American Madness

Intelligent Criticism in the Service of a Better Nation

A basic problem of logic

Posted by Josh Friedlander | 6 Comments

Today’s NYTimes contains an article by Ron Lieber that I’ve read hundreds of times in different guises by different authors.

The basic argument is “buy and hold”: keep your money in the stock market during this blip because you’ll be sorry if you miss a resurgence. Articles like this explain that, yes, it would be great to put money in cash and wait until the market decides to rebound, but there’s no way to know when that is, so you’d better stay invested.

A guarantee of a small loss may sound good right now. But if you’re not bailing out of stocks once and for all, how will you know when it’s time to get back in? The fact is, any peace of mind you gain by being on the sidelines now will turn into a migraine once you see how much you can harm your portfolio over time by missing just a bit of any rebound.

What always strikes me as funny is the basic logical fallacy of this argument. The author says if you sell now you’re market timing and no one can know when to get back in, but at the time time he’s basically telling you to invest NOW. There is no actual distinction between not redeeming money now and putting in new money. They work out to the same result, and both actions involve timing.

Since there was no way for me to comment (why do some NYTimes articles allow comments and others do not?), I wrote the author an email, adapted from my post of the other day:

Sir, yours is a common but seemingly odd argument at this time, given that we’ve just endured a “lost decade” and one would have made more money having been invested in T-bills over the past 10 years. It certainly is the case that missing the big bumps in the market would hurt one’s portfolio as compared to not missing them, but the idea that being out of the stock market would be worse is dubious.

In Maggie Mahar’s “Bull: A History of the Boom (1982-1999),” she cites Martin Barnes who wrote in his newsletter, The Bank Credit Analyst, in March 2003 that “The ratio of equity returns to returns on 30-year Treasuries is now back to the level of September 1980. In other words, a buy-and-hold investor would have done just as well holding Treasuries as investing in the S&P 500 over the past 22 years.” He adds: “Wall Street’s dirty little secret is that in the 34 years from February 1969 to March 2003 stocks outperformed long-term Treasuries by a paltry 1% a year, on average. This is a dismally small gap, given the extra volatility in stocks” (page 356 of Bull).

Of course, if investors had bought stocks in 1982 and sold in 2000, they would have beaten Treasuries handily, but that kind of strategy requires a crystal ball, the same sort of timing genius that we both know no one can rely upon. Perhaps stocks are better over time, but over which time period?

The Wall Street Journal noted in a March 31 article that the S&P 500 then stood at exactly the same level as 9 years ago. They are calling the previous period the Lost Decade. Recently thereafter, I observed that, as of September 17, 2008, the S&P closed at 1,156.39. On September 17, 1998, the S&P closed at 1,018.87. That’s a 13.4% total return over 10 years, an effective compound annual rate of growth of 1.27%. The yield on 10-year Treasuries on Sept 17, 1998 was 4.8%. Of course, the numbers are far worse after this week, even considering dividends.

Dead or negative periods are common. Bear market periods of negative or minimal stock market growth, adjusted for inflation (and with dividends reinvested) include 1882-1897 (3.4%), 1903-1921 (0.6%), 1930-1949 (3.2%), 1967-1982 (0.2%). (Bull, page 5.) Without dividends, those periods are all negative. More from Bull (page 22): “During any particular 10-year period from 1926 to 1998, an investor’s chance of averaging more than 10% a year was only 50-50. Contrary to popular wisdom [an investor] stood a 4% chance of making nothing over 10 years and losing some of his principal to boot. Everything depended on when he got in and when he got out.”

If that’s not marketing timing, I don’t know what is. The problem of knowing when to redeem is just as important, or more important, than knowing when to initially invest. Based on the data, any other conclusion is a product of marketing by fee-based investment companies. Your argument effectively shills for them. I know you didn’t intend to do it, but you’re a smart man, so I hope you stop drinking the mutual fund kool aid.

The standard rejoinder usually comes from the value investor caucus, who have done so well since 1982 (the greatest bull market in our history). Joel Greenblatt in The Little Book that Beats the Market, has a wonderful story to tell. Funny thing: his Compustat data goes back no further than 1988.

Some similar thinking:
Taking stock of profits: Over past seven years, T-bills equal or better stocks.

That article includes this valuable chart taken from Jeremy Siegal’s Stocks for the Long Run:

Holding Period % of time stocks didn’t beat T-Bills
1 year 38.5%
2 years 34.7%
5 years 26.0%
10 years 19.9%
20 years 5.5%
30 years 2.9%


6 Responses to “A basic problem of logic”

  1. Paul Woodland
    October 11th, 2008 @

    Any self respecting financial professional should be able to time the markets to some extent. The dirty little secret is that when the small long term investors bail, as they are doing right now, everything goes to shit on wall street. Therefore, the “experts” have a very strong incentive to lie to the average investor about when they should get out of stocks.

    Further, most people in the finance industry are just salesmen and paper pushers. They don’t have a fucking clue what they are doing or any interest in understanding the machine they are part of. Unfortunately, their plans rarely extend beyond getting drunk at the end of the day.

    Those who knew what they were doing took their money out of stocks a long time ago.

  2. The Unknown Comic
    October 11th, 2008 @

    “So you’re Levine?

    You call yourself a salesmen, you son of a bitch?”

  3. Josh Friedlander
    October 12th, 2008 @

    Coffee is for closers.

  4. American Madness » Blog Archive » Retirement is a foolish goal
    October 12th, 2008 @

    [...] American Madness Intelligent Criticism in the Service of a Better Nation « A basic problem of logic [...]

  5. Maggie Mahar
    October 16th, 2008 @

    Thanks both for citing Bull!–and even more, for the lucid commentary on “buy and hold.”

    The financial media should be warning people: “buy and hold” worked well from 1982- 1999. But in the years ahead, it could be disastrous. Investing is always a matter of tiiming: it’s about buying relatively low and selling relatively high. In the end, it’s all about when you get in and when you get out.

  6. Josh Friedlander
    July 12th, 2009 @

    Jason Zweig writing in the WSJ yesterday demolished Jeremy Siegel’s old stock figures. I am smacking my head because I did not think to take a look at where Siegel got his averages, assuming (stupidly) than an academic of his stature wouldn’t use such fuzzy a methodologically screwy numbers. Apparently, the notion that stocks necessarily beat bonds, even over 30-year periods, is flawed. Check out the column:

Leave a Reply

  • Trust us

    As with Anna Karina, we prefer to remember the U.S.A as she was in the 1960s.
  • Archives

  • RSS Matt Friedlander’s Tumblr Feed

  • RSS Josh Friedlander’s Twitter Feed