This course is no longer eligible for CE credits. Go to cecorner.ca to find eligible courses.
The rules associated with Phase II of the Client Relationship Model (CRM II) were finalized earlier this year. Among other requirements, CRM II requires dealers and portfolio advisors to provide investment performance reports to their clients every year. Implementation of the performance reports will take place gradually starting in 2016.
The performance report will show the annualized percentage return for the client’s account. The return will include dividends and interest in addition to capital gains and losses, be calculated net of charges, and use a money-weighted method generally accepted in the securities industry. When the requirement is fully implemented, annualized returns will be provided for the 12 months, three years, five years and 10 years preceding the date at which the returns are calculated, and since the opening of the account.
A Mythical Investor
Let us fast-forward into the future. Dino, an investor, has just received the first set of performance reports from his portfolio advisors. He firmly believes in the virtues of advisor diversification and has allocated his money among several advisors. Like most investors, he does not feel that he owes them any particular duty of loyalty. He pays them and expects them to perform.
After looking at the various performance reports, Dino muses: “Portfolio Advisor A gave me a decent return of 12% last year, but Portfolio Advisor B only gave me a measly 2%. Quite clearly, B is not performing. I will move my money from B to A.”
Dino’s analysis can be faulted on several grounds. First, portfolio advisors should not be evaluated on the basis of a single year’s performance. In the short term, luck probably plays a greater role than skill. We need a lifetime’s worth of performance data in order to evaluate properly whether an advisor’s performance is due to luck or skill. This is because the analysis of advisor performance relies on statistical inference, which requires a lot of data.
Another issue is that the two portfolios may have different risks. Portfolio Advisor A may have produced better performance last year by taking more risk than B. If so, Dino should realize that Portfolio Advisor A is more likely to lose money in the future when the tide turns.
However, even if we allowed for these factors, Dino’s conclusion would still be unwarranted. Assume that we have many years of performance data on both portfolio advisors and that the two portfolios have the same risk. Dino would still be wrong to use the returns in the performance reports to judge the performance of his advisors. This is because of the way that CRM II requires the returns to be calculated.
What Does a Portfolio’s Performance Depend on?
Over the long term, the performance of a portfolio depends on the following factors:
- the performance of the market as a whole;
- the skill of the portfolio advisor; and
- the actions of the investor
Influence of the Market
Most portfolios do well when the market does well. A common saying is that a rising tide lifts all boats. On the other hand, when the market is experiencing a correction, the advisor would have to be either very skilful or very lucky to generate a positive return for the portfolio.
Influence of the Portfolio Advisor
The composition of an actively managed portfolio is usually different from that of the market portfolio. Portfolio advisors add value by using their analytical skills and knowledge of companies to identify stocks they believe will outperform the market. The additional return, after allowing for risk, is sometimes referred to as alpha, as opposed to beta, which is the return from simply tracking the market. Of course, it is possible for portfolio advisors to make mistakes and choose stocks that end up underperforming. In this case, the alpha is negative.
Influence of the Investor
Most investors probably think that, having entrusted their portfolio to an advisor, they are entitled to hold him or her fully responsible for its performance. This is incorrect. The client’s own actions also have a major impact on the portfolio’s performance, for better or worse.
Most of the time, it’s for worse. Investors contribute to their portfolio’s performance through their decisions to add more money to the portfolio or withdraw money from it. Studies have shown that investors’ actions reduce the annual return on their portfolio by approximately 1.5 percentage points. Think of it as a kind of negative alpha attributable to the investor. The negative alpha is truly enormous, especially when compounded over time. It can make the difference between a comfortable and a frugal retirement.
Investors’ alpha is negative because their decisions are often driven by emotion. Suppose they inject new money. This may be at a time when the market is overvalued and good investment opportunities are scarce. The portfolio advisor will then have to decide whether to buy investments for the portfolio at inflated prices or keep the money in cash—a choice between the devil and the deep blue sea.
Alternatively, suppose investors decide to withdraw money from their portfolios. Perhaps they have been spooked by a recent drop in the market. This is actually the worst time to withdraw money. In order to satisfy the client, the portfolio advisor will need to sell some investments at depressed prices.
The sad reality is that investors consistently choose the wrong timing. They tend to inject new money after a run-up in stock prices, when everything has become expensive, and withdraw from the market after a large drop in stock prices, when everything has become a bargain.
