Sell In May?

I received the following question;

What’s your current thinking this year to the old adage: “Sell in May and go away”? (for the summer that is, for the folks not familiar with the saying)

I’m in a rather large 40% cash position at the present time.

I don’t really act on these sorts of things in a meaningful way for a couple of reasons; first is that there may not a enough of a fundamental case to do it. The second reason is that I think I heard recently that it only works 54% of the time. I may have that wrong.

I can also recall, however, another study that showed the market does much better from November until May by a dramatic amount. The nature of this second study I am vaguely remembering focuses more on better returns during the winter as opposed to declines in the summer. Anyone with specifics on either is encouraged to leave the links.

Obviously in 2006 sell in May was 180 degrees wrong. In looking at the Stock Trader’s Almanac for the last ten years from May to October inclusive sell in May was also wrong in 2005, 2003, 1999 and 1997. Selling in May looks like it was right in 2002, 2001, 2000 and 1998 and I am going to say it was a push in 2004.

I did not look further back due to time constraints and I would not really be able to defend an argument that says the time sampled is flawed, maybe because it includes the bubble and the aftermath. Hopefully you have a Stock Trader’s Almanac and you can study it further if you are so inclined.

The way I might incorporate sell in May into my thinking might be to sell one stock for a little cash if I thought the market was extended. I currently have a stop order for a portion of the position for one across the board holding that is up a lot. As the market has had a big run if the stop order gets elected I don’t think I would redeploy that cash. There are also one or two more other names that are up a lot that I have been mulling partial stop orders for.

Having three stop orders as described above all execute in a downturn seems compelling to me. It would probably raise 3-4% in cash, not a big bet but a way of letting the market reduce your exposure.

Big bet. I think that is what the reader has made by having a 40% cash position, unless of course the other 60% is in double long funds. The dilemma with such a big cash position is knowing how to get back in. I would not know what to do with such a big cash position taken for tactical reasons.

I have been bearish for nine or ten months. I have chronicled many times my skeptically going along for the ride so clearly I have no useful input for anyone with such a large tactical cash balance.

The thing that should be feared is down a lot. While that is the case, the thing to consider is that down a lot doesn’t happen very often but when it does happen it gives plenty of time to get out. I would urge anyone managing their own portfolio to learn the history of this. I would add that down a little, like late February, should not be feared nor traded around by most folks.

The obvious question that would arise from that last paragraph is how do you know that a decline, like the one that started on February 27, won’t be big? Well you can’t know but you can put the odds in your favor. Fast moves down that create fear usually come back fast. Slow moves that don’t bother people have historically been more of a problem for the market.

11 Comments

  1. My tiny mind can not handle hedging with a myriad of options. I like and I need simple. A possibility, though, is using LEAPS alongside a large cash position. The idea, I read, is that if the mkt goes up, you participate and if the market goes down you loose only your premium.

    Roger, any thought on this combo. My goal is to have a large cash position, circa 40%, without suffering your caveat of too large a bet. What do you think of LEAPS as an option? I’d be using them on etfs, if that is possible. If you think there is merit, much appreciate any of the how to details.

    Reply
  2. It’s pretty easy to test the Sell in May type strategy. With all due respect, I don’t think a 5 or 6 year sampling off the top of your head is a real good measurement.

    If you take the DJIA or SPX closing data from yahoo I believe you can go back to 1960. *MOST* of the time, this holds true. As a whole, things seem to be flat to down a bit during May through Sept (my strategy bought again the 1st of Oct). I think the idea here is that you might be a little better off selling your holdings and putting the money in a CD or a high yield savings account until October.

    Obviously, what doesn’t get factored in are transaction costs, taxes and hassle. What also doesn’t factored in my backtesting software is dividends. The divies paid out from May-Oct may make this even less attractive.

    If anyone cares, I can rerun the strategy and produce actual numbers.

