How to determine the proper time horizon to evaluate portfolio performance is always a subject for an interesting conversation. In a recent client survey on investment issues, we asked our clients “What time horizon do you feel is the best time frame to evaluate portfolio returns?” The results varied: 16% said “Monthly,” 43% said “Quarterly,” 37% said “Annually,” and 4% said “Over a complete market cycle.” (As an investment professional, I would have selected the last option.)
For the record, there is no correct answer. Most of the investment industry presents their returns on some kind of calendar basis, with monthly, quarterly, annually, and trailing twelve-month being the most popular. Depending on how long a firm has been in business, they may also show trailing three-year, five-year, and ten-year returns. Notably, Global Investment Performance Standards (GIPS) compliant performance reports require calendar year numbers in the report. As you might expect, each reporting period has its own benefits and concerns. For example, choosing any calendar-based reporting period means that the time frame for viewing performance has nothing to do with the cycles of the news, the economy, or the financial markets.
My favorite example of problems with calendar-based reporting involves trailing twelve-month returns. As the calendar rolls forward, the performance result is impacted equally by the new monthly returns earned by the investment advisor as well as the impact of dropping the data from twelve months prior. The table and charts below illustrate the conundrum.
The table shows a hypothetical portfolio where the portfolio manager “missed” benchmark performance each month for the first three months of the reporting period by 2% each month, and then delivered benchmark returns for the remaining 9 months. Chart 1 shows the resulting scatter chart where portfolio risk and return are dramatically less than the benchmark. Chart 2 shows the same portfolio assuming that the portfolio and the benchmark have the exact same risk and return for the next three months.
You can easily see the impact of rolling the trailing three months of data out of the analysis. The dramatic improvement in relative performance does not mean that the manager got ‘hot’ in the last three months and slayed the benchmark, but it does say a lot about how the portfolio was invested over a year earlier.
Looking At Performance Over A Complete Market Cycle
Turning to Pinnacle’s performance, the following table shows each market cycle from the inception date of our managed accounts and the relative performance of our Dynamic Moderate Growth portfolio for each cycle (a market cycle encompasses either a bull market or bear market, defined as the returns from any market trough to peak or peak to trough, respectively):
|Period||Bull/Bear||Relative Return & Risk vs. Benchmark|
|10/2002 to 10/2007||Bull Market||More return/More risk|
|10/2007 to 3/2009||Bear Market||More return/Less risk|
|3/2009 to Present||Bull Market||Less return/Less risk|
A complete market cycle contains both a bull and a bear market, or a bear and a bull market. With that in mind, we have been managing money for one complete market cycle and one incomplete market cycle. The complete market cycle is from October of 2002 (trough) to October 2007 (peak) and back to the March 2009 (trough). The complete cycle lasted six years and five months. The current incomplete cycle has lasted from October of 2007 (peak) to the trough in March 2009 to the present, or four years and three months. It is critical to view active managers through the lens of a complete cycle so you can see if their returns are earned only in bull markets or bear markets, and you can evaluate how their investment process allows them to opportunistically position the portfolio to earn extra returns in both bull and bear environments. Chart 3 shows Pinnacle’s return and standard deviation (risk) for our first market cycle. Our risk adjusted return (alpha) for the period was 2.89%.
Of course, the problem with evaluating market cycle performance is that you often don’t know if a peak or trough in market price is actually the end of the cycle until well after the fact. For example, Chart 4 shows the current cycle from the market top in October 2007 to the trough in 2009 to the peak in April 2011. You may recall that after the April peak, the S&P 500 index subsequently experienced a 19.5% decline in value. If you thought that the May market top represented the end of the bull market cycle, then you would view Chart 4 as a completed cycle. Today we know that the market subsequently rallied to today’s levels. Chart 5 shows a completely different view of Pinnacle performance for the market cycle through January of 2013. The relative performance “misses” in the fourth quarter of 2011 and the first quarter of 2012 make a big difference in how you evaluate the risk and return for the cycle currently in progress. Interestingly, in both time frames Pinnacle earned positive risk-adjusted returns, or portfolio alpha. The alpha for the period ending April 2011 is 2.19% and the alpha through January of 2013 is 0.61%.
It is important for consumers of investment advice to be disciplined when analyzing portfolio returns. Viewing results using a calendar-based approach can be helpful, but we advocate using a complete market cycle as a more appropriate time frame for analyzing portfolio returns. Consumers must balance the need to “check-in” on their portfolio manager to see how portfolio returns are being earned on an ongoing basis, with the necessity of analyzing manager performance at major market turns between bull and bear cycles. The former is useful to see if a manager’s current portfolio is performing as expected in the current market environment and the latter is most useful in analyzing a manager’s investment process. A complete market cycle is much shorter than the “life expectancy” time-frame used by some buy and hold investors, but it is long enough to reach a sound conclusion about how an investment advisor is implementing portfolio strategy.
New Pinnacle clients whose portfolio inception date is post March of 2009 have only been invested for one half of a complete cycle. They will need to wait until the end of the next bear market before they can fairly evaluate portfolio returns over a complete market cycle.
Copyright: flybird163 / 123RF Stock Photo