Behind the Scenes Constructing the Amerivest Opportunistic Portfolios

Transcription

1 Behind the Scenes Constructing the Amerivest Opportunistic Portfolios Powered by Morningstar Associates The Amerivest Opportunistic Portfolios are constructed to be tactical and more active in their investment orientation and are managed, on your behalf, in a diversified and cost-effective manner. They are designed to be a core holding for investors who desire a fairly active investment program, or to augment a more conservative investor s core assets. These portfolios start out with a core asset allocation followed by an aggressive tactical asset allocation overlay that seek to add return in up markets, while reducing volatility and losses in down markets. Morningstar Associates has access to a wide set of financial tools and market insights to help Amerivest allocate your assets tactically. For example, instead of using the traditional asset allocation breakdown for domestic equities, the models use a sector-based investing to dial in on specific equity sectors in which the manager seeks opportunity. TD Ameritrade s clients can select one of two risk-based asset allocation models from Amerivest Aggressive Opportunistic and Moderate Growth Opportunistic. Both models are constructed using Exchange Traded Funds (ETFs) that are typically more cost effective than actively managed retail funds and allow easy access to a broad set of investment options. As the consultant to Amerivest Investment Management, LLC (Amerivest), we at Morningstar Associates recommend the ETF investment selections, and model allocations for the Amerivest Opportunistic Portfolios. Amerivest, an affiliated investment advisor of TD Ameritrade, retains discretion and on your behalf reviews and validates those recommendations prior to implementation. Once you select the appropriate portfolio based on your suitability and financial goals, Amerivest handles all of the daily portfolio management responsibilities, including the initial account investing, periodic ongoing rebalancing, and other portfolio changes initiated by the client, such as when you deposit more money or request cash withdrawals.

2 Investment Process The following pages describe the investment process we employ which drives our portfolio recommendations to Amerivest. Designed to help investors pursue their long-term financial goals, the process for these portfolios is comprised of five integrated components: projection of market expectations, long term asset allocation, ETF selection, tactical asset allocation, and ongoing monitoring. Each step in our investment process is driven by a rigorous process and a team of Morningstar experts with many years of experience in their field. The overall process is overseen by a team of portfolio managers that use a top-down approach to combine the steps in a way that is unique to Amerivest. 2

3 Capital Market Assumptions The core methodology used to construct our long term strategic asset class models is meanvariance analysis. Mean-variance analysis was developed by Harry Markowitz in the 1950 s and provides a mathematical framework for generating portfolios that maximize expected return for a given level of risk (efficient portfolios), and it can assist investors in making strategic asset allocation decisions. Mean-variance analysis requires three statistical estimates for each asset class: 1. Expected return (Mean) 2. Expected risk (Standard Deviation) 3. Expected relationship between the asset classes (Correlation Coefficients) Morningstar Investment Management (MIM) 1 is widely viewed as one of the leading authorities on the development of capital market expectations. We have written numerous award-winning articles on the subject. Our building block methodology for estimating returns was developed in the 1970s and we continue to improve upon it. We develop capital market forecasts for every asset class by using a combination of historical return data and current market information. We use the building block approach to generate expected return estimates. The building block approach uses current market statistics as its foundation and adds historical performance relationships to build expected return forecasts. This approach separates the expected return of each asset class into three components: 1. Real risk-free rate is the return that can be earned without incurring any default or inflation risk 2. Expected inflation is the additional reward demanded to compensate for future price increases 3. Risk Premia are additional returns that are demanded for accepting additional uncertainty associated with a given asset class. These premia vary for each asset class. Each of these components is re-evaluated annually. We augment our expected return forecasts with a Black-Litterman analysis, particularly for the US Equity Sectors, where we have a short historical track record to analyze. The Black-Litterman forecasting model helps ensure the target allocations that we derived using the econometric mathematical model are consistent across each risk-based portfolio, and are appropriately aligned with current market conditions and investor expectations. Correlations and asset class standard deviations are estimated using a variety of techniques on both relevant historical returns as well as market implied volatility information. When estimating correlations we over-weight more recent observations and underweight more distant history. 1 Morningstar Investment Management is a division of Morningstar. Morningstar Investment Management includes Morningstar Associates, Ibbotson Associates, and Morningstar Investment Services, all registered investment advisors and wholly owned subsidiaries of Morningstar, Inc. All investment advisory services described herein are provided by one or more of the registered investment advisor subsidiaries. The Morningstar name and logo are registered marks of Morningstar. 3

