A Bad Argument Of Why Buy-And-Hold Is Bad Advice

A regular reader Don sent me a post entitled Long Term Buy And Hold Is Still Bad Advice. Okay, fine, everyone and their mom has been telling me this recently. But I read it, and it was such a bad analysis that I had to rebut it here. I think Mish writes a lot of useful and thought-provoking stuff on his popular blog, but he really missed a big error here.

First, a recap of the post. Basically, a guy called “TC” has the idea of comparing S&P 500 returns vs. that of 6-month CDs. I’ll ignore the fact that this has been done many times already. But wait! He comes up with a startling conclusion. For long periods of time, the S&P 500 has actually lagged or been about equal to the returns of safe and steady 6-month CDs. (!!!) His graph:

Keeping my parents in mind, you’re probably wondering how someone did by simply investing in 6 month CDs. The answer is for any holding period of less than 25 years, a stock market investor who made regular and equal contributions has actually underperformed a CD investor! Yes, you read that right for time periods of 1 – 20 years a CD investor outperformed the stock market by 1.6 to 20.1 annual percentage points.

Additionally, if one extends the time window to 50 years (clearly “long term”) CDs again have outperformed the stock market by 0.3 annual percentage points. Even when one extends out the time period to the full 59+ years (the start of the S&P 500 index); the stock market has outperformed short-term CDs by a mere 0.2 annual percentage points – not much of an equity premium.

The sky is falling! Oh wait, there’s a little fine print.

TC is ignoring dividends

Let’s bold that. The analysis and data above completely ignores the dividend return of the S&P 500. This is like buying an investment property and ignoring the rent payments coming in. What? There are checks coming in every month from the tenants? Nah, let’s not cash those.

Let’s take a look at the historical dividend yield of the S&P 500, courtesy of Bespoke Investments:

For the periods compared above, the a true owner of the S&P 500 has earned 2-6% annually from dividends alone, with a long-term average of 3-4%. Now, if you add another 3-4% to the analysis above, you see again the long-term equity premium. Instead of 8% vs. 8%, it’d be more like 12% vs. 8%. That’s an enormous difference.

(I also wonder where TC got his/her data for historical 6-month CD rates. Are these averages, since every bank offers vastly different rates, and doesn’t report them to a central bureau? How does one get the average 6-month CD rate across the country in 1959? Usually studies like this use 6-month US Treasury Bill rates instead, as the data is reliable and widely-accepted.)

Massive Conflict of Interest?
Another argument given as to why buy-and-hold is bad is because there is a conflict of interest between investment advisors and their clients, as they have a “vested interest in keeping clients 100% invested 100% of the time, even if they know it is wrong.”

Actually, brokers get paid the more you trade than anything else. They earn money based on total assets, but a huge chunk is from commissions. This means convincing you to buy stocks when they’re hot (tech stocks)…. and then sell them (cash!)… and then buy others (mortgage-backed securites)…. and then sell them (cash!)… and then buy new ones. Like right now, they’ll happily sell you gold or some non-scary bond funds!

True buy-and-hold means very little trading. At Vanguard, I buy-and-hold(-and rebalance) for a total cost of about 0.20% of assets annually. That’s $20 a year per $10,000 invested. Guess what the average expense ratio of a money market fund is? According to Lipper Inc., it was 0.60% at the end of 2007. The Vanguard Prime Money Market fund (VMMXX) has an expense ratio of 0.28%. The S&P 500 fund (VFINX) charges 0.18%. Even at Vanguard, they actually get less money from me if I hold stocks instead of cash.

Same Old Story
In any case, I grow weary. Bonds have outperformed Stocks both recently and other times in the past, even if people ignored it. This is why investors need to have a balance of both stocks and bonds/cash, not just 100% one or the other. If you needed the money soon, then you should have been at the most 60/40 in stocks/bonds, if not even more conservative. In that case, your portfolio would have dropped about 15% over the last couple of years up until today, and you’d be worried but not broke.

If you use the correct numbers (ahem), stocks still have higher historical returns over extended periods, with many rocky patches. We balance this knowledge with the also-historically steadier but lower returns of bonds and cash. That’s really about it. As for the future, nobody knows, as much as they’d like to suggest they do.

