Wednesday, May 18, 2011

Cut Your Losers and Let Your Winners Run

We frequently here the stock-picking advice “cut your losers and let your winners run” (for example, see this SmartMoney article). This advice sounds smart. After all, isn’t better to own winner stocks that loser stocks? The problem here is that we only know a stock’s past and not its future.

A personal counterexample was my purchase of Apple shares in 2000 and later sale in 2003. I had lost 14% of my investment over those 3 years; so it seems that I followed the rule because I sold a loser. However, Apple stock went on to grow at an average compound rate of about 45% per year since then. My shares would be worth more than a million dollars had I held on.

Of course, one example doesn’t disprove a useful investing rule. And some may argue that I actually sold a winner that had a temporary setback. However, this is something that I didn’t know at the time I made my decision to sell. We’d like to think that we could have known that Apple had big things ahead, but this is just hindsight bias.

I’d be interested in seeing a study of whether cutting losers and keeping winners actually works in the sense that it beats the market. The study would have to precisely define which stocks are winners and which are losers based on past data and not future data. Without a convincing study of this type, I remain sceptical that this advice is anything more than a way to exploit our hindsight bias to make us feel bad about stocks that didn’t work out.

18 comments:

  1. Michael, your blog gave me a chuckle. I have had several such stocks similar to your Apple; which in my mind suggests that selling losers and letting winners run would not in many cases not prove correct. Winners have a momentum affect which I think validates the investing rule in many cases over a short term horizon. However,as you note, i am not sure it is valid over a longer time horizon.

    ReplyDelete
  2. A corollary to your example is how a winning stock can become a dog. Example: I owned E*Trade at one time. Bought at $10 in 2003, sold at $25 in 2007. Today, it is trading at a split-adjusted $1.60. So much for holding on to a winner.

    ReplyDelete
  3. Great point. Perhaps winners are actually more dangerous than losers. Winners will by definition become more expensive over time. If this is due to price rise alone (not backed up by fundamentals), selling a winner is the right thing to do.

    I sold a whole bunch of tech winners in 1999 when I finally noticed P/Es in the 100x range. Like CC's example, it turned out pretty well.

    ReplyDelete
  4. Right up there with fund advertisements which trumpet past performance, then have a mandatory disclaimer telling you past performance doesn't mean anything.

    ReplyDelete
  5. @Blunt Bean Counter: I'm guessing that anyone who has invested in individual stocks can come up with examples where this advice would have worked and others where it wouldn't have worked.

    @CC: I definitely had some good luck of that type with a former employer's stock before it crashed.

    @Gene: Maybe the study I called for would show that the opposite strategy is better.

    @Preet: It seems to be a common theme in ads is for the fine print to say that the rest of the ad isn't true.

    ReplyDelete
  6. Regarding Preet's comment about fund advertising:

    "What the large print giveth, the small print taketh away."

    ReplyDelete
  7. "Nobody ever went broke taking profits" is also something thrown around a bit. Also somewhat dangerous, as it limits your profits.

    ReplyDelete
  8. @Anonymous - well said.

    @Michael - Winners can be very dangerous, easy to get overconfident with them and lose perspective, get greedy. Neither of those are ever good...

    BTW - I moved my site Michael, just in case you wanted to find me on the web. I got a new site, now self-hosted:

    www.myownadvisor.ca

    Cheers,
    Mark

    ReplyDelete
  9. ...reminds me of a saying of the ancient Greeks that went something like this "the only man who can say he has lived a happy life is the one who dies happy". Who knows what will happen to Apple next month/year/decade? I daresay that sooner or later, down it will go. Market timing is everything!

    ReplyDelete
  10. Letting winners run does not mean 'forever' - making it difficult to collect data. You know that at some point you take profits.

    Cutting losses does prevent disasters, and to me that is the key to trading.

    ReplyDelete
  11. Michael, I came across a study that examines long-term returns of a momentum based strategy. It looks like that paper approximates well your case "selling losers and buying winners". In short, it actually does outporform market (index). Check it out here and let me know what you think: http://dorseywrightmm.com/downloads/hrs_research/GMOMomentum.pdf

    ReplyDelete
  12. My link got cutoff. Here's the link again:

    http://dorseywrightmm.com/downloads/hrs_research/GMOMomentum.pdf

    ReplyDelete
  13. http://dorseywrightmm.com/
    downloads/hrs_research/GMOMomentum.pdf

    ReplyDelete
  14. @Dragan: I'm not sure what the problem is, but I can't seem to access the link you provided. I'll try again in a few days.

    ReplyDelete
  15. Michael, check on Google docs, title is Momentum – A Contrarian Case for Following the Herd

    https://docs.google.com/fileview?id=0B9DwlVhskXsoM2NjMzEwYTUtNjBlNi00OGFjLWJkZGQtYzQ5ZTdiOGQ0YzE5&hl=en

    ReplyDelete
  16. @Dragan (or is it Draga?): The study is very interesting. It seems that certain specific momentum strategies worked in the past, although the specific way to implement a momentum strategy may not have been known until after all the data was analyzed. I'm suspicious about the fact that the momentum strategies analyzed stopped working over the last 10 years. As the report says, it may not work any more due to the large amounts of money devoted to employing momentum strategies.

    I also wonder about trading costs, including spreads. Often it is difficult to buy a rising stock at the price you want, and there are commissions to pay (possibly taxes as well). I'd need to see more about the analysis methods used in the paper to form an opinion about how profitable the momentum strategies would have been net of costs.

    ReplyDelete
  17. Michael, what I found interesting in the study is that momentum didn't work for short term past winners which is contrary to the common wisdom that momentum doesn't persist long due to market efficiency.

    ReplyDelete
  18. @Dragan: I'm not a believer in complete market efficiency. However, the manner in which the market is inefficient seems to change over time and we only seem to know about inefficiencies after the fact.

    ReplyDelete