An important precept to bear in mind when designing an investment portfolio is, to quote Jack Bogle, that costs matter. Most mutual funds sold in Canada have management fees - "Management Expense Ratios", or MERs - of 1-3%. The funds' trading costs, bid-ask spreads on the entities bought and sold, and taxes distributed to unit holders as a result of stock sales will all add to this cost burden. As was discussed in the section on risk control, it is likely that returns going forward will only be around 4-5%, after inflation. In this environment, investors can not afford high mutual fund management fees. Instead, they must opt for low-cost alternatives. These include: low-cost index funds; Exchange-Traded Funds (ETFs), which are baskets of stocks that can be bought and sold as if they are single stocks and which have very low MERs; and direct ownership of individual stocks, bonds, or trusts. Although some mutual funds may be necessary in areas where there is no low-cost ETF or index fund, a well-diversified, cost-efficient portfolio can be built with an effective MER of as low as a few tenths of a percent. The cost savings amount to thousands of dollars a year on a moderately-sized portfolio.
Active Management and Costs
In 1991 Nobel-prize-winning economist William Sharpe published a paper entitled "The Arithmetic of Active Management". Using simple assumptions and elementary arithmetic, the paper showed that, if a securities market has a low-cost index fund available, active managers as a group will underperform indexers by the cost differential between active management and indexing. This analysis does not preclude outperformance by individual managers, only outperformance by the set of all active managers. There is no Lake Wobegon in which everybody is above average. Sharpe's analysis does not depend on market efficiency, but only on the cost differential between active and passive management.
The analysis does not apply to all securities markets because not all markets have index funds available. Individual housing and private business holdings are an important exception. However, Sharpe's arithmetic is applicable to essentially the entire U.S. stock and bond markets; most of the Canadian stock market except for small- and micro-capitalization stocks and most income trusts; the entire Canadian bond market; and large-capitalization stocks in the stock markets of most of the developed world. It therefore applies to a significant part of most investor's portfolios, and to all of some investor's portfolios. It follows that investors should only utilize active management in markets in which they are sure they, or the managers they choose, have an edge.
One of the consequences of Sharpe's arithmetic is that outperformers must succeed at the expense of underperformers. To quote Sir John Templeton again, "It's a contest". If you don't have an edge (or aren't sure that your manager does), the most cost-effective decision is to use passive, not active, management.
Tax-Efficient Asset Allocation
Investors must also be concerned about the effect of taxes on their portfolios, since the "tax drag" can amount to 1-2% a year if the portfolio is not properly designed. The following points should be kept in mind for a tax-efficient portfolio:
a. If the US investment is held through an American broker, form W-8BEN must be filed with the brokerage to avoid a 30% tax withholding rate. If a Canadian broker is used, form W-8BEN is often requested for new accounts.
Income-producing investments should be placed within a Registered Retirement Savings Plan (RRSP) during asset accumulation. Bonds should be placed in an RRSP first. If there is room left, high-income equities should then follow. Low-income equities can be left outside the RRSP with minimal tax consequences. Cash can be divided between RRSP and non-registered portfolios.
Because of the tax laws, a major factor in establishing a tax-efficient asset mix for an income-seeking investor is the relative size of the registered and non-registered portfolios. With a large non-registered portfolio, the income-seeking investor may hold a significant portion of the portfolio in preferred shares (which should always be in the non-registered portfolio) because insufficient RRSP room is available to shelter bonds. On the other hand, if the non-registered portfolio is quite small, the investor may hold no preferred shares.
Tax-efficient asset allocation of Canadian assets should be in accordance with the following table:
Tax-Efficient Asset Allocation
Note: The use of the FPX Balanced Index in these examples is arbitrary, and for illustrative purposes only. Other sample portfolios with greater diversification are shown in the section on portfolio construction.
FPX Balanced - $100,000 Total Investment
Registered: Non-Registered 75:25
Registered: Non-Registered 50:50
Registered: Non-Registered 25:75
Tax Loss Harvesting
Tax-loss harvesting is a method of utilizing tax losses to minimize - or defer - capital gains taxes in the non-registered investment account. It can be useful both during asset accumulation and during withdrawal phases.
In order to harvest tax losses, a security in the non-registered account is sold at a time when its price is significantly below its adjusted cost base [i.e. the purchase price plus any commissions or reinvested dividends]. The security is then replaced in the non-registered account by a similar, but not identical, security. After thirty days, the original security can be repurchased, if desired. The previously-available strategy of immediately repurchasing the sold security in an RRSP no longer results in an allowable loss.
CRA may disallow the loss if one index fund is replaced by an identical index fund from a different vendor. Replacing the fund by one tracking a similar, but different, index avoids the problem. The harvested loss can be used to reduce capital gains taxes in the current year. If no capital gains taxes are due in the current year, the loss can be carried backwards three years or forwards indefinitely.
A similar approach can also be used with individual securities, such as replacing one bank stock by the stock of a similar bank at a time when all bank stocks are under pressure, or by replacing one oil-and-gas trust by a similar one when oil prices are low.
Note that these approaches defer tax, but do not avoid it. They are appropriate when the tax is deferred until several years in the future or when the technique moves taxable capital gains into a lower tax bracket.
Further examples of this approach will be given in the section on Withdrawal Strategy.
In order to determine which of several alternate approaches is the most cost effective, it is useful to calculate the payback time associated with a certain move. This entails calculating both the initial cost and then the cost savings per year. The payback timea is how long it would take for projected cost savings to recover the initial expenses.
Similar calculations can be made to determine the minimum amount to invest in a certain approach for a given payback time.