For all the talk on investment performance detriments like management fees and transaction costs, turnover takes the most lethal bite out of a taxable portfolio

Fall, 1996 Canadian Investment Review - It only takes a cursory glance across the covers of investment management journals to note an overwhelming imbalance of research focused on hunting down the factors that positively affect portfolio performance, allocation decisions and security selection to name just two. Yet firmly planted on the opposite side of the spectrum are the detriments to performance, such as transaction and management fees, that seldom receive attention, particularly in the context of a taxable portfolio. And even less concern is given to the rarely discussed yet most impor-tant factor eroding performance-portfolio turnover.

Blame it perhaps on the fact that investment managers are measured and rewarded based on pre-tax performance. And while management fees and other expenses charged to a portfolio clearly affect performance, they contrast with turnover in being more easily identifiable and quantified. Turnover's impact on after-tax performance, however, is far more profound.

While a number of US research studies have quantified and confirmed the importance of turnover in a taxable portfolio, no such studies have encompassed the uniquely Canadian personal tax structure. Consider Figure 1, which demonstrates the after-tax effect of turnover on an initial $1-million investment in a Canadian equity portfolio invested at 12.4% for 20 years. No management expenses are charged to the portfolio and all dividends and realized capital gains are reinvested on a post-tax basis. As the horned-shaped graph shows, when turnover increases, the market value of the portfolio decreases dramatically. Calculating performance from dollar figures, a turnover increase to 8.5% imposes a performance penalty of 1.2%, whereas turnover rates of 35% and 80% reduce performance by 2.5% and 3.0%, respectively.

Figure 1

Indexing in the United States has been lauded by many as the low-turnover alternative to active portfolio management styles. While comprising approximately 35% of the US institutional marketplace, indexing is just now gaining acceptance in Canada, but active management is still regarded as the key to value-added for most institutional investors.

Therefore, we set out to determine whether the net returns of passively managed indexed equity pooled funds, of both Canadian and US securities, would exceed those of an actively managed segregated portfolio and of a no-load mutual fund over the long-term from the perspective of a taxable Can-adian investor. Using the TSE 300 Total Return Index and the S&P 500 Total Return Index as proxies for the Canadian and US equity markets, the results reveal that, once the turnover level associated with various investment management styles is taken into account, the indexed fund generated superior returns. In an era where every basis point counts, investors would do well to rethink the investment management styles they employ with an eye on the accompanying trading frequency.

Making Some Assumptions
In order to estimate returns for each of the three investment vehicles over a projected 20-year period, market value estimates for a portfolio with an initial $1-million investment were made under the following assumptions:

1. Current tax rates and laws, 20-year average historical rates of return and existing investment management fee structures exist over the projected period. As well, an investor is subject to regular Canadian tax treatment and laws, with no other sources of tax credits or capital gains and losses.

2. The expected rates of total and price returns and dividend yield for the TSE 300 Index and the S&P 500 Index over the next 20 years are equal to those of the 20 years ending December 31, 1995. That is, the total return for the TSE 300 Index over the projected period is 12.4%, with the price return accounting for 8.3% of that return and the yield equal to 4.1%. For the S&P 500, the total return is 14.6%, with a price return component of 10.1% and a yield of 4.5%.

3. Turnover of the portfolio is defined as the market value of securities sold out of the portfolio as a percentage of the portfolio's total market value. The turnover rates for the indexed TSE 300 and the S&P 500 funds are 8.2% and 8.3%, respectively. Turnover for the segregated portfolio is 35% and the mutual fund was set at 80%.

4. The management expense ratios used in the model, which encompass all expenses payable by the funds and/or the investor, were set according to our best understanding of current fee levels. In both markets, the indexed fund was charged a fee of 0.5% and the segregated portfolio incurred 1.0% in fees. The Canadian equity mutual fund had a management expense ratio of 2.04% and the US equity mutual fund was charged 2.26%.

5. It was assumed that the gross returns of an indexed fund would match the index returns exactly. That means the portfolio was managed in such a way that its returns had no tracking variance relative to the index-which in real life can be accomplished. Also, all active portfolios maintained small cash balances and did not attempt to "market time" by moving in and out of equities. Therefore, segregated portfolios and mutual funds were assumed to possess returns similar to the index.

Within this framework, net returns were defined as the return realized by the investor after all management fees, fund expenses and tax liabilities were satisfied. In each case, two variations for each investment vehicle were analyzed. The first involved investments that were not liquidated after 20 years and, therefore, no capital gains tax liability existed. The second involved terminal liquidation, whereby the portfolio is sold into the market at the end of the 20 years at the predicted market price and capital gains taxes on the gains are levied on the difference between the market and the book values in the final year. Commissions are expected to reduce this amount further.

After converting these figures into compound annualized rates of return (see Appendix, for details), the results point to a significant difference in returns (Table 1). In all cases, the indexed fund returned the greatest amount, followed by the segregated portfolio and then the mutual fund. The same picture emerges even after adjusting various variables.

