Situation: One of our recurring themes in this blog is that index investing is a more rational and efficient way to save for retirement than stock-picking. But many of you will want to pick dividend-growing stocks that have a good chance of providing you with retirement income that will grow 2-3 times faster than inflation vs. balanced index funds that barely keep up with inflation. If you’re going to try the stock-picking option, you should pick stocks that are likely to have total returns as good as the lowest-cost S&P 500 Index fund (VFINX) after adjusting for cost and risk. We call such hard-to-find stocks “unicorns.”
Mission: Develop a formal plan for identifying unicorns.
Execution: The goal is to bracket your stock choices between two limiting conditions:
(1) performance that is at least as good as the S&P 500 Index (^GSPC); and
(2) price variance that is no worse than ^GSPC.
Criteria are applied over the following periods: 0-20 yrs, 0-16 yrs, 0-10 yrs, Lehman Panic (10/07-4/09), 0-5 yrs and 0-3 yrs. Stock performance is assessed after 10, 16 and 20 yrs. Stock variance is assessed at all 6 periods.
0-20 years (Columns M & N in the Table)
Tool: BMW Method 20-yr data (see Week 199). We screen the ticker list for stocks with CAGRs that are as good or better than the CAGR for ^GSPC, located after “Z” in the alphabetical ticker list. GAGR = 3.1%/yr at this writing. This multi-input CAGR is a “rolling” least-squares logarithmic calculation of trendline growth based on hundreds of weekly prices over the past 20 yrs and refreshed weekly. Note that in the upper left corner RF=0.66, which means there is a 34% drop in price at -2 Standard Deviations (-2SD) below the trendline. That is the extent of loss predicted to occur within 19-20 yrs. To summarize, we’re looking for stocks with a trendline rate equal to or greater than 3.1%/yr and an RF value equal to or higher than 0.66 (i.e., predicting a loss at -2SD of 34% or less).
0-16 years (Columns O & P in the Table)
Tool: BMW Method employed as above but for 16-yr data. ^GSPC trendline growth at this writing is 2.0%/yr and RF is 0.68 (i.e., a loss at -2SD of 32%).
0-10 years (Column C in the Table)
Tool: total return for a single stock purchase made 10 yrs ago, arrived at by consulting the Buyupside website. Because dividend payments are added to the 10-yr CAGR to arrive at total return, we input the Vanguard 500 Index Fund ticker (VFINX) instead of ^GSPC to arrive at baseline total return. Any stock that has a lower total return than VFINX is rejected. Variance is assessed from 18-month Lehman Panic losses (see Column D in the Table).
Lehman Panic
Tool: We use the same Buyupside website, inputting a Start Date of 10/1/2007 and an End Date of 3/31/2009 for each stock that has passed the above tests. Note that the total return listed in Column D (usually a loss) is for the entire 18-month period. Our baseline, VFINX, had a -46.5% total return over that 18-month period. We reject any stock with a greater loss because its price variance exceeds that for VFINX.
0-5 years (Column I in the Table)
Tool: 5-yr Beta. Reject any stock with a 5-yr Beta higher than 0.9 because a higher value is too close to 1.00, which denotes price variance (reset daily) for ^GSPC over the past 5 yrs.
0-3 years
Tool: We use recent BMW Method prices to assess a stock’s price performance relative to the track followed by ^GSPC prices. In other words, both tracks have to be similar. We exclude stocks that deviate more than 1SD from the track being followed by ^GSPC over the past 3 yrs because that indicates variance (deviation from market behavior). For example, in the 16-yr ^GSPC chart you see that prices since the end of 2012 have risen to almost 2SD above the long-term trendline. Any stock that is still tracking its long-term trendline has to be excluded because its price variance is too far removed from market behavior. Remember, traders expect a rising market to “lift all boats.”
Bottom Line: We’ve set up performance windows at 6 time periods over the past 20 yrs to find stocks that outperform the market while incurring less risk of loss. Only 5 S&P 500 stocks met all of our criteria (see Table). All 5 are in one of the 4 “defensive” S&P industries, those being Consumer Staples, Utilities, Healthcare and Communication Services. Past behavior does not predict future performance: At some point in the future, all 5 of these “unicorns” will move outside one or more of the limiting conditions that we’ve set up. If this method appeals to you, remember to periodically reassess your picks. And remember to consider costs. Our baseline stock index fund (VFINX) has an expense ratio of only 0.22%/yr. You can’t do worse than the market by putting your money there. The average investor who picks stocks has an expense ratio of 2.2%/yr, which is 10 times higher than VFINX. That investor’s expenditures on transaction costs, fees, and investment advice are therefore a material consideration. Even if you hold transaction costs and other expenses of stock-picking to only 1%/yr, total return has to exceed 20%/yr before transaction costs become immaterial in the eyes of an accountant.
Risk Rating: 4
Full Disclosure: I own shares of GIS and PEP.
NOTE: metrics are brought current for the Sunday of publication; metrics highlighted in red denote underperformance relative to our key benchmark (VBINX, at Line 18 in the Table).
Post questions and comments in the box below or send email to: irv.mcquarrie@InvestTuneRetire.com