Two Families of Investment Returns
There are two main approaches to calculating investment returns:
- The first approach is to measure the impact of the market together with the skill of the portfolio advisor. The impact of the investor’s actions is excluded. Rates of return calculated using this approach are known as time-weighted returns.
- The second approach is to measure the impact of all the factors that affect a portfolio’s performance, including the actions of the investor. Rates of return in this family are known as money-weighted returns or dollar-weighted returns.
Each of these families of returns has its own use. The mistake of our mythical investor is that he used the wrong rate of return in assessing the performance of Portfolio Advisors A and B.
Performance Reports under CRM II
When developing CRM II, the concern of the securities regulators was to enable investors to figure out how they are progressing towards their investment objectives. Quite rightly in the view of this author, the return that CRM II requires dealers and portfolio advisors to provide to their clients is the money-weighted return.
The money-weighted return is a personal rate of return. It is most unlikely that two investors will have the same personal rate of return even if they happen to use the services of the same portfolio advisor. This is because the money-weighted return takes into account all the factors that affect the return on a portfolio, including the investor’s decisions. It is most unlikely that two investors will add money to or withdraw money from their respective portfolios at precisely the same time.
Dino was wrong to use the performance reports required by CRM II to evaluate the performance of his portfolio advisors because the reported returns are money-weighted returns—they include the impact of Dino’s decisions, for which the advisors cannot be held responsible. To evaluate the performance of his advisors, Dino should use a time-weighted return, which specifically excludes the impact of investor decisions.
More on the Mythical Investor
Like all thoughtful investors, Dino has a personal financial plan. His advisor interviewed him extensively—they discussed investment objectives, investment horizons, financial circumstances, risk tolerance and other subjects. In the end, the advisor used an optimizer to devise an asset allocation. For example:
Canadian stocks 20%
US stocks 20%
International stocks 20%
Using expected returns for each asset class, the advisor calculated that the portfolio would generate an average annual return of 6%. Under certain assumptions, which are beyond the scope of this course, this return would be sufficient to enable Dino to accumulate enough money by the time of his retirement to live a comfortable life. It therefore represents his target return.
Time-weighted returns may be used to judge the performance of the portfolio advisor. They may also be used to evaluate the performance of a portfolio or an investment fund. Time-weighted returns are appropriate for this purpose because they specifically exclude the impact of the investor’s actions.
Here are some situations in which time-weighted returns are encountered:
- Publicly distributed investment funds disclose time-weighted returns. Securities legislation specifically requires them to use the time-weighted method when calculating returns.
- Portfolio advisors usually adhere to Global Investment Performance Standards (GIPS), which require the use of time-weighted returns when reporting to clients and prospects. GIPS are a set of standards issued by the CFA Institute to guide portfolio advisors on how to calculate and report their investment results to clients and prospective clients. The objective is to enable investors to rely on uniformly prepared information when comparing and evaluating portfolio advisors. Note that this objective is very different from that of CRM II, which is to enable investors to figure out how they are progressing towards their investment objectives.
Time-weighted returns are only relevant when the portfolio advisor has discretionary authority to take investment decisions; for example, what stock to buy or sell, when and at what price. If the advisor only has authority to make recommendations the investor is free to accept or reject, the time-weighted return cannot be said to represent the advisor’s performance in any meaningful way.
Portfolio advisors normally have discretionary authority when managing their clients’ portfolio. Certain investment dealers that have implemented appropriate regulatory controls are also allowed to accept discretionary mandates. All other dealers are only allowed to make recommendations, and their clients have the final say.
Evaluating Portfolio Advisors
Time-weighted returns only constitute a starting point when evaluating portfolio advisors. All investments have three dimensions:
Time-weighted returns only provide information on the first dimension. For a proper evaluation of advisor performance, we need to recognize differences in risk. For example, if the portfolio is riskier than the market as a whole, it would not be appropriate to compare the advisor’s return directly with that of a market index such as the S&P/TSX Composite. Again, when comparing the performance of two different advisors, we need to allow for differences in the risk of the respective portfolios.
Liquidity is another factor to consider. A portfolio concentrated in illiquid investments, such as small caps, is significantly different from one invested in liquid investments. Liquidity is important because, the more liquid an investment, the easier it is to undo a past decision.
