MarketRiders' investment approach is based upon Modern portfolio theory (MPT), which seeks to optimize investment return while lowering risk. MPT is rooted in the concept of diversification whereby an investor selects a variety of asset classes that collectively lower risk below the risk profile of any individual asset class. Diversification lowers risk because asset classes are generally uncorrelated, or more specifically, change value in opposite directions. The most obvious example of this is when stocks fall bonds often increase, and vice versa. By adding up to six asset classes to a portfolios' architecture, the stabilizing benefit of diversification can be increasingly felt. A collection of asset classes will have an overall lower risk profile than any individual asset class. Ironically, even adding an asset class that might be perceived as "risky" can lower the overall risk in a portfolio, thus diversification has been described as "the only free lunch you will find in the investment game."
MPT is the philosophical opposite of traditional stock picking. MPT is based upon investing in entire markets or indexes, rather individual stocks recommended by business analysts who look for equities that will outperform the index. With MPT investments are described statistically in terms of their expected long-term return rate and their expected short-term volatility. Volatility is equated with "risk" and provides a measurement of how far below the average an investment's bad years are likely to be. The goal of this investment approach is to identify your acceptable level of risk tolerance, and then build a diversified portfolio with the maximum expected return for that defined level of risk.
More technically, MPT models an asset's return as a variable with a normal distribution and risk as the standard deviation of return. A portfolio is then modeled as a weighted combination of these assets so that the return of a portfolio is the weighted average of the assets' returns. By combining different assets whose returns are not correlated, MPT seeks to reduce the total variance of the portfolio. MPT also assumes that investors are rational and markets are efficient.
To implement MPT, one must first determine an investor's ideal asset allocation across multiple asset classes including US stocks, bonds, real estate, foreign, emerging market stocks, and in some instances, commodities. A large part of financial planning consists of finding an asset allocation that is appropriate for an investor. The variables which effect this determination are the investor's age, timeframe in which one needs to use funds, investment experience and one's ability to maintain a policy allocation in the face of market volatility and risk.
In a given year, different asset classes perform differently making it difficult to predict which asset will perform best in a given year. Thus, although it is psychologically appealing to try to predict the "best" asset classic for any given period of time, we consider this approach risky and scientifically unsustainable. Research demonstrates that investors who "jump" from the one asset to another generally end up with worse results than someone following the disciplined approach defined by MPT. Unfortunately, investors can't consistently predict the future.
Portfolio Recommendation Algorithm
Our portfolio recommendation algorithm is programmed to construct for each investor an optimal asset allocation by weighting the following factors: your age, investment experience, risk tolerance and years until you need the money. By answering these questions through moving the sliders within our software a specific "score" is calculated and mapped to a specific matching portfolio. For example, if an investor was to set the sliders to: 50 years old, do not need to withdraw funds from your portfolio for 10 years (an IRA for example), and limited investment experience and a medium amount of risk tolerance, our algorithm would produce a score of 3.5 which corresponds to a portfolio with 40% bond exposure and a specific asset allocation and with it, a group of recommended ETFs.
If, however, you changed your time horizon to indicate you need funds within 3 years, our algorithm will heavily weight this slider so that your portfolio is skewed towards a bond allocation.
Investors are "fitted" with one of 9 basic asset allocations in portfolios that range in the level of bond exposure - from 90% bond exposure to 10%. The algorithm then evaluates the amount you intend to invest and recommends a specific portfolio with a varying number of ETFs. For example, a portfolio over $100,000 will contain a greater number of ETFs than a portfolio of $25,000. More ETFs give you more flexibility in certain asset classes, but trading costs are higher when the portfolio is rebalanced, or when funds are added or withdrawn. For example, a portfolio of $100,000 will be assigned three ETFs to cover the single asset class of the US Stock market. These three ETFs provide more granularity and accuracy when it comes to rebalancing, but also slightly increased expenses. A smaller portfolio of $25,000, however, would be assigned one ETF for the US Stock Market that includes small, medium and large cap stocks. While you don't have the ability to rebalance between these 3 segments of the US Stock market, trading costs are reduced. With larger portfolios, trading expenses are less of an issue.
