like those invisible dog fences 😉
      -LD

  3. I like the various options for rebalancing you have outlined. Honestly never heard of the 5/25 rule until now so that was a nice nugget of information.

    Personally I just decided on my own that I have “tolerance bands” for each asset allocation. I rebalance when the band is +/- 10% off of the desired allocation. For example, if I have an asset class that I want 20% of total portfolio, then I would rebalance if it is less than 18% or more than 22% (+/- 2%). Even though the percentage seems small, because of the size of the portfolio it really does take a lot of money to move the needle towards a rebalancing event (in fact I have only hit it 1x in several years).

    1. Hey Xrayvsn,

      What you are describing, I think, is what I would call a 10% “relative” trigger using this nomenclature (movement outside the “tolerance band” of 10% relative to itself). Good for a large portfolio with a whole bunch of “small” positions. If holding many well-correlated assets, I suspect people would very rarely hit rebalancing triggers. Of course, holding poorly correlated assets is where diversification and rebalancing becomes the most powerful.
      -LD

  4. I also use the 5/25 rule, but have never had to sell anything, just using incoming cash flows to keep the portfolio enough in balance.

    Bill Bernstein advises that if you have to sell to rebalance in a taxable account to use 10/50 bands to cut down on losses due to capital gains taxes. Larry Swedroe also suggest something similar – I think 10/50 and only selling down to 5/25.

    1. Thanks Grant. I like that idea. The rebalancer that I have been working on is going to work similarly. Rebalance to target with contributions or in sheltered accounts if possible and if not then allow drift and if it goes over the threshold and selling is needed, then just selling enough to hit the edge of the threshold (rather than to target). It likely gives a slight bias to higher equity, but I don’t think that is bad thing as long as close.
      -LD

  5. Hi Loonie Doctor,

    Thank you for your very educational posts. I have learnt a lot from you website and I really appreciate all you put into it for our benefit.

    As a new MD, I am aware there are many things I need to better my future financial grounding. I have meet several wealth management groups to hear them out. I am attracted to get their help in thing like tax management, financial planning, estate planning, IPP, and risk management.

    Is it possible to find a group that can help me with all the other aspect but not the investment part if I wanted to go DIY for the investment part? It seem to me that the groups that I have met so far want to manage my investments in addition with helping with all the other aspects I mentioned.

    Any thought? What do you do yourself if you don’t mind me asking? Thank you kindly

    1. Hi Ahmed,

      Thanks for reading the blog and I am glad you are finding it useful. The right financial advisor is good for all of those things you mention (except maybe IPP – they may or may not be useful compared to a well-planned conventional portfolio for most and is not something a new physician needs to worry about).

      The key is value – are those services worth the fees you pay? They may be for someone who wants to do nothing. It can be very cost effective for someone who is starting out where the financial planning heavy-lifting takes place and the portfolio is small. However, the advisors taking a professional on with lots of front-end work while be justifiably annoyed if you then leave them when the plan is mostly done and the portfolio is starting to pay them really well. Even a %AUM model is expensive with a large portfolio – 1% a million dollar portfolio is $10K/yr. If you can do $10K of work in a few hours managing it yourself, then that is more lucrative than any medical fee I know of. For some, it is still money well-spent but for others it is not. It is really a personal choice – just be aware of what you are paying and what you are getting in return.

      You can split up the financial planner, tax planner, and portfolio management roles.

      Financial planning: A “fee-only” advisor will charge you a flat rate based on a specific task or the time involved. For example, $3-5K to do a comprehensive financial plan. You are then responsible to execute it (buy the investments, appropriate insurance, pay your debt etc). They can also do a re-evaluation for a fee periodically (doesn’t need to be every year!) A “fee-based” advisor charging a %assets under management will cost more, but execute more of the plan for you. A full-service option. I review what to look for in an advisor and the different advisor fee models in the linked posts.

      The biggest issues with tax planning are salary/dividends/incorporation and managing your investments tax efficiently. For the first aspects, you are already going to be paying an accountant. You need a good accountant. The one aspect the accountant can only do their best with the mess that you hand them is around setting up and managing your portfolio tax efficiently. The Robocorp tools that I have put on this sit suggest how you can build a tax-efficient portfolio very simply with four or five to six ETFs. After set-up of accounts at a discount broker, it can seriously be done in a few hours per year.

      What do I do personally? I use my accountant for tax planning advice. He is very responsive via email and we meet face to face once per year as part of the costs for my corporation. I have a financial advisor at MDM that still helps me even though I am DIY for investing via MD Direct there. They were particularly helpful early on. If I did not have an advisor and hadn’t learned so much via writing this blog (or making mistakes already!), I would have found a fee-only advisor to help with an initial plan and then as needed. Probably once every five years or so. They are harder to find in Canada, but they are gaining in popularity. The physician financial independence (Canada) facebook page is also a good peer-to-peer resource for those looking for an advisor and for many general financial planning questions.
      -LD

  6. Hi LD:

    Another great post! FS recently posted 10 year return of various asset classes and REIT was #1. I see white coat investor offering private equity deals in RE and wondering if we have something like it in Canada. I did do a landbanking investment 10 years ago and the money is still tied up, so definitely risk involved.

    WRT to rebalancing, I cannot remember where I read it (Justin bender or his group)? Anyways, it showed very little overall return difference at 15%, 10%,5% rebalance threshold over time. I guess the key word is over time.

    Off topic, I thought capital gain though MF and my own selling are counted the same way for tax purposes. Asked my accountant to do a CDA payout but he suggested MF selling gains could be reduced due to other factors? Not sure if you know.

    Have a good long weekend.

    1. Hey BC Doc,

      I haven’t mentioned it I think, but I do hold a small portion of REITs and am adding that into the options for my more complex portfolio builders and rebalancing tools that I am working on. There are some private equity deals in Canada. I know MD Management has them intermittently. So far, I haven’t tried it because I don’t like having my money tied up like that and now with the corp tax changes, I am leery of anything where I don’t have control or predictable distributions.

      Not sure about the capital gains in a MF being treated differently. I am honestly a bit ignorant on this one.

      The data I looked at didn’t show much difference between 5-15% thresholds either. I like that! Kind of like not having to titrate insulin drips for tight glycemic control in the ICU anymore.
      -LD

Leave a Reply

Your email address will not be published. Required fields are marked *