Suppose we accumulate a large position in a stock and belatedly recognize that we have made a mistake. If the stock is liquid, we will be able to sell it and get out of the position before too much harm is done. However, if it is illiquid, we may be stuck with our position and be forced to watch helplessly as the stock price drops inexorably towards zero. The value of liquidity is most readily apparent when the markets are in a state of turmoil and bids disappear for all but the most liquid stocks.
The Use of Composite Benchmarks
We can judge the performance of an advisor by comparing its time-weighted return with that of peers. Similarly, we can judge the performance of an investment fund by comparing its time-weighted return with that of other funds. We should be careful to compare like with like (similar risk, similar liquidity) and to look at performance data over a reasonably long period of time.
Another way to evaluate the performance of an advisor is to compare its time-weighted return with a composite benchmark. We can build the benchmark using the asset allocation in the investor’s financial plan.
We first need an index to represent each asset class. The indexes should reflect the markets in which the investor intends to invest. Otherwise, the benchmark would not have the same risk and liquidity as the portfolio and comparison would not be possible. For example, the S&P/TSX Composite would be a suitable benchmark for a broad-based portfolio of Canadian stocks but the S&P/TSX 60 would be more appropriate if the investor wanted to invest only in large caps.
For U.S. large caps, the S&P 500 would be suitable and, for international large caps, the MSCI EAFE. In the case of a broad-based portfolio of Canadian bonds, the DEX Universe Bond Index would be suitable. For cash, we could use the yield on treasury bills.
All we need to do is take the return on each index for the past year, calculate a weighted average return using the weights in the investor’s asset allocation, and voilà— we have a personalised composite benchmark which we can use to evaluate the advisor’s return. All index returns are time-weighted and are therefore comparable with the advisor’s return, which is also time-weighted.
Dino’s concern is whether his portfolio advisors are doing at least as well as permitted by current market conditions, as reflected by the composite benchmark. Unless the portfolio advisors do at least as well as the benchmark over time, Dino is unlikely to achieve his target return, which is simply a weighted average of the market’s historical annualized return. However, as we have already emphasized, Dino should not rush to any conclusion on the basis of a single year’s results. Even a good portfolio advisor will underperform the benchmark from time to time.
CRM II does not require advisors to provide time-weighted returns to their clients, but there is nothing to prevent them from doing so. In fact, most of them already do.
In the same way as for advisors, you can judge the performance of an investment fund against a benchmark that reflects its mandate.
What returns should Dino compare with his target return to determine if he is on track to meet his retirement goals? Time-weighted returns are inadequate for this purpose. They provide only an incomplete picture of the portfolio’s performance because they leave out the impact of the investor’s decisions—how he affected the portfolio’s performance through the amount and timing of the money he injected into or withdrew from the portfolio from time to time.
The Target Return as a Benchmark
For the purpose of evaluating their progress towards their investment goals, investors need to compare their target returns with the money-weighted return on their portfolios. Money-weighted returns take into account all factors relevant to the specific investor for whom they are being calculated.
This comparison should not be done by reference to a single year’s return. After all, the target return constitutes a long-term benchmark. Dino should relate the target return to the actual money-weighted return over a period of many years.
Using Money-Weighted Returns
Evaluating the long-term performance of the portfolio is only the starting point. Suppose Dino comes to the conclusion that, in light of his portfolio’s long-term performance, he is unlikely to achieve the target return necessary for his retirement dreams to come true. There is no need to despair, especially if Dino is still years away from retirement. He can implement a number of measures; for example:
- Save more
- Retire later
- Accept a riskier portfolio, though he should think very carefully first
- Accept a lower standard of living after retirement
How Are Investment Returns Calculated?
It remains to explain how time-weighted and money-weighted returns are calculated. The actual mechanics of the calculations are of secondary importance since they are usually handled by computers. However, we need to work through the mechanics for a full understanding of what each family of returns means, what it can properly be used for, and why the results are different.
A Simple Example
At the beginning of the year, Dino hires a new portfolio advisor, Fred, and funds the portfolio with $100,000. Fred invests the money in a mix of shares, bonds and money market instruments. From time to time, he replaces a share or bond with another that he believes to have more potential. Throughout the year, the portfolio receives dividends and interest, which are invested in more shares and bonds or simply left in cash. During the year, Dino does not add any new money to the portfolio or withdraw any money from it.
At the end of the year, the market value of the portfolio is $110,000, so the profit for the year is $10,000. The portfolio’s return simply expresses the profit as a percentage of the initial investment. The return is 10% or $10,000 / $100,000.
