Choosing Mutual Funds: Are iUnits and iShares the Ultimate Solution?
There has long been debate over active and passive management – is there any value in active management? If so, how much return does active management add? If not, why do people buy actively managed mutual funds? Wouldn’t buying iUnits and iShares be the ultimate investing solution?
The Theory
Depending on which study you read (and there are a lot of them) you will get varying answers. If you want my opinion (please stop chuckling) I’d have to conclude there is no definitive answer to any of these questions. Luckily enough, for the average investor, the answer to these questions is not really that important. Let me explain why...
A couple of years ago, our firm was commissioned to analyze the universe of F-class mutual funds, iUnits and iShares (ETFs – Exchange Traded Funds) in order to produce portfolios with the best risk/return characteristics we could come up with. In my original hypothesis I assumed portfolios of all ETFs would produce the very best portfolios because of the ultra-low MERs associated with these securities. After weeks of analysis using proprietary software that mixed and matched the universe of Canadian mutual funds and ETFs together in millions of different combinations, we found the best portfolios were actually combinations of both managed mutual funds, and ETFs.
At first, we were a bit perplexed. How could managed mutual funds give a portfolio a better risk/return ratio than portfolios made up of only ETFs? As it turns out, there’s a very simple reason.
Morningstar released a report some months ago in which they reported managed mutual funds that consistently out-performed their benchmarks, also tended to have a lower R-squared value.
What’s R-squared? R-squared is a number between 0 and 100. The number defines how much of a mutual fund’s returns are due to market fluctuations. In other words, R-squared is a good measure of how much of a correlation there is between a mutual fund’s returns and the returns of the benchmark the fund is compared against.
Why did Morningstar discover managers with lower R-squared numbers tended to have better returns than managers with higher R-squared numbers? There are probably a number of reasons, but I would suggest the following: the higher a mutual fund’s R-squared is, the closer that mutual fund is to mirroring the index (otherwise know in the industry as ‘closet indexing’). Because of the management fees charged in actively managed mutual funds, the more a fund manager closet indexes, the harder it is for the manager to keep up with the index returns. For example, if a closet indexing fund manager’s mutual fund had an R-squared of 100, the fund would have a return equal to the index return minus the management fee.
If all investors had to choose from were ‘closet indexing’ mutual funds with 2.5% MERs, and ETFs with 0.8% or lower MERs, the choice would be very simple; buy the index and forget about managed funds. The problem with making this broad assumption is accounting for managed mutual funds with low R-squared. These funds sometimes differ so greatly from the indexes they are measured against, one could argue, they should be measured against a completely new index. But since there is no closer matching index to choose from, the best an investor can do is to invest directly in the low R-squared managed mutual fund if they want exposure to that management style.
So why were low R-squared, high MER managed mutual funds showing up in our analysis?
Very simply, the universe of management styles covered by all ETFs was smaller than the universe of management styles when managed mutual funds were also considered in the analysis. Some of the managed mutual funds were managed so differently from the index they are measured against, it was like adding a completely different index to the analysis. Bottom line: the managed mutual funds often diversified the overall portfolio risk enough to more than off-set their considerably higher MERs.
The Reality
OK, so now we know that in theory, the best portfolios are combinations of indexes and low R-squared managed mutual funds, but what happens when we apply theory to real life? Mutual funds can be purchased for free (minus the embedded MER of course) so if low R-squared mutual funds show up in a portfolio, we know all costs have already been considered. What about the index component of our portfolio? Are iUnits and iShares really as cost-effective an investment vehicle as we are led to believe, or would an investor be bettor off purchasing index mutual funds instead? The answer really depends on how much money an investor has to invest.
ETFs need to be purchased like a stock or bond with a commission cost on top of the ultra-low MER of the ETFs. This can make transparency of total costs for ETFs a little more cumbersome to figure out than for index mutual funds. Because of this extra cost, investors with smaller accounts may actually be better off purchasing an index fund with a higher MER instead of an ETF. Let’s look at a hypothetical example with numbers an investor should consider before selecting ETFs for the index portion of their portfolios:
Let’s assume we’ve created a mixed portfolio of three low R-squared mutual funds, and 3 indexes. Even if we buy the indexes through a discount broker, we can expect to pay around $40/transaction. If we assume we will rebalance our portfolio from time to time (at least once per year), we could reasonably expect to buy and sell ETFs in our portfolios at least six times per year. Therefore, we should assume we would spend at least $240/year in transactions for our ETFs.
We bought these ETFs to keep costs down right? Based on the MERs of comparable index funds (see table below), we really don’t want our yearly transaction costs to exceed 0.5% of our portfolio’s value invested in indexes per year. That means, with transaction costs of only $240/year, we would need to invest at least $48000 in ETFs to keep transaction costs to 0.5% or less. If we assume the ETFs only make up 50% of the total portfolio then the investor’s total portfolio size should be at least $84000 or more in value to make buying ETFs a good deal for the index portion of our portfolio. Any less money than that, and it might well be cheaper to buy index mutual funds instead.
Note: In this hypothetical example it’s important remember that if any of our transaction assumptions are too low, or if we purchase more than three ETFs we will need an even bigger portfolio to break even.
Now what about if a full-service advisor presents a fee-for-service portfolio to you either with all ETFs, or a combination of ETFs and managed mutual funds? As long as all of the mutual funds being purchased are F-Class mutual funds, an investor can safely pay up to 1% for equity funds and 0.5% for fixed income funds to the advisor without paying more than buying the regular retail version of the fund. Paying anything less than 1% for equity, and 0.5% for fixed income is generally a deal. Again, ETFs are different.
If the advisor is paying for all transaction costs of the ETFs out of the for-service fee, an investor again can safely pay up to around 0.5% of the accounts total value invested in indexes per year to the advisor and still be getting a slight advantage from the lower MER cost. If charged any more, the total cost of the ETFs will be approaching or even exceeding the cost of some index mutual funds (see table below). If transaction fees are charged separately from the for-service fee, then an investor will need to pay careful attention to their account size and the cost of transactions to ensure they can negotiate fees with their advisor low enough to avoid paying too much. If combined for-service fees and transaction costs cannot be negotiated down to under 0.5% per year for the index portion of a portfolio, an investor should consider index funds as a simpler, more transparent solution.
If total fees charged to the investor approach or exceed 0.5%, ETFs quickly lose their competitive cost structure advantage:
Canadian Indexes MER 1 Yr 3Yr
iUnit S&P/TSX capped 0.25 35.9% 24.4%
TD Canadian Index – I 0.88 31.1% 23.3%
RBC Canadian Index 0.78 31.0% 23.3%
National Bank Canadian Index 1.14 33.7% 23.0%
CIBC Canadian Index 0.96 31.0% 23.2%
Altamira Canadian Index 0.54 34.3% 23.6%
Canadian Bond Indexes MER 1 Yr 3 Yr
iUnit Broad Bond Index 0.30 4.4% 7.0%
CIBC Canadian Bond Index 0.96 3.8% 5.8%
RBC Canadian Bond Index 0.74 3.5% 5.3%
Scotia Canadian Bond Index 1.14 3.7% 5.7%
TD Canadian Bond Index – I 0.91 3.9% 5.7%
Conclusion
Although ETFs are very cost effective securities that should be considered for inclusion in an investment portfolio, it is a mistake to think ETFs alone, are the ‘ultimate solution’. Ignoring the vast universe of excellent managed mutual funds that can further diversify risk and increase overall portfolio performance clearly makes little sense. Compounded with the careful consideration to fees that must be paid to ensure account size is large enough to cover all commissions and service fees, ETFs are not always a good solution for every investor.
If an investor is planning to use a ‘buy and hold forever’ strategy, then purchasing ETFs through a discount or online broker is definitely a cheep, simple and effective way to get into the market. If an investor plans to rebalance their investment portfolio from time to time in order to lock in profits and dollar-cost-average into out of favour securities, or if an investor is dealing with a full service or fee-for-service advisor, much more careful research into all costs must be conducted by the investor. A combination of service and transaction fees can easily eat up all of the advantages provided by the low MERs of ETFs, leaving the investor with at best, the equivalent of index mutual funds, and at worst an overpriced, underperforming investment portfolio. However if investors do their research, and are careful to negotiate fees low enough, ETFs can make a very cost effective contribution to a well-diversified portfolio.
The Theory
Depending on which study you read (and there are a lot of them) you will get varying answers. If you want my opinion (please stop chuckling) I’d have to conclude there is no definitive answer to any of these questions. Luckily enough, for the average investor, the answer to these questions is not really that important. Let me explain why...
A couple of years ago, our firm was commissioned to analyze the universe of F-class mutual funds, iUnits and iShares (ETFs – Exchange Traded Funds) in order to produce portfolios with the best risk/return characteristics we could come up with. In my original hypothesis I assumed portfolios of all ETFs would produce the very best portfolios because of the ultra-low MERs associated with these securities. After weeks of analysis using proprietary software that mixed and matched the universe of Canadian mutual funds and ETFs together in millions of different combinations, we found the best portfolios were actually combinations of both managed mutual funds, and ETFs.
At first, we were a bit perplexed. How could managed mutual funds give a portfolio a better risk/return ratio than portfolios made up of only ETFs? As it turns out, there’s a very simple reason.
Morningstar released a report some months ago in which they reported managed mutual funds that consistently out-performed their benchmarks, also tended to have a lower R-squared value.
What’s R-squared? R-squared is a number between 0 and 100. The number defines how much of a mutual fund’s returns are due to market fluctuations. In other words, R-squared is a good measure of how much of a correlation there is between a mutual fund’s returns and the returns of the benchmark the fund is compared against.
Why did Morningstar discover managers with lower R-squared numbers tended to have better returns than managers with higher R-squared numbers? There are probably a number of reasons, but I would suggest the following: the higher a mutual fund’s R-squared is, the closer that mutual fund is to mirroring the index (otherwise know in the industry as ‘closet indexing’). Because of the management fees charged in actively managed mutual funds, the more a fund manager closet indexes, the harder it is for the manager to keep up with the index returns. For example, if a closet indexing fund manager’s mutual fund had an R-squared of 100, the fund would have a return equal to the index return minus the management fee.
If all investors had to choose from were ‘closet indexing’ mutual funds with 2.5% MERs, and ETFs with 0.8% or lower MERs, the choice would be very simple; buy the index and forget about managed funds. The problem with making this broad assumption is accounting for managed mutual funds with low R-squared. These funds sometimes differ so greatly from the indexes they are measured against, one could argue, they should be measured against a completely new index. But since there is no closer matching index to choose from, the best an investor can do is to invest directly in the low R-squared managed mutual fund if they want exposure to that management style.
So why were low R-squared, high MER managed mutual funds showing up in our analysis?
Very simply, the universe of management styles covered by all ETFs was smaller than the universe of management styles when managed mutual funds were also considered in the analysis. Some of the managed mutual funds were managed so differently from the index they are measured against, it was like adding a completely different index to the analysis. Bottom line: the managed mutual funds often diversified the overall portfolio risk enough to more than off-set their considerably higher MERs.
The Reality
OK, so now we know that in theory, the best portfolios are combinations of indexes and low R-squared managed mutual funds, but what happens when we apply theory to real life? Mutual funds can be purchased for free (minus the embedded MER of course) so if low R-squared mutual funds show up in a portfolio, we know all costs have already been considered. What about the index component of our portfolio? Are iUnits and iShares really as cost-effective an investment vehicle as we are led to believe, or would an investor be bettor off purchasing index mutual funds instead? The answer really depends on how much money an investor has to invest.
ETFs need to be purchased like a stock or bond with a commission cost on top of the ultra-low MER of the ETFs. This can make transparency of total costs for ETFs a little more cumbersome to figure out than for index mutual funds. Because of this extra cost, investors with smaller accounts may actually be better off purchasing an index fund with a higher MER instead of an ETF. Let’s look at a hypothetical example with numbers an investor should consider before selecting ETFs for the index portion of their portfolios:
Let’s assume we’ve created a mixed portfolio of three low R-squared mutual funds, and 3 indexes. Even if we buy the indexes through a discount broker, we can expect to pay around $40/transaction. If we assume we will rebalance our portfolio from time to time (at least once per year), we could reasonably expect to buy and sell ETFs in our portfolios at least six times per year. Therefore, we should assume we would spend at least $240/year in transactions for our ETFs.
