INVESTING AT LEVEL3

Level3 Active Strategies

Active approaches involve creating the equivalent of your own mutual fund or funds. You must select one or more approaches that you feel will exceed the 12% to 13% return of a passive approach and then implement it. Some of the approaches examined here require a frequent revision of holdings, and others would have very gradual changes.

They all involve serious continual monitoring and a rigid discipline. Just a little carelessness will offset whatever advantage your approach has. It is so easy to get careless, particularly in up markets. Up 31% instead of up 34% seems immaterial, but it is the same as down 2% instead of up 1% in long-term calculations.

It would be nice to be able to lay out a dozen well-defined approaches that have provably beaten the market by 5% to 7% over the long term. Unfortunately, �provably� is a hard nut to crack. There are many approaches with different levels of proof, but none without some limitations on the certainty that they were not biased in some way.

There are three sources of strategies that may outperform the market: model portfolios with defined rules, enhanced advisory services, and academic and other historical research.

• Model Portfolios
• Enhanced Advisory Services
• Historical Research

Model Portfolios

AAII Model Shadow Stock Portfolio

Throughout “Investing at Level3”, AAII�s Model Shadow Stock Portfolio is mentioned several times. This portfolio comes close to meeting the criteria of Level3 Investing in order to have confidence in the returns. At a long-term return of 15.4% annualized, the Model Shadow Stock Portfolio has a return that is meaningfully better than passive approaches. It is a real portfolio with the costs and problems that entails. It has 23 years of history behind it (it was started in 1993). However, as an estimate of how well you would have done or will do in the future, following its approach has some limitations:

Since the portfolio and any variations of it are real portfolios, the money invested is finite. When the portfolio is out of excess cash, new stocks, no matter how appealing, cannot be bought until there are sales. It is likely that no two portfolios will be exactly the same�even though they are following the same rules�since different investors may run out of excess cash or find new funds at different times. This bias should be neutral on average but indicates that individuals may not achieve equal results.

Since stocks are bought for the Model Shadow Stock Portfolio prior to the actions being announced, the impact of others buying the same stocks later should be beneficial to the model portfolio. This is a positive bias to the portfolio and makes its returns higher than they might otherwise be.

Since the portfolio rules are spelled out in detail and always followed, anyone is free to take positions prior to the portfolio announcement. It could be a day ahead or a month ahead. This bias is negative and makes the model portfolio returns less than they might otherwise be.

Magic Formula Portfolio

Another well-defined strategy that has been monitored through time is Joel Greenblatt�s �The Little Book That Still Beats the Market� (2010). His Magic Formula strategy can be used to provide recommended stocks at different capitalization levels and is available for use by all investors at magicformulainvesting.com.

There are not very many model portfolios maintained in real time and followed over the long run. Active mutual funds could be considered such portfolios if the precise rules for portfolio decisions were spelled out and maintained over the long run. But funds jealously hide their rules and they tend to vary their decision processes as well as their managers over time.

Enhanced Advisory Services

Advisory services generally make buy and sell recommendations and maintain a list of their recommendations. Very few show what an actual portfolio based on their recommendations would look like and what the actual return would be over time. Over the years, the Hulbert Financial Digest (HFD) has turned most advisory letters into model portfolios and tracked their performance.

Unfortunately, HFD stopped publishing in early 2016. The last data published showed that John Buckingham�s The Prudent Speculator has produced a long-term (35-year) average return of 15.8%.

In his 2016 Honor Roll, Hulbert showed several advisory letters that have significantly outperformed the market over the long term. These include:

• Bob Brinker�s Marketimer
• The Buyback Letter
• Fidelity Investor
• Investment Quality Trends
• Investor Advisory Service
• Morningstar StockInvestor
• Outstanding Investments
• The Oxford Club
• Sound Advice
• Zacks Premium

These are only a few of many advisories that have outperformed the market for 15 years or more. [Note: We do not receive any compensation for listing these newsletters.]

In most cases, advisory letter portfolios, or HFD�s interpretation of them, equally weight the selected stocks. This in itself should account for 1% to 2% of the return, so to beat Guggenheim S&P 500 Equal Weight ETF or other equally weighted indexes, I would suggest that an effective advisory must beat the market by 3% over the past 15 years.

While the Hulbert Financial Digest is gone, the results for the long-term winners will be valid for quite some time. A list of stock advisory letters that outperformed the S&P 500 by 3% or more for at least 15 years as of December 31, 2015, and met the other recommendations of Level3 Investing can be found here.

Historical Research

To uncover other approaches, we must rely on simulations used over historical periods. Voluminous academic research has been done on the so called anomalies. Unfortunately, much academic research is aimed at explaining stock behavior rather than developing effective investment strategies.

Typically, the research shows that a particular investment strategy makes a positive difference, but it is difficult to translate the findings into return estimates. The list of stock characteristics that increase returns over the market return introduced in Chapter 1 of the book have all been supported by academic research, but in most cases long-term estimates of the actual returns are not easily inferred.

While all of the strategies listed in Chapter 1 have shown some improvement over market returns, many have only slight impact or there is contradicting research.