    Reply
  3. Here’s a hedge suggestion. Take a SMALL position in a 2x fund such as UAPIX and use a short term moving average (50 to 100 day) as a timing trigger.

    http://tinyurl.com/ywg8xj

    As you can see in the chart, a stategy like this would have been a wash during the 2000 to 2002 timeframe but you would have made some excellent gains during prolonged bullish periods from 2003 to today.

    A few caveats. This strategy is susceptible to single day plunges like Feb 27. Also, you will take the wrong position on 2/3 of your trades. If you’re not comfortable with this, do not use this strategy.

    The idea is to miss extended bear moves but to participate in a good part of any upward movement.

    Reply
  4. I tend to agree Roger.
    Here is what I wrote this morning with links to several sites discussing this very theme.

    http://tinyurl.com/2b5c95

    REW

    Reply
  5. Hi Roger

    I decided to test this one a while back using actual data. If I recall correctly (it’s been a while) I downloaded daily S&P 500 level from 1950 or so from Yahoo. Then I calculated daily returns by calendar day throughout the year, and averaged the 50+ years. Result is a historic average of what you would have done throughout the calendar year.

    Bottom line was that since 1950, just about all of the year’s gain comes between roughly October 1 and first week of May. Summer months give you break even result. So you would have done slightly better in cash getting some small interest during summer. The backtest suggests that the “sell in May” suggestion is true. You can get almost exactly the same long term result with half the risk (out of market 50% of the time) using seasonal timing. The only thing that really bugs me is since I can’t really point to a fundamemtal reason why this should work, it’s hard to be too enthusiastic believing it’s not some major fluke, or that the past driver will continue to work in the future.

    Also, to really trade this every year you would have to pay taxes (short term holding rate) every year, incur transaction costs, slippage, etc. So not practical for most small investors. Not to mention that it’s a tough one to stick to if you get several positive summer years in a row – average of breakeven comes from a few big down years averaged with majority of years up a little.

    On the other hand, time and volatility premiums are fairly low right now, so an investor can hedge some or all of their exposure using put options at fairly modest cost. Not right for everyone, but the possibility is there for anyone who like me is nervous about valuations generally.

    Reply
  6. Michael, thanks for the heavy lifting.

    You raise a question to which I am not sure there is an answer.

    There is some sort of statistic floating around that just about all of the markets return comes from just a few days every year.

    Miss those days and you miss the market. I wonder what the overlap of the two would be.

    Further, in the ten years I casually glanced at there were some huge months (not days) between May and October.

    This doesn’t feel like the sort of thing I can wrap my hands around.

    Reply
  7. Buy and hold proponents sometimes argue that if you miss the best “X” days in a year, you’ve missed the majority of the year’s upside. But of course they neglect to point out that if you additionally miss the worst “X” you can do much better. If distribution is symmetric (I’m guessing the tail is fatter on downside, but just guessing) than it’s a red herring. What you’re really doing is simply describing the tails of return risk you take when taking on market exposure.

    John Hussman has done some interesting work suggesting that when the market is overvalued, overbought, and overbullish (he contends it is all those things right now) you’re much more likely to get one or more of those X-worst than X-best days. What I like about his approach is that he’s ruthlessly data driven and refuses to accept simple platitudes in his portfolio strategy. (Full disclosure – I invest in his fund.)

    Reply
  8. Here’s a recent article by Hulbert on Fosback’s month-end seasonality timing system:

    http://tinyurl.com/3dos26

    Hulbert calls it “The best performing stock market timing system by far over the past 25 years, among any that the Hulbert Financial Digest has followed”

    Reply
  9. There’s an interesting article from the “Social Science Research Network” (SSRN) entitled, “The Halloween Effect in US Sectors” Written by two college professors from Massey University, Auckland, New Zealand. It’s well worth the effort of signing up at SSRN to read their research.

