Investment Strategy

September 18/19, 2013

    Day before yesterday, I estimated the long-term, inflation-adjusted total rate of return on the U. S. stock market to be around 4% per year. Yesterday, I set about checking that out.
    Starting in 1871, the market indices rose at an inflation-adjusted rate of 1.4%-to-1.5% until the latter 1990's. 

    However, as the chart shows, the S&P 500 moved dramatically above its upper trend line in 1997. 
Where the upper trend line would have been in 2000, measured in constant 2000 dollars
    It looks to me as though the S&P 500 upper trend line intercepts the right side of the chart (January 1, 2000) at a value of about S&P 500 = 750. Allowing for three years of inflation between 1997 and 2000, its inflation-corrected year-2000 intercept might have been 850 (in year-2000 $). In reality, the S&P 500 peaked at 1,535 toward the end of 1999... 1.8 times its normal upper bound of 850, and much higher-priced than it had ever been in its previous 130 years of stock market history! (After correcting the chart for year-2000 dollars, at the beginning of 2000, the S&P 500 curve would have touched the top of  the aqua "picture frame" directly above the 2000 axis on the chart.)
Where we would expect the upper trend line to be in 2013, measured in constant 2013 dollars
    Since 1997, inflation has run 46%. A 1.5%-per-year rate of rise in the S&P 500 upper trend channel would place it today 26.9% above where it was in 1997 (720?). By now, we'd expect it to be about 914 in 1997 dollars. Correcting this for inflation gives us 1334 as an upper bound for the trend channel today in 2013 dollars.
    Obviously, at yesterday's close of 1725, the S&P 500 is way out of its trend channel, and way above its normal bull-market-to-bear-market turning point. But the strange thing about this is that the P/E ratio at S&P 500 = 1725 is 19... high, and near the top of its normal range, but not in bubble territory.

    But getting back to the subject of the total rate of return on equities:
    Dividends typically range from, maybe, 5 % at typical market bottoms to 2 % at typical market tops. But guess what? Yep. During the bubble of the 90's, dividend yields plumbed all-time lows, bottoming at 1.11% in 1999! At the time, the argument was made that dividends are "so yesterday"... that it made more sense to retain corporate earnings and use them to buy back corporate stock, thereby boosting the price of the stock. Given that corporate compensation includes company stock options, corporate executives had incentives to boost the prices of their companies' stock rather than give excess money to the owners of the companies. An examination of the recent history of dividend payouts shows that this practice of pouring money into boosting stock prices rather than paying it out in the form of dividends may have continued to the present time.
    This may explain the puzzle mentioned above that the S&P 500 is at 1725 when the upper limit for its trend channel lies at 1334 (give or take a few). A rise in stock prices may have substituted for a rise in dividends. Let's estimate how much of an offset higher stock prices might represent. If we take S&P = 1875 for the S&P price corresponding to the upper-limit, turning-point P/E ratio of 21, then stocks are 1875/1334 = 1.4 times their historic price at a normal upper turning-point. That would correspond to a dividend yield of around 2.75%.
    Another way to assess this is to derive the additional boost provided by running the S&P 500 index 40% above its historic upper trend boundary without pushing its P/E above its normal trend boundary value of something like 21. We can estimate that by calculating ln(1.4)/16 = 2.1% per year. This means that the total return on the S&P 500 since 1997 is its historic 1.5% per year plus 2.1% per year because the upper boundary of its trend channel has moved up faster than it did over the previous 126 years. (For the past 16 years, the top, and presumably the bottom lines in the S&P trend channel have been moving up at 1.5% per year + 2.1% per year = 3.6% per year.) In addition, there have been some dividends returned by the S&P 500... maybe 2% a year? That would bring the total inflation-adjusted rate of return to 5.5% a year.
    Looking at this another way, over the past 16 years, the stock market has returned value to us investors in two ways: through the distribution of dividends, and through the lowering of P/E ratios from their 1999 peaks. The dividends have been historically low, but they've been real money delivered to our accounts by the S&P's stocks. The lowering of P/E ratios has been more subtle, as we've gradually transitioned from dangerously overpriced stock valuations to values that, while on the high side, are at least back within normal limits.
    Jeremy Siegel has shown that since 1802, the stock market has returned 6.5% to 7% per year after inflation. However, he expects real returns of 4.5% to 5.5% through 2020.

    The formula below assumes an after-inflation total rate of return  of 6% a year.

Annual Retirement Income = Annual Retirement Savings Rate X (2t/12 - 1), where t is the retirement savings period, in years.

