Portfolio Turnover
We know that true couch potato investors who just buy and hold indexes are rare. One indicator of deviation from couch potato investing is the percentage of a portfolio invested in individual stocks or other non-index securities. Another such indicator is portfolio turnover.
High portfolio turnover can be a sign of market timing. Even investors who think they are just changing their minds about asset allocation percentages may be doing it so often that they are effectively making active market timing decisions.
I decided to look back at my own trading history to examine my portfolio’s turnover since I switched to mostly indexing nearly six years ago. It turns out to be surprisingly difficult to decide how to measure portfolio turnover for this purpose. To show what I mean, I’ll go through the different categories of trades I’ve made.
1. Adding new money
By far the most common trade I’ve made is to buy more of the index ETFs I own with new money I’ve saved or with dividends. However, it hardly seems fair to count this as deviating from passive investing. In each case, I added money to ETF positions that were below their target percentages.
At the very least, for someone like myself still in the accumulation phase, I should probably just focus on selling transactions to measure portfolio turnover.
2. Currency exchange with Norbert’s Gambit
A few times over the years I’ve needed to exchange Canadian dollars for U.S. dollars or vice-versa to make an ETF purchase. Most brokerages hide high fees in their exchange rates. To save money, I use Norbert’s Gambit to exchange currencies. This involves finding a security that trades in both Canadian and U.S. dollars (I’ve used DLR and RY in the past), buying the security with one currency, and then selling it in the other currency.
Technically this is portfolio turnover, but it hardly seems reasonable to count it as active investing. It just happens that using Norbert’s Gambit for large currency exchanges usually costs less than hidden brokerage currency exchange fees.
3. Rebalancing
Occasionally, my portfolio tracking spreadsheet emails me to indicate that 2 or more of my ETFs have deviated by enough from their target percentages that I need to sell an ETF that is too high to buy and ETF that is too low.
Again, this is technically portfolio turnover, but should it count as active investing? If I was tinkering with my asset allocation percentages or chose the timing of the rebalancing then it should count as active investing. But if the rebalancing is triggered by a purely mechanical set of rules, it’s hard to see why it’s not part of a sound passive strategy.
4. Selling individual stocks
Now we’re getting to some real portfolio turnover. After I sold off my 20 or so stocks nearly six years ago, I hung onto two of them (BMO and BRK). A couple of years later I sold all the BMO stock to buy more ETFs. Over the years, my wife and I have been leaking out the BRK stock as well. The problem in this case is that we had substantial capital gains in taxable accounts, and we didn’t want to recognize all the gains in one year. But it’s still portfolio turnover.
5. Changing ETFs
At different times I’ve held Canadian ETFs XIU, XCS, and VCE. I’ve since sold them all to buy VCN. But how much of this is actual portfolio turnover? VCN has a large overlap in underlying stocks with both XIU and VCE. It only makes sense to count the degree of change in the underlying stock holdings as portfolio turnover. However, pretty close to all of the XCS sales are portfolio turnover.
On the U.S. side, at one point I decided to trade U.S. small cap stocks in the form of ETF VB for U.S. small cap value stocks in the form of ETF VBR. Roughly have of these trades should count as portfolio turnover.
Conclusion
Initially I intended to report my portfolio turnover each year since I started mostly indexing, but I’ve decided now not to do this. This is because some of the thinking I explained above seems too close to rationalizing.
If I go back to a stricter definition of portfolio turnover that counts all stock sales except for Norbert Gambits, my 2015 turnover was about 8%. Add to this the less than 5% I still have in BRK, and we get some sort of measure of my active investing in 2015. Most other years were higher than this (except 2011 where I had zero turnover).
I hope to keep my active investing at lower percentages than this in future years. If you think of yourself as a passive index investor, you may find the exercise of calculating your portfolio turnover illuminating.
High portfolio turnover can be a sign of market timing. Even investors who think they are just changing their minds about asset allocation percentages may be doing it so often that they are effectively making active market timing decisions.
I decided to look back at my own trading history to examine my portfolio’s turnover since I switched to mostly indexing nearly six years ago. It turns out to be surprisingly difficult to decide how to measure portfolio turnover for this purpose. To show what I mean, I’ll go through the different categories of trades I’ve made.
1. Adding new money
By far the most common trade I’ve made is to buy more of the index ETFs I own with new money I’ve saved or with dividends. However, it hardly seems fair to count this as deviating from passive investing. In each case, I added money to ETF positions that were below their target percentages.