The following are the factors that seem to provide the most significant impact on returns, but even these sometimes must be used in combination in order to make a significant improvement over passive strategies.

• Price-to-cash-flow ratio (P/CF): Cash flow seems to be more important than earnings. Earnings are calculated through the principles of accrual accounting, which attempts to match expenses to revenues when the revenues can be expected or recognized. Accrual accounting introduces many interpretations and estimates by management into the financial statements. Decisions regarding the capitalization of expenses, the recognition of revenues, the creation of reserves against losses, and the write-off of assets are just a few of the factors that may vary from firm to firm. Cash flow can be measured directly from the cash flow statement or adjustments can be made to the income statement. When adjusting the income statement, non-cash expenses are added back to net income after dividend payments are subtracted. Non-cash expenses such as depreciation, depletion and amortization are expenses that appear on the income statement but require no cash outlays. They represent the accountant�s attempt to measure the reduction of the book value of assets as these assets are depleted. While dividends are a discretionary item, they are a real cash outlay that is not tax deductible and is not reflected in net income. There is evidence that there are positive anomaly earnings associated with high cash flow; therefore, low price-to-cash-flow ratios are favored.
• Price-to-sales ratio (P/S): Usually used to eliminate stocks with high price-to-sales values, but may require adjustments for typical industry profitability. Unlike earnings and book value, sales are less subject to management assumptions and more difficult to manipulate. Furthermore, all companies that are going concerns have sales and positive ones at that. Therefore, the vast majority of companies will have meaningful price-to-sales values. Lastly, sales tend to be less volatile than earnings, making the price-to-sales ratio a more reliable means of valuation. However, price-to-sales ratios do not generally work well for financial firms such as banks, where sales are not a driving force. Furthermore, the price-to-sales level is driven by profit margins, which tend to vary from industry to industry. Companies in industries with low profit margins, such as supermarkets, tend to sell with very low price-to-sales ratios. For this reason, it is best to compare price-to-sales ratios across companies in similar industries or lines of business.
• Shareholder yield: The combination of dividend yield and stock buyback yield seems better than either figure alone. A shareholder yield shows what percentage of total cash the company is paying out to shareholders, either in the form of a cash dividend or as cash spent to repurchase its shares in the open market and reduce the number outstanding shares. While there are a couple of ways to calculate the buyback yield, the easiest is to look at the change in the number of outstanding shares. A stock�s buyback yield is determined by comparing the average shares outstanding for one fiscal period with the average shares outstanding for another fiscal period. If a company is paying a 5% dividend yield and has a buyback yield of 10%, its shareholder yield would be 15%.
• Earnings surprises: Reported earnings above expectations have a positive impact, and earnings below expectations have a negative impact that lasts into the future. Not an easy approach to use without close monitoring, but it appears to be very effective.
• Relative strength/momentum: Stocks going up tend to keep going up. If a stock has displayed significant positive relative strength against the market, it tends to continue to display positive relative strength into the future. Improving relative strength or momentum may signal that the market recognizes that a stock is attractive and help pull in other investors that continue to bid the stock price up. While relative strength is a worthwhile analytical technique on its own, its usefulness increases when combined with other stock selection criteria. Notably, requiring low valuations or strong earnings growth with high relative price strength ranks tends to lead to better stock price performance than when the factors are used independently.
• Company capitalization size: The smaller the capitalization, the better. A strong factor, but it can have long periods (two to three years) of underperformance.
• Price-to-book-value ratio (P/B): Lower values are a strong predictor of excess returns. Works best with factors that provide confidence in stated book value. The price-to-book ratio provides a relatively stable measure of value, which can be compared to the market price. It is a valuation metric that can be used even if a company has negative earnings, which renders the price-earnings ratio not meaningful. However, over the history of the firm, many events occur that can distort the book value figure to where it bears little resemblance to current economic values. For example, inflation may leave the replacement cost of capital goods within the firm far from their stated book value, or the purchase of a firm may lead to the establishment of goodwill, an intangible asset, boosting the level of book value. Different accounting practices across industries may also come into play.
• Return on assets (ROA): A higher ROA is a strong value factor. In the earnings-to-assets ratio, earnings before interest, taxes, depreciating and amortization (EBITDA) can be substituted for earnings and enterprise value can be substituted for assets. EBITDA is used to analyze and compare profitability between companies and industries, since the effects of financing and accounting decisions are eliminated. EBITDA is a good metric to evaluate profitability, but not cash flow. Enterprise value (EV) is a measure of a company�s value and is thought by some to be a more accurate measure than market capitalization because it includes debt in its calculation. Enterprise value is often calculated as the sum of market cap (historical), short-term debt, long-term debt, minority interest and preferred equity less cash. Enterprise value is sometimes referred to as a company�s theoretical takeover price. In the event of a buyout, an acquirer would have to take on the company�s debt, but would pocket its cash. Whereas multiples that use the stock price look only at the equity side of a stock, enterprise value includes a company�s debt, cash and minority interests and therefore gives the total value of a company.
• Gross profit-to-assets ratio: A newer anomaly finds gross profit (revenue minus cost of goods sold) more predictive than earnings. Gross profit is revenue less the cost of goods sold and reflects the pricing decisions of a company as well as the costs of the materials needed to produce their goods and services. A higher relative or absolute ratio is attractive.