4 Strategic Asset Allocation When creating long term portfolios at the asset class level, we aim to identify long term targets that will attempt to position investors to achieve their investment goals. Portfolios that provide the best risk/return characteristics may not be practical or acceptable to many clients. Furthermore, the most quantitatively efficient portfolios may not take into account possible errors in the input forecast, or the practicality due to liquidity constraints or size (market capitalization) of the asset class. All of these factors are incorporated into our portfolio recommendations. Morningstar Associates employs a variety of techniques and criteria when developing the long term asset allocation targets for the Opportunistic Portfolios. These analyses help during the construction of the long term models, and also help us to be more effective in our tactical asset allocation program. Mean-Variance Optimization Mean-variance analysis was developed by Harry Markowitz in the 1950 s and provides a mathematical framework for generating portfolios that maximize expected return for a given level of risk (efficient portfolios). It can assist investors in making strategic asset allocation decisions. The optimization considers the expected risk and return of each asset class, plus the covariance among asset classes, to determine the combination of asset classes that will provide the highest expected return for any risk level. As it is a mathematical framework, mean-variance optimization does not take into account investor preferences or investability of the asset classes. Often an unconstrained optimization can lead to highly concentrated allocations, often suggesting portfolios that are either non-diversified, highly impractical, or both. A second set of optimizations are performed incorporating asset class constraints. An example of a constraint is limiting the allocation to Non-US asset classes. Re-sampled Mean Variance Optimization In a forward-looking context, capital market assumptions are estimates. The true capital market assumptions are not known with certainty; therefore, it is more appropriate to use an optimizer that accounts for the uncertainty in the estimated capital market assumptions. Conceptually, re-sampled mean-variance optimization is like a giant scenario test in which multiple small adjustments to the starting capital market assumptions are made, and the resulting asset allocations from all of the different scenarios are averaged. Sensitivity Analysis Morningstar Associates asset allocation inputs are based in part upon historical measurements intended to capture the average performance and volatility over various economic and market cycles. Since the future may not conform to our model, sensitivity analysis is employed to evaluate the stability of the asset allocation policy s performance through a variety of alternative input assumptions. It is important that the sensitivity analysis shows that reasonable changes in the inputs do not cause the models to deviate dramatically from established risk targets (although a certain degree of risk deviation is unavoidable). 4

5 The scenarios are purely hypothetical situations that can be used to test the stability of the models in different market environments. Value-at-Risk Traditional Mean Variance Optimization does not account completely for all of the downside risk that we have observed in markets since To construct better models, we use two measures of downside risk; value-at-risk (VaR) and conditional value-at-risk (CVaR). VaR is the estimate of the loss on a portfolio that we expect to be exceeded with a given level of probability (5%) over a time period. CVaR is derived by taking a weighted average between VaR and losses exceeding VaR. We are able to incorporate these enhanced modeling and measures into the asset allocation recommendations. For example, we lower allocations to asset classes that are more prone to downside risks in conservative models. The end result is a set of strategic asset class targets that are diversified across a range of investment categories. We use a broad range of asset classes and many different strategies to help improve the overall risk-return characteristics of our portfolios. For the Amerivest Opportunistic Portfolios which comprises two risk-based models we recommended, and Amerivest approved, the following long-term sub-asset classes and weights for 2013: 5

6 Target Opportunistic Portfolio Allocations (Allocations are subject to change) U.S. Equity Sectors Moderate Growth % Aggressive % Consumer Discrete 3 6 Consumer Staples 3 4 Energy 3 4 Financial 4 6 Industrial 3 4 Health Care 3 6 Materials 1 2 Technology 5 10 Utilities 1 1 Small- and Micro-Cap U.S. Equity Small-Cap Stocks 5 9 Micro-Cap Stocks 0 1 Non-U.S. Equity Large-Cap Stocks 9 17 Small-Cap Stocks 2 4 Emerging Markets 4 7 Specialty U.S. Real Estate (REITs) 2 4 Non-US Real Estate 1 2 Private Equity (Stocks) 0 1 Commodities 2 3 Total Equity and Specialty Fixed-Income Emerging Markets Bonds 3 0 Developed Non-U.S. Bonds 5 2 High Yield Bonds 6 2 TIPS 3 0 Intermediate-Term Bonds 24 4 Short-Term Bonds 7 0 Cash & Equivalents 1 1 Total Fixed-Income