Table 1

Scenario 1: The base case
The first scenario, analyzing the investment vehicles without liquidation after 20 years, calculated the pre-tax performance required for the higher turnover options to achieve the same net (after taxes and fees) returns earned by the index fund. Since it is difficult to change dividend yield; it was assumed that this outperformance would only be possible in the price appreciation component of total return.

The results in Table 2 indicate that for Canadian equity, a segregated portfolio must outperform an index benchmark by 2.3% and a mutual fund requires 4.3% outperformance if an investor is to realize the equivalent net returns of an indexed fund. The required outperformance for US equity is greater: 2.6% for a segregated portfolio and 5.1% for a mutual fund.

Table 2

A huge spread
For a mutual fund and a segregated portfolio to match the net returns earned by an indexed strategy, they must outperform the benchmark by a range of about 200 to 500 basis points in gross returns over a 20-year projected period (without liquidation).

	TSE 300 Index		S&P 500 Index
	Total Return	Required Outperformance	Total Return	Required Outperformance
Benchmark Index Fund	12.44%	-	14.61%	-
Equivalent Segregated Portfolio	14.70%	2.26%	17.25%	2.64%
Equivalent Mutual Fund	16.77%	4.33%	19.67%	5.06%

Scenario 2: A fee squeeze
The effect of fees on final returns was also determined by slashing fees for a segregated portfolio and a mutual fund to the 0.5% rate of an index fund. Setting fees for all investment options at an equal level does help, as the degree of outperformance required by a segregated portfolio and a mutual fund is reduced versus the base case (Table 3).

Table 3

Still, in order to match the final returns of an indexed strategy in this revised scenario, the required outperformance by a Canadian equity segregated portfolio declines slightly from 2.3% to 1.9%, and from 4.3% to 2.9% for a mutual fund. For US equity, similar reductions occur. The segregated account must now generate an additional 2.3% in returns relative to the benchmark versus 2.6% in the base case, while the mutual fund must outshoot the index by 3.5% versus 4.9%.

It is important to note that the reduction in fees does not result in a direct one-for-one in required returns over the base case, since investment management fees are tax deductible and, therefore, lower fees reduce the tax credit that would have otherwise applied given a higher fee level.

Scenario 3: Targetting turnover
Turnover levels can vary from time to time or among various funds. Therefore, to test the sensitivity of the results to the level of portfolio turnover, the turnover levels of the mutual fund and segregated portfolio were reduced by half of their original amounts, to 40% and 17.5%, respectively. The indexed fund remained at the standard turnover level.

Once turnover is cut, the performance gap between the benchmark index and a mutual fund and a segregated portfolio does narrow (Table 4). However, the mutual fund is still required to outperform the index by 3.4% in Canadian equity and by 3.9% in US equity to match the indexed strategy, while the segregated portfolio must outperform by 1.3% for Canadian equity and by 1.4% in US equity. That the segregated portfolio experienced a more dramatic improvement in returns than the mutual fund suggests, as seen in Figure 1, that a reduction in portfolio turnover has less of an impact on after-tax returns the higher the initial turnover level.

Table 4

Sooner Better Than Later
The fee structures and taxable consequences of turnover in segregated portfolios and mutual funds make an indexed investment management style an attractive option for a taxable portfolio. For both Canadian and US equity, an indexed fund outperforms a segregated portfolio and mutual fund by a margin ranging from about 200 to 300 basis points on an after-tax basis.

During any comparison of an indexed fund to an actively managed mutual fund or segregated portfolio, the performance of the index relative to manager universes should not be ignored. While all active managers strive to beat the index, the performance margin required to equal net after-tax returns of an index fund is extremely difficult to earn, given the track record of institutional equity managers. Although the median Canadian equity manager outperformed the TSE 300 Index by 0.4% over the past decade, and first-quartile and fifth-percentile managers managed to outstrip the index by 1.4% and 3.6%, respectively, by year-end 1995, the tax and fee consequences would easily wipe out these margins. That means that even if an investor was able to consistently select the fifth-percentile manager of a mutual fund, this manager would have still underperformed an indexed strategy on a net after-tax basis.

Top managers south of the border have fared even worse. The first-quartile and the fifth-percentile foreign equity managers outperformed the S&P 500 Index by a mere 0.2% and 2.0%, respectively. This smaller outperformance suggests that the returns reported by an actively managed US equity portfolio are even lower on a net after-tax basis. Consequently, it would appear that adopting an indexed strategy for US equity investments makes even more sense than for Canadian equity.

Portfolio turnover is the largest detriment to after-tax returns, yet may be the hardest portfolio variable for active managers to control. Achieving excess returns with the high turnover associated with actively managed portfolios may be even more difficult when the transaction costs of trading are considered.

As the shape of the curve in Figure 1 illustrates, the greatest performance hit occurs in the early turnover stages. As turnover becomes progressively larger, its impact on after-tax returns declines. In other words, most of the damage occurs up front, making prospective turnover figures that much more important in the investment manager hiring process. Selecting an investment strategy which maximizes after-tax returns through low turnover is, therefore, critical in determining ultimate performance.

Reprinted with permission.