    Suppose that my wife and I can salt away $20,000 a year in Annual Retirement Savings. And suppose that after retirement, we continue to invest our retirement savings in a stock market that returns 6% a year after inflation. Then after saving for 12 years, we'll be able to withdraw an annual retirement income equal to our annual rate of retirement savings ($20,000 a in two $5,500 Roth IRA contributions plus $9,000 a year in a Roth 401k account) X (212/12 - 1) = $20,000 a year (in 2013 dollars) on $360,000 in retirement savings (in 2013 dollars). We can withdraw this in perpetuity without shrinking the purchasing power of our savings. (Enough money is held back each year to keep our nest egg even with inflation.)
    Suppose that we save $20,000 a year for 24 years. In that case, our annual retirement income will be our rate of savings X 3 = $60,000 a year (2013 dollars) on $1,080,000 in retirement savings (2013 dollars).
    Suppose that we save $20,000 a year for 36 years. In that case, our annual retirement income will be our rate of savings X 7 = $140,000 a year.(2013 dollars) on $2,520,000 in retirement savings (2013 dollars).
    Suppose that we save $20,000 a year for 48 years. In that case, our annual retirement income will be our rate of savings X 15 = $300,000 a year.(2013 dollars) on $5,400,000 in retirement savings (2013 dollars).
    Does 48 years sound like an unrealistic length of time to work? Today, if you worked from the age of 22 to the age of 70, your employment career would stretch over 48 years.
    It's too big a topic to include here, but I suspect that most of the readers of this page will be able to see their "youth spans" extended by 10 or more years compared with what's been available for my generation. In that case, 48 years of employment might not be so bad if it afforded a married couple $300,000 a year in perpetual income without drawing down the couple's retirement savings accounts. 
    When radical life extension finally occurs, Social Security will be (I should think) a wild card. Many of us have already paid a lot of money into Social Security, and a lot more money will probably flow into Social Security before radical life extension happens. So I should think that all this prior investment in Social Security will need to continue to be distributed to those of us who have invested in Social Security.
    Actually, if we paid in $7,500 a year to Social Security, plus $7,500 a year from our employers for 35 years at this 6%-a-year total rate of return, our perpetual income from Social Security could be about $105,000 a year. But this assumes that the government would invest its money in the stock market in order to get this hypothetical 6% a year real rate of return, and that may not be what the government would do. But Social Security could provide a major source of income, especially if the Social Security Administration got to keep our principal after we died as it does now.
    One thing to remember is that in retirement, we no longer have to save for retirement. In the example given above in which we sock away $20,000 a year in Roth accounts, we have to pay income tax each year on that $20,000 we're saving, which brings its total annual cost up to about $28,000 a year. The same thing goes for Social Security where it costs us about $10,000 a year to put $7,500 a year into Social Security. This means that once we're retired, we need $38,000 a year less than we did when we were working.
    Another expense we generally don't have in retirement is life insurance. With a guaranteed annual income, our significant others are still supported even if we're not around to earn money.
    Robotics may contribute to retirement financial scenarios, along with more and more do-it-yourself goods and services. What role will 3-D printers play? 
    It's hard to know how things will play out over the next 20, 50, or 100 years, especially given the exponentially rising rate of change that we're currently experiencing..

    Want to add two years to your "health span"? Eat an ounce of raw almonds, walnuts and/or pistachios each day. (Loma Linda University is responsible for this claim.)
    Want to add another two years to your "health span"? Take a baby aspirin a day.
    Want to add still more time to your "health span"? Eat a four-ounce serving of wild Alaska salmon twice a week, and check your serum levels of B12, folate, and vitamin D to make sure you're not deficient in these vitamins. Drink two or three cups of green or white tea with two heaping teaspoons of high-polyphenol "raw" cocoa a day. Add a daily dose of Longvida curcumin as an anti-inflammatory. Add astaxanthin and the tocotrienols to your daily regimen.
    Will this add years to our "youth spans" and life spans? That's what I'm reading. What do you think?.

    Tonight's (September 19, 2013) update of Tuesday's model portfolio is given below. It's up 3.63%


Price Purchase Change Change, % Value Shares Gain/Loss Gain/Loss, %
NSM $57.35 $54.45 +$0.21 +0.37% $10,532.61 183.655 +$532.61 +5.33%
KORS $77.49 $74.30 +$1.77 +2.34% $10,429.38 134.590 +$429.38 +4.29%
LL $113.13 $108.62 +$0.32 +0.28% $10,415.20 92.064 +$415.20 +4.15%
XRS $13.84 $13.09 +$0.83 +6.38% $10,572.96 763.842 +$572.96 +5.73%
ADUS $28.21 $25.84 +$0.28 +1.00% $10,917.19 386.997 +$917.19 +9.17%
SYNA $42.32 $41.40 -$0.45 -0.45% $10,222.23 241.546 +$222.23 +2.68%
GMCR $84.32 $84.51 -$2.64 -2.64% $9,977.50 118.329 -$22.48 -0.22%
WLK $106.42 $104.77 +$0.60 +1.15% $10,157.47 95.447 +$157.47 +1.57%
ISBC $21.75 $21.45 -$0.22 -1.00% $10,139.85 466.200 +$138.86 +1.39%
SYNT $78.54 $76.52 -$0.52 -0.66% $10,264 130.685 +$263.98 +2.84%


$103,628.38   +$3,628.36 +3.63%

    Last night's (September 18, 2013) update of Tuesday's model portfolio is given below. It's up 2.95%


Price Purchase Change Change, % Value Shares Gain/Loss Gain/Loss, %
NSM $57.14 $54.45 +$2.11 +3.83% $10,494.05 183.655 +$494.05 +4.94%
KORS $75.12 $74.30 +$0.70 +0.93% $10,191.15 134.590 +$191.12 +1.91%
LL $112.81 $108.62 +$2.58 +2.34% $10,385.74 92.064 +$385.75 +3.86%
XRS $13.01 $13.09 +$0.06 +0.46% $9,038.89 763.842 -$61.12 -0.61%
ADUS $27.93 $25.84 -$0.26 -0.92% $10,808.83 386.997 +$808.82 +8.09%
SYNA $42.51 $41.40 +$0.50 +1.19% $10,268.12 241.546 +$268.12 +2.68%
GMCR $86.61 $84.51 +$0.76 +0.88% $10,248.24 118.329 +$248.24 +2.48%
WLK $105.21 $104.77 +$0.60 +0.57% $10,041.98 95.447 +$42.00 +0.42%
ISBC $21.97 $21.45 +$0.14 +0.64% $10,242.41 466.200 +$242.42 +2.42%
SYNT $79.06 $76.52 +$0.37 +0.47% $10,331.96 130.685 +$331.96 +3.32%


$102,951.36   +$2,951.34 +2.95%