Announcement

Collapse
No announcement yet.

Survivorship Bias

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Survivorship Bias

    Just a quick reminder that I'm sure most on this forum already know.  I was using https://www.portfoliovisualizer.com/ to backtest the asset allocation that I've chosen for my portfolio going forward and almost made a huge mistake...

    I have a fellow physician friend that recommends investing in the dividend aristocrats rather than ascribing to modern portfolio theory.  He says that the only reason to buy stocks is to receive the dividends, and companies that have consistently increased their divided for over 25 years represent a great investment.  I decided to test this and chose a somewhat random selection of 20 dividend aristocrats with high current dividend yield:  HCP, T, CVX, EMR, ED, NUE, ADM, PG, XOM, TGT, KO, PEP, MCD, TROW, KMB, MMM, GPC, CL, LEG, and MKC.  When I backtested these across several different 5 and 10 year periods in the last 20 years, they almost always came out with a higher yield over time as well as less volatility when compared to my portfolio and the S&P 500.  This made me really question my asset allocation, but, then I remembered survivorship bias.

    While these may still be great investments going forward, the stocks that I was backtesting are the stocks that survived the test of time.  I looked at http://www.suredividend.com/list-of-dividend-aristocrats-from-1989-to-2014/ to find companies considered dividend aristocrats in 2000.  Several of these companies are no longer in business.  Others that I had chosen for my backtest only became aristocrats in the last couple of years.  If I had performed the same exercise in 2000 and actually bought the stocks, I'm sure a fair percentage would be out of business, split or consolidated into other stocks.  So, long story short, I'm sticking to my plan and remembering that even companies that have been in business for a long time still fail and it's safer to own the whole market.

  • #2
    Good points, no doubt.  Also, as I pointed out in the thread you started yesterday, a high dividend taxable portfolio is the last thing you want when you are in higher tax brackets.  I would much rather have Zero dividends and an equal total return to avoid the tax drag.

    Comment


    • #3
      I agree with you, PoF.  I was just playing with backtesting and was really surprised at how well those stocks did over time until I realized my bias.  BTW, thanks for your comments on the other thread.  I'm spending the week doing my taxes and now, I'm creating my investment policy statement and setting up my taxable investment account.  Trying to get things in order before I start doubling down on savings and start putting away 17k/month in addition to maxing tax sheltered accounts...

      Comment


      • #4
        Dividend investing is so popular at this point in time its in a bubble all things considered. The rate that dividends have been raised and buybacks performed since the crisis are unlikely to be seen again as that was an artifact of cheap capital. Many slow to no growth companies, say PG, are bid to levels that make zero sense, even KO is contracting at best yet is valued otherwise. People have rotated to these names in the recent downturn and furthered increased their bids and thus further decreased their forward returns. If/when this dividend bubble pops, it will be the final kick for this market and its real capitulation.

        What your friend doesnt tell you, or more likely doesnt understand is the idea of risk adjusted returns. Buying these names now when they have very high PE/PEG ratios makes no sense and basically will commit you to understated future returns. Make your best risk adjusted return or tolerable volatility portfolio. If the return is less than you desire, leverage it until it is or the volatility is in line with your goals or benchmark. You will have steadier and in all likelihood greater returns than your friend, all the while not being exposed or concentrated into any of the risks out there.

        I wish I had saved this great sequence of return risk paper/volatility to throw on that bond thread. Lemme google this for a minute. Two very good papers I've read recently about just this sort of thing.

        Ok, here is the one about sequencing. Its great, and shows you why you may not want a 100% stock portfolio, unless you are certain to be drastically over investing for retirement or can tell whether or not volatility will be in your favor or not the last 10 years of your career or so.

        https://www.gmo.com/docs/default-source/research-and-commentary/strategies/asset-allocation/who-ate-joe-s-retirement-money-sequence-risk-and-its-insidious-drag-on-retirement-wealth-.pdf?sfvrsn=20

        Comment


        • #5




          I agree with you, PoF.  I was just playing with backtesting and was really surprised at how well those stocks did over time until I realized my bias.  BTW, thanks for your comments on the other thread.  I’m spending the week doing my taxes and now, I’m creating my investment policy statement and setting up my taxable investment account.  Trying to get things in order before I start doubling down on savings and start putting away 17k/month in addition to maxing tax sheltered accounts…
          Click to expand...


          You're off to a great start, and clearly an excellent salary.  With that kind of savings rate and a little advice (which you are wise to request), you are going to be in great shape! Keep up the good work.

          Best,

          -PoF

          Comment

          Working...
          X