Joined Nov 11, 2007
1,458 Blog Posts

Developing an Edge from Window Dressing

The following article is lifted from Trading Markets. I think it is very interesting and could definitely provide an edge.

Here is the link: When Do Money Managers Like to Buy Stocks? 

Do money managers pile into the stocks at the end of the month in anticipation of 401k and savings money coming in? Well it looks like they do…big time!

In 1999, Kevin Haggerty wrote about this phenomenon on the TradingMarkets site. Kevin had just retired as the head of trading for Fidelity Capital Markets, and he discussed the “end-of-the month” phenomena. Up until that time, I had not heard about it nor had seen anyone else write about this.

Since Kevin first mentioned it, I and others have observed it time after time, especially when a stock was rising (meaning above its 200-day moving average).

Recently we decided to look further into this. We decided to see if we could quantify and potentially profit from the behavior. The findings were a bit eye-opening, and I’ll share them with you today.

We looked at over seven million trades going back to January 1995*. We then broke these trades down to the day of the month, meaning we asked “if we bought every stock today, how would we have done over the next 5 trading days?”

We did this for every day of the month for every stock that was trading in a longer-term uptrend, meaning above its 200-day moving average.

Here are the some notes of our findings:

  1. The average gain for all days (meaning you randomly bought a stock and held it for five days) when it was above its 200-day MA from January 1995-September 2006 was 0.27%.
  2. Now, let’s look at what happens as we start approaching month end. The average gains rise…significantly. On the 23rd day of the month, the average gain for these stocks more than doubles. On the 24th it triples. On the 25th it almost quadruples! And this type of behavior holds through the end of the month (see table below).
  3. Let’s go further. Let’s look what happens at the early part of the next month. In spite of a strong upward move in stock prices over the past 11 3/4 years, stocks on average have “lost money” for the six consecutive days from the 3rd through the 8th. Money spent is money spent, and it looks like the fund managers spent their money near the end of the previous month.
  4. And now let’s go even further and look at something which is even more eye-opening. Let’s look at what happens after a stock drops the previous day. The returns for those stocks go up even more near months-end. And if the stock has dropped two days in a row, the gains become extreme. On the 25th and the 26th, the stocks which have dropped two days in a row have risen more than 1.5% on average over the next 5 trading days. Annualize that out. One can make a heck of a living with returns like this.
  5. More to number 4. After a stock has dropped two days in a row going into the 25th, 26th and 27th day of the month, more than 60% of the stocks have closed higher five trading days later. Markets are efficient…uh huh.

Kevin was right. And considering his position in the trading world, he certainly had the opportunity of seeing this behavior first-hand. Over the past 11 3/4 years, stocks have been much stronger near months-end than they have been any other time of the month. And, stocks which have dropped a day or two near the end of the month, have out-performed even more (think about what happened this past Tuesday morning after many stocks dropped two days in a row). The fund managers like to buy near months end. And it looks like they especially like to buy stocks as cheap as possible near months end.

Baseline Days of the Month, Above 200-Day MA, 5-Day HoldDays which have outperformed by a 2:1 margin or greater are in green.

1 or More Down Closes, Above 200-Day MA, 5-Day HoldDays which have outperformed by a 2:1 margin or greater are in green.

2 or More Down Closes, Above 200-Day MA, 5-Day HoldDays which have outperformed by a 2:1 margin or greater are in green.

This research can go much further. The size of the drop, ETF and futures trading, short selling, pairs trading, combining PowerRatings with the day of the month, further intra-day drops, etc. are all there for a deeper look.

In the coming weeks I’ll share some of this research with you here on the TradingMarkets site.

Have a great week trading!

Larry Connors

* Our research looked at 7,050,517 trades since Jan 1, 1995. We applied a price and liquidity filter that required all stocks be priced above $5 and have a 100-day moving average of volume greater than 250,000 shares.

Larry Connors is CEO and Founder of TradingMarkets.com.

If you enjoy the content at iBankCoin, please follow us on Twitter


  1. MarketRaider

    Thats kick ass Woody, great info even if it was lifted! Now let’s see if I can put it to practice.

    • 0
    • 0
    • 0 Deem this to be "Fake News"
  2. colonic


    Thanks for posting this article. I have in my “archives” an article discussing a similar strategy, called “Seasonality and Trading System”, created by Norman Fosback in the early 1970s.

    Here’s the URL, but I’m not sure if it’s still good:


    Anyway, here are some excerpts:

    “Of course, advisers over the years have claimed to discover numerous seasonal patterns in the market. Many of these “discoveries” were nothing more than unscrupulous data mining and predictably failed the moment that they were followed in real time.
    And even some patterns that had genuine statistical significance had little investment significance, as they were not profitable to follow when you factored in transaction costs.
    But one pattern that has shown incredible significance, both statistically and financially, is the Seasonality Trading System created by Norman Fosback in the early 1970s.”