7 ETF Research The process for selecting ETFs is different than for selecting active mutual funds. ETFs generally seek to match the performance of a specific market index, asset class, or sector. They usually have lower annual expenses than mutual funds as they require little or no manager oversight. One additional benefit of ETFs is that they tend to be tax efficient; however, because we expect a high degree of turnover as a result of our tactical approach to Amerivest Opportunistic portfolios, that tax benefit is likely to be negated. In making our ETF recommendations we spend a great deal of time evaluating the particular risk characteristics of ETFs (such as trading volume, liquidity, and discounts). We also spend a great deal of time determining how best to combine ETFs, as their strategies can be much more narrowly focused than mutual funds and may offer less asset class coverage. We obviously don t have to consider manager history in our ETF evaluations. ETF Selection Criteria The ETFs were selected based on a variety of factors including: Tracking error Net expense ratio Average daily volume Net assets Average 12-month premium/discount Number of owners, buyers, and sellers Correlation versus a set of asset allocation benchmarks Attribution to a set of asset allocation benchmarks Benchmark construction methodology Although ETFs are a lower-cost solution than mutual funds, ETFs have some limitations. The main limitation is that they don t have a manager at the helm who can dynamically execute the fund s strategy through upturns and downturns. 7

8 Tactical Asset Allocation The Amerivest Opportunistic Portfolios combine a fundamental (qualitative) tactical asset allocation program and quantitatively driven tactical strategy. The Fundamental / Qualitative Program Morningstar s fundamental asset allocation program is managed by a multi-national team of asset allocation experts (our Dynamic Asset Allocation Sub-committee) that have been managing tactical allocation programs for many years. They meet at least monthly to identify and discuss opportunities in markets that are too short term in nature to be included in an annually updated strategic asset allocation program. The key elements of the Fundamental Program s investment philosophy are to understand the different drivers of long-term returns and risks and the importance of valuation on shorter-term investment returns. The committee focuses on economic, profit, credit and interest rates cycles to help determine reliable valuations and understand investor expectations and pricing. Ultimately, the objective is to improve the long-term risk adjusted returns of investors by maneuvering the portfolio when shorter-term returns are expected to differ from long-term returns. The investment process has three key elements: 1. Assessment of the Economic Cycle Assessment of inflation, growth, monetary policy, fiscal policy and fiscal fundamentals help determine where we are in the cycle and what this may mean for investment fundamentals. The idea is not to forecast economic variables but to understand where we are in the cycle and what the general consensus is expecting. 2. Estimate of Fair Returns and Risks Across Assets Determine fair returns across asset classes using fundamental drivers and sustainable valuations to help determine whether assets are mispriced and whether investors can be rewarded from the current pricing. Valuation and fundamentals drive medium-term returns, and when valuations are extreme, medium-term returns can differ from long-term expectations. This assessment also includes a number of measures designed to assess the riskiness of assets and the potential changes in our underlying fair return assumptions. 3. Assessment of Investor Sentiment, Expectations and Positioning Assess asset classes to identify extremes in investor behavior that may suggest a directional impact on asset prices. Sentiment and expectations can lead investors to adjust their positions to extremes, creating the potential for shifts in investor behavior over the short-term if expectations are not met. This can cause sentiment to change and prices to reverse. 8

9 Investment Positions Investment positions are determined using our TAA investment process, with valuation and sentiment central to the decision making process. Investment positions may be adjusted should the relative opportunity set alter by a change in valuation (price movements or reassessment of the underlying fundamental value), and/or a change in the level of investor sentiment, positioning or expectations. We also consider the diversification of investment positions and may seek to alter a position should we become concerned with the concentration or dependence on any one underlying factor, assumption or risk driving the investment thesis across positions. The table to the right is for illustrative purposes only. Some of the over/under-weights in the asset classes listed in the table may not apply to your specific portfolio. Strategic asset allocation, or diversification, do not guarantee a positive investment result nor do they protect against loss in periods of declining markets. For illustrative purposes only. Portfolio construction This stage combines the above steps to help determine the most appropriate investment allocation given investment objectives, risk constraints, investment line-up, fair returns, and market cycle for investors. Macroeconomic forecasts do not always map clearly into the available ETFs, so it is vital to have a strong, experienced team that is capable of understanding of how macro-forecasts may ultimately translate into ETF returns and opportunities for investors. Investment Time Horizon The overall focus of the fundamental tactical asset allocation process is to outperform over a threeto-five year period, seeking consistent outperformance and ideally preserving capital during bear markets and recessions. We believe that medium-term investment opportunities are created from investors using a short-term focus on macro-economic events, momentum and earnings revisions, as well as a focus on short-term relative performance. This can provide an opportunity for the 9

