Finding Investment Ideas By Screening Stocks and ETFs Marc H. Gerstein mgerstein@yahoo.com June 14, 2008
About me • Joined Market Guide in February 1999 as director of investment research. Market Guide was acquired by Multex, which in turn was acquired by Reuters. • Worked as a securities analyst since 1980 and has analyzed stocks across a wide variety of industries and sectors, including household products, specialty retail, restaurant, mining, energy, hotel/gaming, homebuilding, airlines, railroads, and media. • Managed a high-yield (“junk”) bond mutual fund during the 1980s. • Created, managed and authored much of the Ideas & Screening section of Reuters.com from 2003-06, which provided analysis and stock selection strategies to investors. • Designed indexes in anticipation of use for ETFs • Author of two books, Screening the Market and The Value Connection, and appears periodically on CNBC, USA Today , CBS.MarketWatch, The Wall Street Journal, Money Online, and The New York Daily News • Has an MBA in finance from New York University, a JD from Brooklyn Law School, and a BA in social science from Ohio State University.
Today’s Agenda • Stocks • Finding ideas using rules-based protocols • Screening • Ranking • Testing your ideas • Evaluating individual situations • Exchange Traded Funds (ETFs) • Making sense of the jumble • Finding ideas • Simple screening • Interpreting available information
Stocks
Rules: What We Are Trying To Do • We want to create mathematical expressions of concepts that are well established as sound stock selection criteria • The key is the phrase “well established” • All of out rules should be consistent with good common sense • We are not trying to turn the world upside down with exotic notions • Start with sensible verbal expressions: Look for … • Shares of companies that are growing briskly • Reasonably priced stocks • Financially sound companies • Translate them to mathematical expressions: Consider stocks if … • Trailing 12 Month EPS growth rate is greater than 15 and above industry average • P/E less than 25 and less than industry average • Latest long term debt ratio and trailing 12 month return on equity above industry average
Benefits Of Using Fundamental Rules – 1 • See past market buzz • “If I could avoid a single stock, it would be the hottest stock in the hottest industry, the one that gets the most favorable publicity, the one that every investor hears about in the car pool or on the commuter train — and succumbing to social pressure, often buys.” • Peter Lynch, Once Up On Wall Street, Chapter 9
Benefits Of Using Fundamental Rules – 2 • Focus us on merit • “Stock screening is absolutely, positively the best way to find investment ideas…. No other approach can match screening when it comes to calling stocks to your attention based on at least some objective showing of merit….” • Page 1, Screening The Market: A Four-Step Method to Find, Analyze, Buy and Sell Stocks by Marc H. Gerstein, Director of Investment Research, Multex (John Wiley & Sons, July 2002)
Benefits Of Using Fundamental Rules – 3 • Effectively Articulate subtle and promising numerical stories • Example Operating Margin comparison ABC Co Industry Average Trailing 12 Months (TTM) 8.6% 8.8% 5-year average 9.0% 11.5% Even though ABC’s margins are slipping and were consistently below the peer average, we see that ABC’s “relative comparison” has become less unfavorable. This may reflect something uniquely positive at ABC We can design and test rules that capture situations like this
Benefits Of Using Fundamental Rules – 4 • De-emphasize popular but generally unanswerable lines of inquiry • We do not chase vague (and often impossible to solve) puzzles about management talent, customers, suppliers, employees, proprietary technologies, and so forth • Few, if any, outside investors can reliably and consistently assess such issues based on publicly available information. • Notice how Wall Street tends to praise or criticize management based not on inherent talent but on who well they “delivered” in the latest period • Notice how dramatically stocks move in response to estimate revision and earnings surprise, analysis of which has grown to become a significant sub-industry All based on the inability of highly trained Wall Street analysts to translate efforts with such questions into reasonable earnings estimate even for the next three months
Types of Rules – Screens • Stocks selected based on data-oriented tests • All stocks in universe are rated “yes” or “no” depending on whether they pass or fail the complete set of tests • In an 8,000 stock universe, we may see 40 passing stocks and 7,960 fails.
A Simple Screening Example • Screening for Value • Stocks pass the screen if . . . • P/E is below industry average, and • Price/Sales is below industry average, and • Price/Book is below industry average
Pros and Cons of Screening • Pro • This is a good way to narrow an overly broad universe • Con • We cannot control the number of passing stocks we’ll have • All tests are considered to be of equal importance
Types of Rules – Ranks • Based on tests similar to those used in screens • But rather than seeking yes/no answers, ranks aim to classify each stock in the universe on a best-to- worst scale
A Simple Ranking Example • A Value Rank • Rank all stocks in the universe from best to worst in terms of • P/E • Price/Sales • Price/Book • Calculate an overall value score based on • 0.50 times P/E rank, plus • 0.30 times Price/Sales rank, plus • 0.20 times Price/Book rank • Rank all companies from best to worst based on overall value score • Determine, for example, that companies whose value scores are in the top 10% are eligible for consideration
Pros and Cons of Ranking • Pro • This allows us to address every stock in the universe and decide in advance how many passing stocks we will have. • We can assign different levels of importance (weights) to each criterion • Con • Used on its own, this technique may strain the capabilities of statistical probability • We may feel comfortable saying the top 10% of the universe is better than the next 10% • But we may hesitate to say stock number 25 is better than stock number 26, that stock number 26 is better than stock number 27, and so forth
Combining screens and ranks • This involves using a rank to identify a broad, appealing, portion of the overall universe, and following up with a screen to make more precise selections from this “subset.” • Continuing with the previous example, we would . . . • Consider stocks that pass the value screen and have value ranks in the top 10%, and • Make our final selection by eligible choosing stocks with the 30 highest value ranks • The previously mentioned probabilities strain is no longer troubling since we also used a screen to narrow the universe
Criticisms against rules-based investing • Rules are based on data from the past and we all know past performance is not necessarily indicative of what is likely to occur in the future • Rules cannot capture qualitative factors such as brand image, management experience, economic “moats,” patents, competitors, etc. • Rules often tend to contain systematic biases against certain kinds of companies • i.e. a heavily earnings-based screen is not likely to give a fair shake to biotech or cable TV or real estate companies, or energy exploration companies whose merits depend on proved reserves • Rules cannot capture every individual situation. • All they can do is establish probabilities, which means we know we’ll have some wrong answers (although we don’t know, in advance, which specific decisions will be the ones that go sour) • Rules tend to be un-sexy • Professional investors interviewed on TV tend to sound a lot less guru-like if they say they picked their winners because of mathematical rules
Deconstructing rules critiques: past performance • Assuming we ignore historic data and just look ahead, are we REALLY nearly as good at forecasting as we like to think we are? • As of 3/7/05, there were 3,945 estimates of quarterly EPS for the current fiscal quarter • Of these, 2,169 (55%) are equal to where they stood 4 weeks ago • Only 1,084 (28%) still stand where they stood 8 weeks ago • Only 851 (22%) still stand where they stood 13 weeks ago • In other words, there’s a 78% probability that the life span of a near-term earnings estimate, the group that get the most attention, will have a shelf life of less than three months • There’s a reason why concepts such as earnings surprise and estimate revision have become standard fare in the investment community • Companies are more like ocean liners than rowboats • They can turn in the opposite direction, but this tends to happen in gradually rather than in an instant • Evolutionary, rather than revolutionary change • So rather than pretend the past is irrelevant, we may as well use it as constructively as we can to help us make rational assumptions about the future
Recommend
More recommend