We bought these ETFs to keep costs down right? Based on the MERs of comparable index funds (see table below), we really don’t want our yearly transaction costs to exceed 0.5% of our portfolio’s value invested in indexes per year. That means, with transaction costs of only $240/year, we would need to invest at least $48000 in ETFs to keep transaction costs to 0.5% or less. If we assume the ETFs only make up 50% of the total portfolio then the investor’s total portfolio size should be at least $84000 or more in value to make buying ETFs a good deal for the index portion of our portfolio. Any less money than that, and it might well be cheaper to buy index mutual funds instead.
Note: In this hypothetical example it’s important remember that if any of our transaction assumptions are too low, or if we purchase more than three ETFs we will need an even bigger portfolio to break even.
Now what about if a full-service advisor presents a fee-for-service portfolio to you either with all ETFs, or a combination of ETFs and managed mutual funds? As long as all of the mutual funds being purchased are F-Class mutual funds, an investor can safely pay up to 1% for equity funds and 0.5% for fixed income funds to the advisor without paying more than buying the regular retail version of the fund. Paying anything less than 1% for equity, and 0.5% for fixed income is generally a deal. Again, ETFs are different.
If the advisor is paying for all transaction costs of the ETFs out of the for-service fee, an investor again can safely pay up to around 0.5% of the accounts total value invested in indexes per year to the advisor and still be getting a slight advantage from the lower MER cost. If charged any more, the total cost of the ETFs will be approaching or even exceeding the cost of some index mutual funds (see table below). If transaction fees are charged separately from the for-service fee, then an investor will need to pay careful attention to their account size and the cost of transactions to ensure they can negotiate fees with their advisor low enough to avoid paying too much. If combined for-service fees and transaction costs cannot be negotiated down to under 0.5% per year for the index portion of a portfolio, an investor should consider index funds as a simpler, more transparent solution.
If total fees charged to the investor approach or exceed 0.5%, ETFs quickly lose their competitive cost structure advantage:
Canadian Indexes MER 1 Yr 3Yr
iUnit S&P/TSX capped 0.25 35.9% 24.4%
TD Canadian Index – I 0.88 31.1% 23.3%
RBC Canadian Index 0.78 31.0% 23.3%
National Bank Canadian Index 1.14 33.7% 23.0%
CIBC Canadian Index 0.96 31.0% 23.2%
Altamira Canadian Index 0.54 34.3% 23.6%
Canadian Bond Indexes MER 1 Yr 3 Yr
iUnit Broad Bond Index 0.30 4.4% 7.0%
CIBC Canadian Bond Index 0.96 3.8% 5.8%
RBC Canadian Bond Index 0.74 3.5% 5.3%
Scotia Canadian Bond Index 1.14 3.7% 5.7%
TD Canadian Bond Index – I 0.91 3.9% 5.7%
Conclusion
Although ETFs are very cost effective securities that should be considered for inclusion in an investment portfolio, it is a mistake to think ETFs alone, are the ‘ultimate solution’. Ignoring the vast universe of excellent managed mutual funds that can further diversify risk and increase overall portfolio performance clearly makes little sense. Compounded with the careful consideration to fees that must be paid to ensure account size is large enough to cover all commissions and service fees, ETFs are not always a good solution for every investor.
If an investor is planning to use a ‘buy and hold forever’ strategy, then purchasing ETFs through a discount or online broker is definitely a cheep, simple and effective way to get into the market. If an investor plans to rebalance their investment portfolio from time to time in order to lock in profits and dollar-cost-average into out of favour securities, or if an investor is dealing with a full service or fee-for-service advisor, much more careful research into all costs must be conducted by the investor. A combination of service and transaction fees can easily eat up all of the advantages provided by the low MERs of ETFs, leaving the investor with at best, the equivalent of index mutual funds, and at worst an overpriced, underperforming investment portfolio. However if investors do their research, and are careful to negotiate fees low enough, ETFs can make a very cost effective contribution to a well-diversified portfolio.
posted by Edward Iftody, BA, CIM at
9:37 AM
6 Comments:
FYI, I think the $40 is high.
I pay $27.95 at Action Direct but get tremendous access to free research on-line. The broker commission is tax deductible as you know.
Also, you can write covered calls on most ETF's for extra income/risk mngt. or even short them. Index doesn't have to mean passive investing.