ENDNOTES

1. See for example "Is Your Alpha Big Enough to Cover its Taxes?: The Active Management Dichotomy," by R. Jeffrey and R. D. Arnott, The Journal of Portfolio Management, Spring 1993, pp. 15-25.; and "Ranking Mutual Funds on an After-Tax Basis," by J.M. Dickson and J. B. Shoven, 1993, National Bureau of Economic Research Working Paper No. 4393.

2. Back-testing could not be conducted because data for these funds does not exist for the past 20 years.

3. The yield figures calculated from this method are higher than expected since they include both stock and cash dividends declared and paid. As the tax treatments for both types of dividend are the same, this does not affect the study's predicted returns.

4. According to data provided by the TSE 300 and the S&P 500, the average turnover amounts for the indexes are 1.95% (in the last four years) and 2.91% (in the last seven years), respectively. The turnover for the Canadian equity indexed pooled fund was set equivalent to the average turnover over the last four years for an institutional pooled fund, TD Asset Management's Emerald Canadian Equity Fund. The US indexed pooled fund turnover figure is based on an average of 22 US indexed mutual funds as reported in the S&P 1995 Directory (Standard & Poor's: New York, 1995).

   The turnover figures for the segregated portfolio of 35% was set equal to common industry practice. SEI Financial Services Ltd. reports that from 1989 to 1995, the average annual turnover of an institutional Canadian equity portfolio was 33.6% (purchases) and 30.4% (sales). As some of the portfolios measured by SEI hold units of pooled funds, and as SEI does not "look through" these units to the underlying portfolio's turnover, this number is likely to be significantly understated. Turnover figures for mutual funds were sourced from "How Mutual Funds Work," by A. J. Friedman and R. Wiess, New York Institute of Finance: Englewood Cliffs, 1993, pg. 40. Comparable research in the Canadian market is not readily available, although consultations with brokers and financial planners confirm 80% as a reasonable figure.

5. The fees for an indexed fund reflect prospective fees on a portfolio of $1 million charged by TD Asset Management. Fee levels for segregated portfolios are based on a survey conducted in 1992 by the asset manager. The mutual fund fee level is derived from data provided by the Financial Times Bell Charts in December 1995.

6. All manager universe return data has been provided by SEI Financial Services for periods ending December 31, 1995. Performance is calculated before fees, which tends to overstate an active manager's performance relative to an indexed fund. Also, no consideration has been given to survivorship bias in this study. According to a recent study, this may artificially increase the median manager's returns by as much as 0.5%. See "Survival of the Fittest," by B. Curwood, P. Halpern, S. Hadjiyannakis and K. Taylor, Canadian Investment Review, Winter 1995/96, pp. 9-12.

APPENDIX
The following formulas were applied for the indexed and segregated portfolios. In the case of a mutual fund, investment management fees directly reduced dividend income.

Market Value:
MV₂=MV₁ + Capital Appreciation1 + Dividend Return1 -Investment Management Fees₁ - Dividend Tax Liability₁ -Capital Gains Tax Liability₁ + Investment Management Fees Paid Tax Credit₁

Book Value:
BV₂=BV₁ + Realized Capital Appreciation₁ + Dividend Return₁ - Investment Management Fees₁ - Dividend Tax Liability₁ - Capital Gains Tax Liability₁ + Investment Management Fees Paid Tax Credit₁

Capital Appreciation:
CG₁=MV₁ x Price Return %

Dividend Return:
DR₁=MV₁ x Yield %*

Investment Management Fees:
IM₁=Investment Management Fees x Average(MV₁, (MV₁ + CG₁ + DR₁))

Capital Gains Tax Liability:
CT₁=RC₁ x Inclusion Factor (75%) x Tax Rate (50%)

Realized Capital Gains:
RC₁=((MV₁ + CG₁)-BV₁) x Turnover %

Canadian Equity Dividend Return Tax Liability:
DT₁=(MV₁ x Yield % x Dividend Gross Up (125%)) x (Tax Rate (50%) - Dividend Tax Credit (13.33%))

US Equity Dividend Return Tax Liability:
DT₁=(MV₁x Yield%) x Marginal Tax Rate

Investment Management Fees Tax Credit:
FC₁=Investment Management Fees₁ x Marginal Tax Rate

In the case of the mutual fund, investment management fees directly reduced dividend income. The following formulas were used:

Market Value:
MV₂=MV₁ + Capital Appreciation₁ + Dividend Return₁ - Dividend Tax Liability₁ - Capital Gains Tax Liability₁

Book Value:
BV₂=BV₁ + Realized Capital Appreciation₁ + Dividend Return₁ - Dividend Tax Liability₁ - Capital Gains Tax Liability₁

Canadian Equity Dividend Return Tax Liability:
DRT₁=((Dividend Income₁ - Investment Management Fees₁) x Dividend Gross-Up x (Marginal Tax Rate - Dividend Tax Credit))

US Equity Dividend Return Tax Liability:
DRT₁=(Dividend Income₁ - Investment Management Fees₁) x Marginal Tax Rate

Copyright © TD Bank
Click here to view The Fine Print
Click here to view Internet Security Information.