Combinations of these factors, particularly value factors paired with accounting measures, provide a strong interactive effect. This is because firms headed for disaster possess attractive value measures as the value inputs are lagging in time, so an added check on financial strength helps to eliminate these weak stocks.

O�Shaughnessy Research

Fortunately, �What Works on Wall Street� (2012) by James O�Shaughnessy has taken these various factors that promise market-beating returns and simulated long-term portfolio returns. He tests the individual factors, as well as many combinations of them. This book is an essential source for any investor wishing to develop actively managed portfolios.

The only problem with the book is that it eliminates, with a few exceptions, the use of smaller-capitalization stocks because O�Shaughnessy feels that only institutions will be doing this type of active investing. His cutoff of $200 million in market capitalization would be appropriate for institutions, but most individual investors can effectively include stocks with market caps of $50 million, and in many cases even lower. The Model Shadow Stock Portfolio goes down to $30 million if there is reasonably active trading and acceptable bid/ask spreads.

Any of the approaches O�Shaughnessy is testing should perform even better with a lower market-cap limit. As it turns out, the few tests he did with micro-cap stocks turned out to be the most effective approaches. When O�Shaughnessy does use micro caps, he goes down to $50 million with restrictions that eliminate stocks individual investors wouldn�t be able to buy.

�What Works on Wall Street� isolates some additional decision factors but really expands past research by looking at multiple combinations of factors used in portfolio decision-making.

A word of caution: When putting together this type of simulation, it is difficult to include the cost of investing in the real world. Researchers sometimes add estimates of the impact of bid/ask spreads and commissions, but they generally are too low, particularly for small- and micro-cap stocks.

Most simulations use �last� price in calculations, which distorts reality for less-active stocks. The price that determines your return is the price you would get if you sold and the price you would pay when you buy; this is the �ask� price when you buy and the �bid� price when you sell. For example, if a stock is trading with an inside spread of $8.44 bid and $8.66 ask, a purchase of 100 shares would cost $866 ignoring the commission cost. If you were to turn around and sell the stock, you would only get $844 for your 100 shares (again ignoring commission costs and now SEC transaction fees). This quick change from a �buy� or ask price to a �sell� or bid price represents a 2.5% reduction in the reported price of last two transactions, even though the price of the bid/ask spread did not change.

The market can move dramatically since a last price occurred. Simulations should use the bid price, the ask price, or an average of the two rather than last price. Many rules require frequent turnover and adjustments, which multiply the cost impact of the bid/ask spread.

Based on real-world experience, an approach that indicates a return of 20% in O�Shaughnessy�s research might be closer to 16% when actually implemented, but there is no way to verify this.

Accordingly, focus on O�Shaughnessy�s approaches that indicate returns of over 15% in order to exceed both the market return and the return possible from a passive approach. There are over 60 approaches that provide that return or better in his book, although many of them are very similar. His book covers most of the combinations of strategies that have been discussed in research.

The bottom line is that O�Shaughnessy�s research shows that stocks with small and micro capitalizations, value factors and momentum are the winners. However, the devil is in the details and the specific combinations of value factors is important to get the results he shows.

In addition, there are value components that, when combined with larger-cap stocks, can also provide significantly above-market returns and useful diversification at the same time.

Studying �What Works on Wall Street� is a first step to any active portfolio management.

The most recent edition (fourth) of the book is 2012 and the data is through December 31, 2009. That includes the Great Recession and the successful strategies have been consistent over the years. A list of those characteristics O�Shaughnessy has tested in simulations through the years that seem to be the best can be found here.

Some of O�Shaughnessy�s simulations are based on data back to 1927 and some only to 1965. Some of this is due to the databases with CRSP (Center for Research in Security Prices, Booth School of Business, University of Chicago) going all the way back to 1926 and Compustat, which has additional details only going back to 1965 for quarterly data. Wilshire data goes back to 1971 for both cap-weighted and equal-weighted portfolios.

AAII Stock Screens

The American Association of Individual Investors (AAII) also evaluates multiple strategies through its stock screening series available to members. AAII stock screens include the most researched of the anomalies in academic literature and simulations of the approaches of numerous gurus such as Warren Buffett, Benjamin Graham, Joseph Piotroski, Josef Lakonishok, William O�Neil, Peter Lynch, John Neff and many more.

Wayne Thorp�s evaluation of the more than 60 strategies in the January 2016 AAII Journal, showed more than half besting the S&P 500 over the past 10 years. Even after adding the 4% transaction cost estimate I mentioned above, 17 strategies beat the S&P 500. However, when dealing with micro-cap stocks or strategies that require very short-term adjustment periods (monthly or quarterly), the transaction costs may be higher than 4% a year.

The AAII stock screens research is available to AAII members online and is updated regularly. There is certainly overlap between the AAII stock screens and the simulations in O�Shaughnessy�s book, and factors that have shown to be successful in both should be strong contenders for any active approach.