10 patient and fundamentally orientated medium-term investor. At times, this means underperforming, standing apart from others, and holding positions for lengthy periods; however, it is this approach that is likely to lead to rewarding medium-term returns. The Quantitative Strategy (GTAA) Our Global Tactical Asset Allocation (GTAA) is a quantitative process with two steps. Our Big Signal sets our overall target to equities and fixed income. The second step uses return momentum to set allocations to intra-stock and intra-bond asset classes. Our Big Signal is comprised of a momentum signal and a valuation signal. The momentum component looks at the markets from a short-term perspective, while the valuation signal looks at long-term market dynamics. The two measures result in under- or over-weight signals to equities, which are then averaged to produce the final equity positioning of the portfolio. Big Signal Example Historical Equity % 90 Equity Base Model 50 Equity Base Model This chart shows an example of how our raw Big Signal program drives portfolio positioning. Actual equity / fixed tilts are significantly reduced before they are implemented in the portfolios as part of our risk budgeting process. Source: Morningstar Investment Management. For illustrative purposes only. The first Big Signal component measures the strength of the stock market s momentum whether it was positive or negative. The intuition here is that short-term market performance tends to be auto-correlated, which means those periods of out- or under-performance are more likely than not to be likewise followed by periods of out- or underperformance. In particular, we calculate the market s most recent performance from one up to 24 months, and for each of these periods we record whether it was positive or negative. These positive or negative reads are then averaged in such a way that the most recent and longest periods receive relatively less weight, while the periods whose length is between six and 10 months are weighted most heavily. This becomes our momentum signal, whose strength is then calibrated to translate into an equity over- or underweight of the portfolio. The valuation component measures the market s valuation relative to its own history. This measure relies on extensive academic research documenting that the degree of market under- or overvaluation is inversely related to its subsequent performance over the following several years. To 10

11 measure the degree of under- or over-valuation, we use Professor R. Shiller s 10-year Cyclically Adjusted Price to Earnings (CAPE) ratios, which we standardize relative to their mean. If the market s CAPE is within one standard deviation of its mean, the valuation signal is neutral. If the market s CAPE is more than one standard deviation above its mean, an equity underweight is instituted, which will last until the market gets back within one standard deviation of it s long-run average; similarly, when the market s CAPE is more than one standard deviation below its mean, an equity overweight is instituted. How far out of the normal range the valuations are will determine the proportional degree of the under- or over-weights. After the Big Signal gives us a recommended overall equity level, we use a price momentum methodology to determine tilts within equities, sectors and fixed income asset classes. Total return momentum is measured over the previous six months and momentum signals are allowed to persist for six months. This results in an eleven month signal window. Asset classes that have performed relatively well are over-weighted and asset classes that have done relatively poorly are underweighted. We include an overall tracking error limit to this program to avoid concentrating too many tilts on the same underlying market factors. Tactical Asset Allocation from Momentum Model Source: Morningstar Investment Management. For illustrative purposes only. Combining the Tactical Programs A team of investment professionals meets monthly to combine the signals from both tactical programs. The investment professionals discuss the unique drivers of each program s tilts one-byone and on the occasions where the signals conflict, they make calls on whether they want to negate the tilts or go with one model over the other. 11

12 The models are not allowed to use leverage (go over 100% allocation to ETFs in aggregate), use shorting or invest in leveraged ETFs. We expect the ranges of exposure in the models to stay within these ranges. 50% Equity Opportunistic 90% Equity Opportunistic Minimum Maximum Minimum Maximum Equity U.S International Fixed-Income U.S International Cash Other REITs Commodities

13 Ongoing Allocations Morningstar Associates will make monthly reallocation recommendations in response to its analysis of changing market conditions and current short- and medium-term market opportunities. Amerivest will review and validate those recommendations prior to implementation. We also revisit ETF selections quarterly and our long-term strategic asset allocation targets at least once a year. Because asset class returns, risks, and correlations evolve over time, we want to keep our strategic targets fresh with the most recent data integrated into our various estimates and forecasts. We also periodically incorporate enhancements to our asset allocation policy that grow out of our ongoing experience working with the portfolios. In some cases, we may take into account some near-term reality affecting the market or our forecasts, such as a long term bubble in a certain asset class, which may lead us to temporarily adjust a long term target for an asset class until the dislocation dissipates. Each month, we evaluate the performance of each step in our process. We construct multiple benchmarks to generate attribution reporting. 1. Our Basic Asset Allocation Benchmark is built from stock, bond, and cash indexes. This very simple benchmark is our starting point for performance evaluation and never changes. Our Long Term Strategic Asset Allocation Benchmark is built by blending indexes at the weights of our long term asset class models that change annually. Subtracting the Basic Benchmark from this benchmark helps us understand the impact of our long term asset class modeling. 2. Our Tactical Asset Allocation Benchmark is built by blending indexes at the weights of our tactical asset allocation models that change monthly. Subtracting the Long Term Strategic Benchmark from this benchmark helps us understand the impact of our tactical asset allocation program. 3. Our Model Portfolio Benchmark is built by blending monthly ETF returns at the weights of our ETF model portfolio weights. Subtracting the Tactical Benchmark from this benchmark helps us understand the impact of our ETF selections. ETF fees, premium/discount changes, and tracking error each affect the model in different levels. 13