At the very least, for someone like myself still in the accumulation phase, I should probably just focus on selling transactions to measure portfolio turnover.
2. Currency exchange with Norbert’s Gambit
A few times over the years I’ve needed to exchange Canadian dollars for U.S. dollars or vice-versa to make an ETF purchase. Most brokerages hide high fees in their exchange rates. To save money, I use Norbert’s Gambit to exchange currencies. This involves finding a security that trades in both Canadian and U.S. dollars (I’ve used DLR and RY in the past), buying the security with one currency, and then selling it in the other currency.
Technically this is portfolio turnover, but it hardly seems reasonable to count it as active investing. It just happens that using Norbert’s Gambit for large currency exchanges usually costs less than hidden brokerage currency exchange fees.
3. Rebalancing
Occasionally, my portfolio tracking spreadsheet emails me to indicate that 2 or more of my ETFs have deviated by enough from their target percentages that I need to sell an ETF that is too high to buy and ETF that is too low.
Again, this is technically portfolio turnover, but should it count as active investing? If I was tinkering with my asset allocation percentages or chose the timing of the rebalancing then it should count as active investing. But if the rebalancing is triggered by a purely mechanical set of rules, it’s hard to see why it’s not part of a sound passive strategy.
4. Selling individual stocks
Now we’re getting to some real portfolio turnover. After I sold off my 20 or so stocks nearly six years ago, I hung onto two of them (BMO and BRK). A couple of years later I sold all the BMO stock to buy more ETFs. Over the years, my wife and I have been leaking out the BRK stock as well. The problem in this case is that we had substantial capital gains in taxable accounts, and we didn’t want to recognize all the gains in one year. But it’s still portfolio turnover.
5. Changing ETFs
At different times I’ve held Canadian ETFs XIU, XCS, and VCE. I’ve since sold them all to buy VCN. But how much of this is actual portfolio turnover? VCN has a large overlap in underlying stocks with both XIU and VCE. It only makes sense to count the degree of change in the underlying stock holdings as portfolio turnover. However, pretty close to all of the XCS sales are portfolio turnover.
On the U.S. side, at one point I decided to trade U.S. small cap stocks in the form of ETF VB for U.S. small cap value stocks in the form of ETF VBR. Roughly have of these trades should count as portfolio turnover.
Conclusion
Initially I intended to report my portfolio turnover each year since I started mostly indexing, but I’ve decided now not to do this. This is because some of the thinking I explained above seems too close to rationalizing.
If I go back to a stricter definition of portfolio turnover that counts all stock sales except for Norbert Gambits, my 2015 turnover was about 8%. Add to this the less than 5% I still have in BRK, and we get some sort of measure of my active investing in 2015. Most other years were higher than this (except 2011 where I had zero turnover).
I hope to keep my active investing at lower percentages than this in future years. If you think of yourself as a passive index investor, you may find the exercise of calculating your portfolio turnover illuminating.
Yeh, it is kind of like a vegetarian figuring out how many cheeseburgers they eat over the year :-). Interesting self reflection.
ReplyDelete@Alan: Unfortunately, continuing your analogy, my cheeseburger count for the year was 13% of 365, or about 48.
DeleteIt's not religion man. Don't sweat the small stuff.
ReplyDelete@Anonymous: I agree that it's not religion; it's money. There are various estimates of behavioural drag on portfolios, but if we go with a low estimate of 2% per year, that's 40% over 25 years, or $400,000 lost out of each million I would have had. It's not small stuff.
DeleteAnd if it was a religion, you could either buy or get absolution for your sins, in this area, there is no Absolution.
DeleteI've read that the largest ETFs have annual turn over rates of up to 800%. ETFs are definitely not a passive investment.
ReplyDelete@SST: I'm guessing you mean the turnover rate of the investors who trade these ETFs rather than the internal turnover of the ETFs themselves. I blame the investors for this rather than the ETFs. I own ETFs that I hold passively even though other investors trade them hyper-actively.
DeleteThat's a problem, though. ETFs are a hyper-active investment held within a passive strategy. This will effect your personal risk/return, to what degree, I have no idea.
Deletehttp://www.cfapubs.org/doi/pdf/10.2469/faj.v72.n1.5
@SST: Bogle's remarks indicate there is danger for investors who trade too much or who buy narrow ETFS. But, I'm not aware of any way that buy-and-holders of broad index ETFs are harmed by the existence of others who trade the same ETFs actively. If you know of such a harm, I'm interested.
Delete