Comments

  1. Way to challenge another blog. Lets get a blog war started. I’m on your side.

  2. I have been seeing a lot of “buy and hold is dead” type posts lately. Most center around the argument that you should be buying and selling asset classes depending on what’s hot or not as you stated above. I guess the feeling seems to be that it’s ok to miss the top or the bottom, but you need to get in on the slide.

    I.e. it’s ok to sell S&P at 1200 instead of 1400 when it was headed to 700. However, you’d stick with more specific asset class funds.

    Then there is the whole thing that stocks are a sham and overly inflated due to tax incentives to invest and that we should all be in quality bonds like our grandparents.

  3. Ephraim says:

    I actually wanted to ask you about a similar topic to this when I came across this review for the Little Book of Common Sense Investing from Amazon, specifically this one: http://www.amazon.com/Little-Book-Common-Sense-Investing/product-reviews/0470102101/ref=cm_cr_dp_synop?ie=UTF8&showViewpoints=0&sortBy=bySubmissionDateDescending#R42VLAO6M9K3A

    That also says Buy-and-Hold is a bad strategy (more specifically, through Index funds).

    It’s the most recent 1 star rating for the book I found, and I actually can agree with some of the things he says, although I’d like to hear your perspective on it myself.

  4. xreflux says:

    i’m not. jon is writes very well and thoroughly about many subjects, but the one subject he doesn’t treat fairly is market timing or analysis. and its obvious in seeing the results of his own home purchase and investments. and fine, that’s his style, but it seems like he never gives the serious research a chance, because he always falls back to the line well you can’t predict the future. well i say you can’t predict it, but you can get a pretty good idea of what might happen and place your bets accordingly, IF you do you research. and i think the amounts you’re investing (home and stocks) grow, the more attention should be spent do the research on where to put them.

    the ironic thing in his promotion of buy and hold is that he DOES make assumptions of the future and allocate accordingly. it just happens to be that his data source is past data. either way, i wish jon the best.

  5. I agree, ignoring dividends is a bad anti-buy-and-hold argument. However, the primary argument used FOR buy-and-hold, which is “you might miss a big market up day”, is equally bad. Missing the worst down days will put you way ahead of missing the best up days. Unfortunately, it’s impossible to predict when these days will come. However, the market does move in cycles of uptrends and downtrends, and the best uptrends have occurred during bubbles (energy, real estate, etc). These are typically followed by severe downtrends. Sticking by a staunch buy-and-hold philosophy when the market is showing signs of a clear downtrend (lower highs, lower lows) ends up costing a lot of money in order to stick to a principle. Personally, I believe there aren’t too many bubbles left to inflate, so I use sector ETFs which allow me to ride the uptrends and avoid major losses when an individual sector turns over. No commissions (other than the small trading fee), and no 40% losses.

  6. Wow, this blog post is really silly. Like this part:
    “One simple, active strategy that would have avoided the stock market holocaust in both the recent recessions would be to get out of the market when the yield curve inverts and stay out until the NBER announces the recession has ended. [...] Using the above two charts one would have exited the stock Market in Spring of 2000 and reentered in November of 2001. One would have exited the stock market in Summer of 2006 and would still be out.”

    Ok, that sounds great. Except one problem. The NBER didn’t announce the 2001 recession ended until late 2003! How would you have known that November 2001 was what they would pick without a time machine?

    Great advice!

  7. christina says:

    I adore your blog. Thank you.

  8. You really should read this white paper on asset allocation. Very simply, be well diversified and only own asset classes when they are above the 200day moving average (otherwise, cash/treasuries). Best part, you only need to check it once a month. The result is slightly better than S&P 500 returns with half of the volatility/risk.

    http://papers.ssrn.com/sol3/papers.cfm?abstract_id=962461

    This is a VERY solid argument against buy and hold.

  9. Thanks, Jonathan. This stuff needs to be said.

    And if anybody else on here hates listening to the phrase, “buy and hold is dead”, simply don’t watch cnbc between 7 and 10pm weekly because that’s all you’ll hear.

    Note to the whole Najarian family…………I CAN’T TRADE BAZILLIONS OF SHARES OF BAZILLIONS OF DOLLARS WORTH OF STOCKS FOR FREE EVERY DAY AND NOT WORRY ABOUT WHETHER OR NOT THEY GO UP OR DOWN WHEN I JOKE ABOUT IT ON MY NATIONALLY SYNDICATED TV SHOW AT NIGHT!!!!!

    Ugh!