    “That system attempts to exploit two notable market behavioral patterns — abnormal strength around the turns of the month and right before exchange holidays. More specifically, it recommends being in the market for the last two trading days of each month and the first five of the subsequent month, as well as for the two trading days before exchange holidays. At all other times investors remain in cash.”

    “But what makes this timing system’s performance so impressive is not that it beat a buy-and-hold strategy by 1.4 percentage points on an annualized basis. Instead, the system lets investors keep pace with the market even though they’re in cash more than half the time.
    This is why on a risk-adjusted basis this system is far and away the best-performing market timing system of any that I track at the Hulbert Financial Digest.
    Over the last 20 years, for example, the performance of the second-best market timing system (from Richard Russell, editor of Dow Theory Letters) on a risk-adjusted basis is barely half that of this seasonality timing system.”

    “But before you rush out to your broker and change your entire investment strategy, you should be aware of at least four potential Achilles’ heels that could plague this timing strategy in the future.
    First, the strategy involves a lot of transactions — around 17 round trips a year, in fact. So you need to pursue it using investment vehicles that have little or no transaction costs.
    This probably means using one of the handful of no-load index funds that levy no additional fees for frequent switching, such as those offered by the Rydex, Potomac, and ProFunds families of funds.
    The strategy’s second vulnerability, related to the first, is that it generates lots of short-term capital gains. So, ideally you should follow this strategy in a tax-deferred account. The strategy’s risk-adjusted returns are still attractive on an after-tax basis, but a lot less so than on a tax-deferred basis.
    Third, the strategy may stop working once enough investors follow it. Success eventually will undermine all profitable strategies, of course, and there is no reason to expect this one to be an exception.
    Finally, the pattern may be a statistical fluke. I point this out not because I believe it’s true, but because you should always ask whether what has happened in the past is a matter of luck or coincidence.
    One good defense against this possibility is to submit the pattern to a reality check. Can we tell a plausible story for why it should exist? If not, then we should be more skeptical of its continued success in the future.
    Fortunately, such explanations appear to exist for the month-end and pre-holiday seasonality patterns. People typically put retirement plan contributions into the market at the end or the beginning of each month.
    And traders tend to cover short positions right before exchange holidays, since they don’t want to be short during an extended period when the markets are closed.
    But the bottom line is, over a 20-year period the system works, which is good enough for most investors.”

    Thanks again.

    • 0
    • 0
    • 0 Deem this to be "Fake News"
  3. Gunners

    vewy interesting

    • 0
    • 0
    • 0 Deem this to be "Fake News"
  4. BPOE

    Great Article:

    With this ression you have to see if the good old USA become savers again or go broke. This will change money managers game plan. If they go broke or don’t save money managers will follow some other money inflows. China 401k plans ETC ?

    • 0
    • 0
    • 0 Deem this to be "Fake News"
  5. BudFox

    I did a quick stock screen based on these parameters. Here were the top volume picks:


    • 0
    • 0
    • 0 Deem this to be "Fake News"
  6. yo

    I think in some weisenstein book they have a chart that has about the same sort of data…
    It also shows days of the week. I think Monday is traditionally down, Tusday is relatively flat, Wednesday is the best day, and I don’t remember, but it would be interesting to check. I think the strategy is buy on monday or tuesday, sell on thursday or friday.

    • 0
    • 0
    • 0 Deem this to be "Fake News"
  7. astrologer

    Good post. You guys are noticing a trend in the time of the month, but originally the gregorian calander was made based on astrology… 12 months, 12 signs in the zodiac, 30 days in a month, 30 days in the moon cycle, 7 days of the week, 7 known planets at that time. It should come as no suprize that there are monthly trends. The monthly trend is a trend based on the moon. That data agrees with mine. Observe.


    Mars has an even stronger alpha, but I can’t find the site where I found that info.
    Buy just before a new moon, sell quarter moon.
    I guess maybe a “time machine” would let you look at where the planets are and compare to the charts, and thus, predict the future. Just as the moon effects the tide, it effects our emotions, our bodies being made out of 75% water. Plus, planets emit electromagnetic waves, as does our own. Either way you can test the numbers and notice that patterns do exist, and at the very least, planets are a good reference points to “real” time and historical patterns, because afterall, Einstein showed that time is all relative to how fast we’re moving in comparrison to others.

    • 0
    • 0
    • 0 Deem this to be "Fake News"