14 Summary Our investment process is designed to help investors pursue their financial objectives by striving to create portfolios with asset allocation targets that aim to maximize returns for a given risk level. We incorporate a complex tactical program that seeks to take advantage of identifiable trends and mispricing in asset classes, sectors and world regions, while avoiding overly risky markets where we identify a significant risk of loss. It is impossible to predict with certainty which investments will do best at any given time and past performance never guarantees future results. By including a broad range of sectors, sub-asset classes, market caps, and regions, we may help increase the likelihood that some part of the portfolio will benefit from broad diversification. Our use of ETFs helps avoid driving up expenses and unnecessarily disrupting the portfolios. At the heart of our investment process, however, is an experienced investment team that is skilled at calculating the strategic asset allocation targets that help position investors to strive to reach their long-term investment/financial objectives or goals. That team is also skilled at identifying the funds that can help drive performance and weighting those funds in order to hit the asset allocation targets with requisite accuracy. We believe the Amerivest Opportunistic Portfolios are ideally suited for investors seeking a tactical asset allocation strategy process that is overseen by investment professionals with a long history of putting the interests of investors first. About Morningstar Associates For more than 25 years, Morningstar, Inc. has helped investors make more informed investment decisions, resulting in a brand that is trusted by investors worldwide. Morningstar Associates builds on this tradition by combining its expertise in analyzing managed investments with sophisticated, institutional-level portfolio management practices. As a registered investment advisor and wholly owned subsidiary of Morningstar, Inc., Morningstar Associates provides investment management, and consulting services on more than $31.5 billion in assets (as of December 31, 2012). Morningstar, Inc. and Morningstar Associates, LLC are not affiliated with Amerivest and TD Ameritrade. 14

15 Important Information Before investing in an Amerivest portfolio, be sure to carefully consider the underlying funds objectives, risks, charges, and expenses. For a prospectus containing this and other important information about each fund, contact an Amerivest Specialist at Please read the prospectus carefully before investing. Past performance does not guarantee future results. There is no assurance that the investment process will consistently lead to successful investing. Asset allocation and diversification do not eliminate the risk of experiencing investment losses. ETFs can entail risks similar to direct stock ownership, including market, sector, or industry risks. Some ETFs may involve international risk, currency risk, commodity risk, and interest rate risk. Trading prices may not reflect the net asset value of the underlying securities. Amerivest Portfolios is an investment advisory service of Amerivest Investment Management, LLC (Amerivest), a registered investment advisor. Brokerage services provided by TD Ameritrade, Inc., member FINRA/SIPC/NFA. TD Ameritrade, Inc. and Amerivest Investment Management, LLC are both wholly owned subsidiaries of TD Ameritrade Holding Corporation. Amerivest and TD Ameritrade are trademarks of TD Ameritrade IP Company, Inc. Amerivest provides nondiscretionary and discretionary advisory services for a fee. Risks applicable to any portfolio are those associated with its underlying securities. For more information, please see the Amerivest Disclosure Brochure (ADV Part 2) Morningstar Associates, LLC is a registered investment advisor and wholly owned subsidiary of Morningstar, Inc. Morningstar Associates provides consulting services to Amerivest but is not acting in the capacity of advisor to individual investors. Morningstar Associates provides recommendations to Amerivest regarding asset allocation targets and selection of securities appropriate for the Amerivest Portfolios; however, Amerivest retains the discretion to accept, modify or reject Morningstar Associates recommendations. Morningstar Associates selects securities for the Amerivest portfolios from the universe of investments made available through TD Ameritrade. Asset Allocation target allocations are subject to change without notice. Morningstar Associates establishes the allocations using its proprietary asset classifications. If alternative classification methods are used, the allocations may not meet the asset allocation targets. The Morningstar name and logo are registered marks of Morningstar, Inc. Morningstar Associates is not affiliated with Amerivest or TD Ameritrade. TDA 2590 SS 02/13 15