  10. If you bought Vanguard’s S&P 500 Index (VFINX) exactly 20 years ago at 20.02 and held it through today it would have gone from 20.02 to 84.68 (price adjusted for dividends).

    That equals an annualized return of 7.477%, which is not very likely to be achieved in CDs. Furthermore today’s interest rates are extremely low, and will likely stay low for another couple of years. In my opinion CDs are almost certainly an even poorer investement (for someone wanting to achieve growth) relative to common stocks today than they have been anytime in the last 5 years.

  11. dirty_dirty says:

    jon, i’ve been following your blog for several yrs now. and i really bought into the “ultimate buy and hold strategy” you wrote about and follow.

    and even though i’m losing my shirt, i whole heartedly believe it is the right strategy. it’s funny how people always look short-term and can’t seem to grasp the big picture. the s&p 500 is a great way to invest and all these nay sayers just make me buy more. things will eventually turn around, whether it be 5, 10, 15 yrs from now. and i want to be invested when it happens.

    thanks jon, and keep up the good work.

    dirty_dirty in FL

  12. Thanks. It’s valuable to see such incomplete arguments combusted.

    The author may have had good intentions, but by giving partial information he will lead people into making bad decisions. Peer review is accountability, not war.

  13. Buy-and-hold is certainly based on the past information. The question is what you deem to be a more reliable assumption.

    1. Stocks will have an overall future return based on earnings growth and dividend yield. Bonds will have an overall future return based on current yield. Average investor return will be this return minus costs. An index fund will capture this average return no matter what.

    2. Market timing in an out based on publicly available signals or exceptional stock selection, which has been shown to be ineffective even by large pension funds spending millions on “advice”, will improve your return in a net positive way forever. This means that to average out, others will underperform you.

    You can believe 1 and/or 2, but I think 1 is much more reliable. If you go with #2, you could win, or you could lose. And if your costs are higher, on most people who try this will lose.

    But that’s just me. If you do have a clever way to “beat the market” consistently that really works, you will be massively rich, so go for it. :)

  14. Richard says:

    Sincere thanks for taking the effort to educate people. Unlike many other blogs, this is the only one where I see some time and effort taken on most posts. Most other blogs just talk about their views on a published article or so…Most of them lack, detailed analysis and true technique revealing (not on this case, but on those applicable).

    Love your blog. keep up the good work.

  15. Cramer has an entire book on how buy and hold sucks.

  16. What? He forgot to add dividends to the total returns? Well according to Jeremy Siegel dividends accounted for 97% of US total stock market returns between 1871 and 2003.

    Touche ;-)

  17. Jonathan, read the paper and try to refute it. You’re a methodical, well read person. Don’t simply dismiss it by lumping it into the snake oil that others sell. There is a reason the paper is the top downloaded of 190,000 white papers on http://www.ssrn.com in the past 12 months. His findings have been validated in real time (i.e. post backtest and paper publishing), particularly in 2008 when it was flat vs -36% S&P decline. I understand your skepticism toward “systems”, but all great investors apply a diligent process. Meb’s process being well diversified and avoiding protracted market downturns. It is fully transparent, easy to follow, and doesn’t involve the purchase of anything (though he has written a fantastic book following up the paper). Give it a read – you’ll be pleasantly surprised.

    http://papers.ssrn.com/sol3/papers.cfm?abstract_id=962461

  18. sauropod says:

    You’re right that it’s crazy to ignore dividends in assessing *past* performance of the stock market. But the problem in relying on dividends for *future* stock growth is that many companies have reduced or eliminated their dividends.

    The book “Bonds,” by H. and S. Richelson, makes this point well (pp. 14-15):

    “The classic explanation of stock appreciation is that it is principally driven by two factors:
    - High dividend yields
    - The growth of dividends over time

    “From 1926 to 1954, the dividend yield on large company stocks was always above 5 poercent…. From as late as 1975 to 1985, the dividend yield on large company sticks was generally around 5 percent. The dividend yield on large company stocks in 2006, however, was about 1.8 percent, and at least the majority of midsize and small-company stocks paid no dividends at all.

    “The reinvestment of dividends has been a major driver of large stock appreciation. When stock performance includes the reinvestment of all dividends paid on stocks, $1 invested in 1824 grew to $3.2 million in 2005. However, if dividends are left out of the calculation and not reinvested, $1 invested in 1824 grew to only $374 in 2005…. [Not a typo. $374.00] Since high dividends were the main driver of stock appreciation in the past, why would we expect high appreciation in the future when dividends are much lower?”