We express the return as a percentage because this conveniently allows us to compare portfolios of different sizes. A profit of $10,000 may be reasonably satisfactory on a portfolio of $100,000 but much less so on one of $1 million. A 10% return has the same interpretation, irrespective of portfolio size.
The profit of $10,000 consists of a mix of dividends, interest, net realized capital gains and net unrealized capital gains. The sources of profit are all reflected in the ending portfolio value of $110,000. We do not take them explicitly into account when calculating the return because this would be tantamount to counting the same thing twice.
It is important to note that Dino did not add new money to his portfolio during the year or withdraw any money from it. This means that the 10% return for the year is both a time-weighted return and a money-weighted return.
Calculating Time-weighted Returns
Let us now complicate matters slightly. Assume the same facts as previously. In addition, assume that, on June 1, Dino injected an additional $5,000 into the portfolio. At the close of business on May 31, the market value of the portfolio was $95,000. We want to calculate the time-weighted return for the year. This will enable us to evaluate the portfolio advisor’s performance by comparing it with that of its peers and with a benchmark.
Time-weighted returns are designed to remove the impact of the investor’s actions. We therefore need to find a way to calculate the portfolio’s return as if the $5,000 injection never took place.
First, we calculate the return for the period up to May 31, just before the injection. The loss for the period is $5,000 or ($95,000 less $100,000), which is equivalent to a return of -5% (-$5,000 / $100,000) for the five-month period.
Injecting $5,000 on June 1 makes Dino’s behaviour unlike that of most investors. He replenished his portfolio after losing money, which certainly required some courage. We add the $5,000 injection to the market value of the portfolio on May 31. This gives $100,000 ($95,000 plus $5,000) – the starting value for the second period, which begins on June 1 and ends on December 31. The profit for the period is $10,000 or ($110,000 less $100,000), which translates into a return of 10% for the seven-month period.
The final step is to calculate the time-weighted return for the full year. We do this by linking the returns for the two periods of five months and seven months, respectively. Start the year with a notional portfolio of $1. The portfolio earned a negative return of 5% for the first five months. At the end of the period, its market value was therefore $0.95 ($1 × (1 – 0.05)). In the second period of seven months, the portfolio’s return was 10%. At the end of the year, its market value was therefore $1.045 or ($0.95 ×1.10). The profit for the year is $0.045 or ($1.045 less $1), which is equivalent to a return of 4.5% on the initial portfolio of $1. This is the time-weighted return for the year.
We spent some time on the mechanics of the calculation because it sheds light on the nature of the time-weighted return. The time-weighted method tracks the value over time of a fixed investment of $1. (Any fixed number will do but $1 is convenient enough.) Injections and withdrawals of money have no impact on the calculation because we are tracking the progress of a fixed $1.
One drawback of the time-weighted method is that we must value the portfolio every time there is an injection or a withdrawal. We first calculate the return for each period preceding an injection or withdrawal, and then obtain the return for the full year by linking the periodic returns. The method we use to link the returns is known as geometric linking. Geometric linking effectively weighs each return with the length of the period of time over which it is earned. This is why this method is known as time-weighted.
In our example, the weights are five months for the -5% return and seven months for the 10% return, allowing for compounding. Note that there is no need to adjust for the fact that the two periods are of unequal length. If we made an adjustment, we would be implicitly assuming that returns are a linear function of time, which is clearly not the case. A 1% return for one month cannot be extrapolated into a 2% return for two months or a 12% return for 12 months.
If an investor makes multiple injections and withdrawals during the year, it will be necessary to value the portfolio a corresponding number of times. Perhaps the portfolio advisor should have a candid conversation with the client and inform him that an investment portfolio is not meant to be used as a chequing account.
Calculating Money-Weighted Returns
Let us now turn to money-weighted returns. These differ from time-weighted returns in that they do take account of the impact of the investor’s decisions in the form of injections and withdrawals.
The money-weighted return owes its name to that fact that, by recognizing all external flows of money into and out of the portfolio, it weighs periodic returns according to the assets under management during the period. Injections and withdrawals usually constitute the most important external cash flows.