    The gist of it is, America is the least seasonally differentiated of the worlds major stock markets, according to their study. They also break down seasonality by sector. They have this to say about domestic sectors:

    “If we take a closer look at which sectors show a strong summer effect or a strong winter effect, a pattern becomes apparent. One might say that sectors with no strong Halloween effect tend to be defensive consumer oriented sectors, generally related to products which have short life spans (Food, Consumer and Utilities). Sectors with a strong Halloween effect seem to be pro cyclical related to the raw material and production economy (Construction, Steel and Machine). However, in both cases there are exceptions. The Clothes sector shows strong winter effect, whereas the Oil sector has no significant effect.”

    There’s also a graph of different nations and how they perform during the summer and winter months. The US, Hong Kong, South Africa and Denmark are the least differentiated between Summer and Winter. Italy, Singapore, France, Austria and Japan are the most differentiated.

    Check it out. You may want to trade ETFs of other countries, instead of gaming the American markets. But if you do trade America, you might want to game by sector.

    Here is the link:

    http://papers.ssrn.com/sol3/papers.cfm?abstract_id=901088&high=%20halloween

    Reply
  10. Thanks everyone for your input.

    When I posed this question I was somewhat wondering about this year specifically. Another thing to consider in this day and age is the the added risk of terrorism, that I believe is more prone to happen on our soil (the US) during the summer months due to more people being out and about, and therefore more casualties to tally.

    So it may be somewhat advantageous to also consider that x-factor as well for the buy and hold folks.

    Reply

Submit a Comment

Your email address will not be published.

WP-SpamFree by Pole Position Marketing

Sell In May?

I received the following question;

What’s your current thinking this year to the old adage: “Sell in May and go away”? (for the summer that is, for the folks not familiar with the saying)

I’m in a rather large 40% cash position at the present time.

I don’t really act on these sorts of things in a meaningful way for a couple of reasons; first is that there may not a enough of a fundamental case to do it. The second reason is that I think I heard recently that it only works 54% of the time. I may have that wrong.

I can also recall, however, another study that showed the market does much better from November until May by a dramatic amount. The nature of this second study I am vaguely remembering focuses more on better returns during the winter as opposed to declines in the summer. Anyone with specifics on either is encouraged to leave the links.

Obviously in 2006 sell in May was 180 degrees wrong. In looking at the Stock Trader’s Almanac for the last ten years from May to October inclusive sell in May was also wrong in 2005, 2003, 1999 and 1997. Selling in May looks like it was right in 2002, 2001, 2000 and 1998 and I am going to say it was a push in 2004.

I did not look further back due to time constraints and I would not really be able to defend an argument that says the time sampled is flawed, maybe because it includes the bubble and the aftermath. Hopefully you have a Stock Trader’s Almanac and you can study it further if you are so inclined.

The way I might incorporate sell in May into my thinking might be to sell one stock for a little cash if I thought the market was extended. I currently have a stop order for a portion of the position for one across the board holding that is up a lot. As the market has had a big run if the stop order gets elected I don’t think I would redeploy that cash. There are also one or two more other names that are up a lot that I have been mulling partial stop orders for.

Having three stop orders as described above all execute in a downturn seems compelling to me. It would probably raise 3-4% in cash, not a big bet but a way of letting the market reduce your exposure.

Big bet. I think that is what the reader has made by having a 40% cash position, unless of course the other 60% is in double long funds. The dilemma with such a big cash position is knowing how to get back in. I would not know what to do with such a big cash position taken for tactical reasons.

I have been bearish for nine or ten months. I have chronicled many times my skeptically going along for the ride so clearly I have no useful input for anyone with such a large tactical cash balance.

The thing that should be feared is down a lot. While that is the case, the thing to consider is that down a lot doesn’t happen very often but when it does happen it gives plenty of time to get out. I would urge anyone managing their own portfolio to learn the history of this. I would add that down a little, like late February, should not be feared nor traded around by most folks.

The obvious question that would arise from that last paragraph is how do you know that a decline, like the one that started on February 27, won’t be big? Well you can’t know but you can put the odds in your favor. Fast moves down that create fear usually come back fast. Slow moves that don’t bother people have historically been more of a problem for the market.

Submit a Comment

Your email address will not be published.

WP-SpamFree by Pole Position Marketing