    Having practiced buy-and-hold religiously with Vanguard since 1992, and not having a lot to show for it now, I would sincerely advise you to avoid my mistakes. At the very least, learn to follow long-term moving averages so you can ease out of stocks and into cash near the start of a downturn. If I had done this, I would have been out of stocks by late 2007 and would have at least 30% more capital today.

    The Meban Faber paper linked by a commenter above (the “papers.ssrn” link) would have saved me a small fortune, if I’d known about the MA strategy at the time.

    The problem with buy-and-hold is that (as someone once said) “the market can remain irrational longer than you can remain solvent.”

  19. I’ve skimmed the paper before on investing based on moving averages, and will post some of my amateur thoughts shortly. The idea is not new to that paper (see Merriman or Shiller for example), but always seems to pop up in consciousness after a rise and crash, as in 2001.

  20. I downloaded and read the paper on market timing from http://www.ssrn.com. I’m confused with one particular calculation. So when you’re buying and selling based on simple moving average method, did the author take into account dividends? If you out of the market for a short period of time, don’t you miss the dividend payment?

  21. The analysis was also done on another site using the S&P500 + dividends and comparing it to CDs:

    http://www.ritholtz.com/blog/2009/07/sp500-vs-cds-1994-2008/

    1994 – 2008

  22. AS – Yes, that pretty much aligns with this study. Just take the IRR of S&P 500 and add the dividend rate for last 15 years, it is still less than the IRR of 6-month CDs. There have and will continue to be 15 year periods where bonds (basically same as CDs) outperform stocks.

    But over 40 years? No, bonds have not outperformed S&P 500.

    What would be better would be to plot a 60/40 stock/bond portfolio with annual rebalancing. I suspect that would be my preference, even in hindsight.

    Thanks for the link though, now I see where they get the 6-month CD rates (Federal Reserve, based on secondary market).

  23. Jasan Mundur says:

    There is a good discussion on Mebane’s paper on bogleheads.org
    http://www.bogleheads.org/forum/viewtopic.php?t=27460&start=0
    Some points to consider
    1. Different timings resulted in different returns.
    2. The outperformance/underperformance reported is just a little bump to be counted as significant.
    3. Systems when followed widely do have a tendency to underfperform
    4. Already mentioned above – the study avoids dividends lost when out of the market.
    5. Taxes are not considered when getting out of the market.
    6. Technical traders talk about whipsaws that can take a psychological toll that can be more frequent than a Buy_hold_rebalance_stock_bond strategy.
    7. Always beware of strategies that are resurrected right after a abnormal market action.
    Your choice.
    Jasan Mundur

  24. @Jasan – Your comment is eerily similar to my post draft! I like your use of the term resurrection, because as I said using moving averages is not a new technique.

  25. Money God says:

    Here is an article that does show CD’s beating stocks with the dividends included and also includes the great years for being in stocks between 1995-1999.

    http://www.ritholtz.com/blog/2009/07/sp500-vs-cds-1994-2008

    The fact is an investor can’t really get good returns in the stock market by simply buying and holding unless he is lucky to start buying at the begining of a secular bull market and gets lucky by having to sell before the next secular bear market begins. Someone buying in 1982 and having to sell for retirement in 2000 would have done very well. But buying half way through a secular bull market and riding out a secular bear market would result in sub-par returns compared to bonds and CD’s as shown in this article. Most of these “set and forget” investor’s do not get lucky. What they usually get is burned. I know of no long-term buy and hold investor that is happy with their 401k returns that they had for the last 15 years. They made very little money.

  26. Chicken Littles everywhere! This isn’t surprising because it happens every time there’s a big market downturn. In the 30′s and 70′s, stocks were declared dead. Then in the 50′s, 60′s, 80′s and 90′s people were buying them like crazy. History repeats itself.

    Is it really surprising that after a big downturn the short and medium term (less than 20 years) comparison of returns to the market won’t be super favorable to the stock market?

    All the link above this comment makes me want to do is buy more stocks. It shows that, most of the time, you will more more money investing in stocks than with CD’s. The is also not counting taxes — which makes the analysis skewed more towards CD’s.

Speak Your Mind

*