The Treatment of Fees
If the portfolio advisor’s fees are paid out of the portfolio’s assets, the payment is reflected in a lower portfolio value at the end of the period. We should not include the fees explicitly in the calculations since this would be tantamount to counting them twice. However, if the investor pays the advisor’s fees directly, say, by means of a cheque drawn on his chequing account, the payment should be treated as an external cash outflow. Think of the direct payment of fees as an injection into the portfolio combined with a simultaneous payment of the fees out of the portfolio’s assets.
Using the XIRR Function
We calculate the money-weighted return by finding the internal rate of return (IRR) associated with the portfolio’s cash flows. The starting value of the portfolio and injections are treated as negative cash flows. The ending value and withdrawals are treated as positive cash flows.
It is easy to calculate the internal rate of return by using the XIRR function in Microsoft Excel. In order to use the XIRR function, you first need to install the Analysis Toolpak add-in. The calculation is illustrated in the screenshot shown as Table 1.
The XIRR function requires us to specify:
- Each cash flow; the cash flows are shown in Column F.
- The date of each cash flow; the dates are shown in Column A.
- A guess for the value of the IRR – 10% is usually a convenient guess.
The XIRR formula is in Cell F10. It reads:
F6:F8 reference the cells where the cash flows are entered.
A6:A8 reference the cells where the corresponding dates are entered.
0.1 or 10% is our guess for the IRR.
The formula returns the value 4.85%, which is the money-weighted return. This compares with the time-weighted return of 4.50%, which we computed earlier.
Why Do the Two Methods Give Different Results?
In this particular example, the money-weighted method gives a better return (4.85%) than the time-weighted return (4.50%). Why is this?
After the injection, the portfolio advisor obviously has more money to manage. This happens to coincide with an improvement in returns, from -5% for the period before the injection to 10% after the injection. The money-weighted method weighs periodic returns according to the money under management and therefore assigns a greater weight to the 10% return than the time-weighted method, which tracks a fixed $1. The result is to boost the return for the full year.
The decision to inject new money at a precise time was taken by the investor. He reaped the benefit for being a contrarian in the face of adversity. In the end, his personal rate of return was 4.85%, whereas he would have earned only 4.50% if he had taken no action. The additional return may be thought of as positive alpha contributed by the investor. Of course, this is not to suggest that contrarian investing always pays off. The point is to illustrate how the investor’s actions can affect a portfolio’s return, for better or for worse.
For his part, the portfolio advisor deserves no accolades for having more money under management precisely at a time when returns are good. The additional money is the result of a decision by the investor. The advisor’s performance is properly measured by the time-weighted return of 4.50%, which removes the impact of the investor’s decision.
For the first time in the history of securities legislation, CRM II will make it mandatory for portfolio advisors and dealers to provide personal rates of return to their clients. If used properly in conjunction with complementary information, personal rates of return have the potential to be extremely useful to investors. The information will enable them to take corrective action as required while there is still time. The challenge will be in educating investors to use the information properly.
When advisors and dealers start to release performance reports under CRM II, there is a very real risk that investors will misinterpret the information. Receiving correct information and using it for the wrong purpose is possibly worse than receiving no information at all. As an advisor, you should make sure that your clients understand the uses and limitations of the performance report and of performance information generally:
- Investors should use the money-weighted returns in the performance reports mandated by CRM II to evaluate their progress towards their investment objectives. A suitable benchmark for this purpose would be the target return in their personal financial plan. If the actual return trails the target return over a reasonably long period of time, this may indicate a need for corrective action on the part of the investor.
- They should use time-weighted returns provided by portfolio advisors and investment funds, allowing for differences in risk and liquidity, to decide which advisors or funds to use for their portfolios. Advisors can be compared with their peers and funds with other funds in the same category. Advisors can also be judged against a composite benchmark based on the investor’s target asset allocation. Funds can be judged against a benchmark that reflects their mandates. If an advisor or a fund underperforms over a reasonably long period of time, this may indicate a need for replacement.
- By comparing their personal (money-weighted) return with the advisor’s or fund’s (time-weighted) return, investors can ascertain how their own decisions are contributing to their portfolio’s overall performance. If the money-weighted return trails the time-weighted return, hopefully this will motivate them to adopt better investing practices.
André Fok Kam, CPA, CA, MBA is a consultant to the securities industry. He has advised regulators, fund managers, advisors and dealers. He is the author of From Conflict to Trust: How Mutual Funds Manage Conflicts of Interest (Toronto: Carswell, 2009) and several industry courses, including “Understanding Investment Returns,” offered by the Smarten Up Institute.