ETF Recommendations
At MarketRiders, we have culled through a universe of over 900 ETFs to carefully select the best 20 ETFs as building blocks for our portfolios. We used several criteria in selecting these ETFs as follows:
1. Exposure to Core Asset Classes. Since our methodology is based upon MPT, we wish to gain exposure to broad asset classes which have 3 basic characteristics: a) they give basic, valuable differentiable characteristics to an investment portfolio. For example, while real estate protects against the effects of inflation, bonds protect against a financial crisis, b) they rely fundamentally on market generated returns, not on active management of portfolios. In most asset classes, active managers do not outperform the market and because satisfying investment objectives is critical, a core asset class won't rely on luck or serendipity, and c) they are derived from broad, deep, investable markets. As a core building block of a portfolio, we are focused on well-established marketplaces instead of trendy concoctions promoted by Wall Street financial engineers (ala all of the new ETFs currently coming out). Thus, we do not recommend a sector fund like technology, or an index based upon some new quant method of reweighting stocks.
Rick Ferri, the author of "The ETF Book" has developed a helpful way to categorize ETFs according to how they are designed and how they are managed. We use his categorization table below to further distinguish between ETFs in our recommended portfolios. They are all market capitalization weighted and passively managed approach.
2. Volume. After identifying the list of broad based asset class ETFs, we further filter the list by looking at volume. We will only employ the largest ETFs with an average size of $6 billion in net assets where daily volume runs on average over a million shares per day. This affords sufficient volume and liquidity so that the bid/ask spreads are narrow.
3. Fees. Our ETFs are inexpensive with an average fee of .19%. Low fees are key to achieving higher returns.
4. Turnover. We evaluate the annual turnover of the securities within the ETF. More turnover means more taxable income. We require that the turnover be very low, about 10% in the equity ETFs. For bond ETFs, the turnover is naturally greater as bonds mature and need to be continually replaced.
5. Sponsors. We primarily use ETFs from Vanguard and iShares with several from State Street Bank. These 3 firms account for close to 90% of all ETF assets and are the most highly regarded ETF providers in the industry. Since Vanguard is a not for profit institution, with the lowest fees in the industry, they tend to keep the other ETF providers "honest." In addition, the skill sets and technology for managing and constructing ETFs are highest among these 3 firms.
6. Index Construction. We have evaluated each ETF for the way the provider has constructed an index around the asset class and seek to find the best ETF that most precisely replicates a given index's performance over a long period of time. There are various ways to index a market. For example, small cap US stocks, are generally indexed using the popular Russell 2000. But the alternative S&P 600 small cap index is superior due to its construction and how it avoids the drag of what is known as the annual Russell 2000 trade.
Toxic ETFs
The ETF industry has dramatically expanded since the SEC changed the definition of an "index" in 2003. Prior to that time, ETFs were purely limited to baskets of stocks that indexed broad markets and asset classes like the S&P 500 or MSCI EFA for Foreign Developed Country Markets. After 2003, the SEC began allowing an ETF provider to create any set of guiding rules and thereby create a new "index". This has changed the historic definition of an index.
With nearly 900 ETFs today, the original "purity" of ETFs as suitable building blocks for asset allocation has been polluted. One of the most extreme examples of this is an ETF released by a new player, FocusShares, which developed an index of mid- and large-sized companies consisting of casinos, producers of beer and malt liquors, distillers, vintners and producers of other alcoholic beverages, as well as cigarette manufacturers - and called it a "sin" index.
Rebalancing Algorithm
Rebalancing is an essential discipline to the MPT approach to investing. MTP requires specifically defined target allocation percentages based upon the investor's objectives. These percentages need to be rigorously maintained through the ups and downs of the market. The practice of rebalancing is essentially a buy low sell high discipline that is counter intuitive and somewhat contrarian. When an asset class goes up, you trim your gains and add to underperforming asset classes to bring your portfolio in line with target allocations. Although this may feel like the wrong thing to do, research and our back testing software demonstrates that this discipline can add significantly to your portfolio performance year-over-year.
When it comes to rebalancing, our software is designed around a proprietary algorithm that guides investors to rebalance their portfolios back to target allocations when specific conditions are met. These algorithms can be manually adjusted within the software to take into consideration your interest in more active or less active rebalancing. The default setting is optimized so that when an asset class is greater than 15% out of balance, a rebalancing alert is generated. For example, if one's allocation to US Stocks is to be 10%, and based upon market changes, the allocation is now 11.6% (slightly greater than 15% variance), an alert will be triggered. We believe that for most investors, this setting is ideal. Our software, however, allows you to increase or decrease the sensitivity of this setting. If you have a large tax deferred IRA, you may want to increase the rebalancing sensitivity. Likewise, if you have a small portfolio of $25,000, you may want to avoid trading costs and decrease the rebalancing sensitivity. The rebalancing levels are set as follows:
