The First Day of the Month Trade - Why Am I Still Working

Posted by: Howard Silverblatt on May 4, 2010

I continue to get significant inquires into the First Trade of the Month portfolio. The calculation is basic - you are in the S&P 500 for the first trading day of the month (close-to-close change) and are credited with only the stock price change, all proceeds are reinvested each month, and there is no interest given for the time out of the market.

The beliefs range from ‘just a bazaar stat’ to that additional funds are invested into the market on the first day and therefore create upward pressure which accounts for the over performance.

I have three items of note, however. First in a constant Bull rally, even if the one-day trade is better than most, it can not make up for the other 20 trading days in aggregate. Equally, the same is true of a Bear market, even if the one-day is down, it should not, in aggregate, be worse that the remaining 20 days in aggregate. So, based on some soft research, the higher the directional consistency and volatility, the wider the spreads between the first day of trade and the overall time span. Great since 12/1999 (28.62% vs. -18.17%), but not so good since the Bull rally started in 3/2009 (3.44% vs. 45.57%) or 4/2009 (8.49% vs. 63.55%). Second, I ran the same portfolio based on the last day of the month trade and the result was a gain of 2.26%. Better than the overall index, but considerably less than the 29% first day trade results. And third, outside of my first point about consistency and volatility, I can’t really explain it. But, do I really need to know why it works when it continues to work so well. It’s hard to argue with a box which has produced a 29% return, verses the markets 18% decline, but it’s ridiculous to invest in something you don’t understand. Hopefully some academic will go into it more (sounds like a global dissertation if anyone is ABD), or eventually I will - of course, at that point it will cease working.

Deatials in file
first_day_20100503.doc

Reader Comments

azure

May 19, 2010 10:03 AM

Interesting article Howard but please learn to spell. You seem to have meant 'bizarre' as in weird not 'bazaar' as in shop. I'm not going to comment on your 'details in file' ...its clearly a typo.
OK, I just did. Seriously, correct spelling adds credibility to your story.


My responce: I am guilty as charged. English is my second language, numbers are my first; I'll try and do better.

Charles Hodges

May 22, 2010 8:55 AM

You are looking at a subset of the "turn of the month effect." From Ariel (1987) and Lakonishok
and Smidt (1988), it is the fact that U.S. equity returns are unusually high over the four-day period that begins with the last trading day of the month and ends with the third trading day of the following month. If you can trade mutual funds with no transaction costs, such as your 401-k, the effect works with lower risks and higher returns thatn pure buy and hold. If you have transaction costs and taxes, such as whne using ETFS in a taxable account, the returns are as you state. Good in a bear market and bad in a bull market, mildly postive in a sideways market.

Alem

April 7, 2011 4:42 PM

However the percentage change of first trading day of the month and the last trading day of the month for major indexes have negative correlation coefficient, not very big but negative. is this consistent with Charles comment?

Post a comment

 

About

Bloomberg Businessweek’s Ben Steverman focuses on the latest moves in financial markets and emerging trends in stocks, bonds, and funds, always with an eye toward giving readers a better understanding of the sometimes confusing and often chaotic world of money. Standard & Poor’s senior index analyst Howard Silverblatt will also provide his take on companies’ finances and the markets. Voted one of the “Top 100 Finance Blogs” in 2007.

BW Mall - Sponsored Links

Buy a link now!