As to taxes, Prod. Milevsky found that actively managed funds are far less tax- efficient than ETF's .( no problem if in an RRSP)
Average equity fund T/O about 100%!
Bond funds are totally a waste of time , MER's uncompetitive.
For asset allocation I'd think the iunits bond ETF's or even laddered GIC's would be hard to beat over time.
A real world problem is that fee-only advisors are conflicted and sometimes double dip i.e they don't buy F class for their clients because of the reduced or NIL trailer commission.
Notwithstanding my comments I think you've raised some useful food for thoght.
ken k
By
Anonymous, at February 27, 2006
[I] agree with you about the risk-diversification advantage of active funds. In fact, I’ve written on the benefits of a 50-50 passive/active partnership. I think there are psychological benefits here to investors that lead to better returns, or at least prevent bad decisions that lead to poor returns.
Specifically, a portfolio with a mix of conservative active funds and index funds might weather a down market better thanks to the active funds and it might offer enough in returns during an up market to prevent investors from getting fed up with a slow-performing active fund.
Best,
Rob Carrick
By
Anonymous, at February 27, 2006
Your discussion on "Choosing Mutual Funds..." falls far short of a rational - or even reasonable - analysis of what you purport to be doing. You seem simply to be "proving" a pre-adopted (self-serving?) position by using selected facts and faulty rationalizations. There is so much "missing" information and specious explanations in your article as to render your conclusions meaningless.
By
Anonymous, at October 09, 2006
Michael:
I don't actually care if you prefer investing your money in iShares. I think it's great you are actually researching your investments. However, I think I should respond to your post.
I don't make my living selling mutual funds or iShares. I lease software to financial advisors and dealers across Canada. The software is designed to help advisors explain risk to investors in a graphical format so investors can make informed decisions on how much risk they would like in their portfolios BEFORE buying anything. It just happens that in my line of business we obviously run into interesting information (sometimes by accident) when testing our software prior to release. I don't think its a duty, but we freely share our findings with investors who could benefit from our research.
You say my article is not rational or reasonable, I use faulty facts and there is so much missing information that my conclusion is meaningless. Unfortunately you didn't bother to mention which facts, but I assume from the tone of your comments you disagree strongly with all of them.
You should note the two other posters to this article are a respected financial journalist Rob Carrick and Ken Kivenko - famous for fighting for investors' rights. Both men are not fans of actively managed mutual funds, but both agree there are some points in the article they hadn't considered.
I'm sure you must have considered all relevant points to be so sure of yourself, so please feel free to forward your personal iShare portfolio to me, and I will send back a mixed portfolio that smashes it on a risk-return basis.
Edward Iftody
President
PureLogix Corp.
By
Edward Iftody, BA, CIM, at June 14, 2007
Edward:
Well, its 2009 and much has happened since your most recent post. Such as the incredible (to some!), poor performance of actively managed funds which were touted by many financial advisors to produce better investment performance in "down" markets. (That is, where there has been a reluctant admission by most active managers now that actively managed funds could not do as well as passively managed funds in "up" markets but should do better in a down market).
History has demonstrated conclusively that actively managed funds (some 85% of them) have also done worse in down markets.
Now, your 'offer' in your previous post clearly shows the non-sense of your information. Of course you (or anyone) can configure (or identify) an actively managed fund that has done better than the 'market'. Why don't you provide an actively managed fund that WILL do better in the future?
Your business seems to be based on selling "promo" material (masquerading as "analyasis") to active managers so they can convince their clients to pay unnecessary management fees for substandard performance. The standard of course is the market index.
Financial advisors know no more than the market - but as long as they can fool their clients in this regard, they will have a good future - but, unfortunately at the expense of their clients.
'
By
Michael, at July 13, 2009
Michael:
Same toothless and rude ranting as three years ago! Repeating the same arguements made by indexers for years, does not make them true.
Our analysis still stands, since obviously we did our analysis on both bull AND bear markets.
Clearly, from your post, you do not do any sort of analysis. That is sad and certainly costly you a lot of money over time.
Let me make my 'offer' more clear for you. Send me the portfolio YOU invested in over the last 3 years and I will compare it directly on a risk return basis to the portfolio I invested my own money in.
By
Edward Iftody, BA, CIM, at July 13, 2009
Post a Comment
<< Home