Historical Pattern: Stock Prices Increase on the First Day of the Month

Are there certain specific days that are better than others, when it comes to investing? It just might be that there’s historical evidence of a pattern of  better stock market returns on certain days than others.

As regular readers might notice, I occasionally like to dive into the data to find trends about stock market returns.  While I generally focus on more of a passive, index fund strategy, I do think that it’s also interesting to find patterns that might give us an edge if we choose to be a bit more active.  This way, we may have something to tip the odds in our favor a bit.

A prior analysis I did on monthly stock market returns showed how there are some historical patterns that seem to indicate better stock performance in some months than others. A subsequent analysis on the September effect delved further into that particular month, indicating that September is historically poor performing relative to others.  The premise held up the very next month after the post, as the market plummeted by over 7% in that 30 day period.

This got me thinking – if we could go beyond identifying monthly stock returns, and instead looked at daily stock returns, could we find any discernible patterns? In other words, are there certain days that provide better returns than others?

As I did before, I pulled raw historical stock market data from the S&P 500 index. This time, I just looked back to 2000, but I changed the data points to daily. I had to set up the spreadsheet with a variety of formulas, so it took a bit of time to convert the raw data into a place where I could generate some insights.

But when I got done, I found something interesting:  Stocks tend to go up, higher than they otherwise would, on the first trading day of the month.

Essentially, looking back since 2000, it’s clear that there have been more first days of the month that went up versus those than went down. So, from January 2000 to November 2011, the breakout was as follows:

  • “Up” first trading days of the month: 90
  • “Down” first trading days of the month: 52.

Interesting. It became clear that first trading days of the month tend to have stock price increases, more so than stock price declines. But what about the actual returns themselves?

I broke out the different types of classifications of trading days and totaled them as follows:

  • First trading days of the month: 143
  • Other trading days of the month: 2,837

Then, I did a simple calculation of taking the average daily return (up or down), summed them, and divided by the total number of days in each respective category. Granted, this isn’t a weighted average, but a direct simple sum and divide approach. Here are the results:

  • First trading days of the month: 0.192% average daily return
  • Other trading days of the month: -0.005% average daily return

Now, this was even more interesting. The first trading days of the month yielded an impressive average of a 0.19% daily return. If that seems small, remember that it’s for an average day. Just a day. By projecting out over a larger number of days, we can see that getting returns like that is pretty good! All the while, when removing these first days from the mix, and looking at all other days – we can see that the market fell short of breaking even.

Implications: It seems like while there’s certainly variability in the data, the first day of the trading month tends to outperform the average of the other days. Thus, perhaps there should be consideration given to this data when deciding when to buy or sell market baskets of stocks. Maybe buying toward the end of the month makes more sense, and selling soon after the first few days of the month makes sense.

I’m not completely sure why this phenomenon takes place, but it could be because there are more institutional investors putting money in the market at the beginning of the month, 401(k) plans putting in money at predetermined times, or maybe inherent optimism at the beginning of a new time period – much like people get fired up with resolutions at the beginning of a new year. Who knows?

It turns out that some others have discovered this concept of the first day of the month effect too, as I subsequently found out after doing this analysis. Well, that’s ok  :)  Of course, I still think that there’s a potential opportunity here that’s really been flying a bit under the radar. This isn’t discussed too often in personal finance circles. 

My Questions for You: 

What do you think of this effect? Is it something you’ve considered when making buying/selling decisions?

Why do you think that this clear pattern exists? There must be a logical explanation, right?

Comments

  1. PKamp3 says

    You should have kept it to yourself, haha. If there is a historic trade-able pattern, once it becomes public it can get arbitraged away by people betting with the expectation it can hold up. (Think “Sell in May and Stay Away”, or the Price to Sales Ratio – both have weakened as investors have adopted them.).

    If I had to throw theories out there, I would guess it has something to do with automatic investing. Perhaps on the 1st and the 15th of every month people are automatically dumping money into the stock market. It might be weighted towards the start of new quarters, so mutual funds are making their purchases (then they can easily sell the stock before they have to report it at the end of the quarter). It could be index funds re-balancing as well.

    To MoneyCone above – I bet a lot of that is due to options expiration day, third Friday of the month. Google ‘option pain’ for an interesting theory on that, haha.

    Nice analysis!

    • Squirrelers says

      PKamp3 – I think you have some really plausible explanations as to the source of the effect. It seems like this is a truly underreported phenomenon by the mainstream media, probably no secret whatsoever with serious traders one would think. Glad you liked the analysis, time well invested then!

  2. Nicole says

    These are fun. We had a bunch of these in one of my grad classes. I should have paid more attention then… All I really remember is to try to invest in Sept if you’re going to lump sum invest once a year.

    • Squirrelers says

      Nicole – yeah, these are fun. It took me an analysis of my own to figure out the September issue. Over time and on average, that’s the month with the worst stock returns and actually stock declines as of the last 40 or so years. While there’s variability of course, a pattern of 40+years seems like a decent amount of data points to substantiate it. So, I can get you idea of buying in Septemeber if you had one puchase to make.

  3. cashflowmantra says

    I hadn’t heard about the first of the month, but had about Mondays and Fridays. I don’t really consider it when making buy or sell decisions. It is nice to look at from a market perspective, but individual stocks vary enough during the day that it probably doesn’t matter as much.

    • Squirrelers says

      Cashflowmantra – I think this phenomenon is worth considering, particulary for the bigger investors. While that’s not me, I think that there might be some applicability here with smaller investors. For example, if I going to make investments automatically, perhaps making them right at the end of the month could give some slight benefit over time?

  4. Cat says

    I would not be surprised if there aren’t a lot of conversations like this at the horse races also. I feel the stock market to be the same toss up these days.

    • Squirrelers says

      Tushar – I disagree as far as that being a blanket statement for all passive investors. Really, passive investors in some cases can choose when they make automatic investments. It could be worth considering in those cases for passive investors.

    • Squirrelers says

      Andy – I don’t agree. I’m willing to say that there are probably investors much bigger than us that take patterns like this into account. That said, I can absolutely see how it would be hard for many people to take advantage of this. Some can, some can’t.

  5. 101 Centavos says

    I’d have to devote a ton of time in order to capture the returns. In the end, I’m not sure that the return on the small amount of risk capital would be worth. Very interesting analysis, though. I wonder how the black-box trading computers in the backrooms of brokerage houses take advantage of this.

    • Squirrelers says

      101 – Good point on the brokerage houses. I have to think that historical patterns like this are considered by such larger, or institutional investors. For smaller investors, I tend to agree with you that it would be a tradeoff of time in order to capture returns, and mignt not be worth it for everybody. But for some, it might – particularly for larger investors. For smaller investors, it might help in terms of timing for automatic investments. For example – if someone is going to automatically put money into a 529 account on the same day every month, it might help to set the deposit date at a time just before the fist of the month. This way, there’s little effort, and perhaps an opportunity to take advantage of these patterns – even if in a small way.

  6. JP @ Novel Investor says

    Have to agree with 401k, IRA money flowing into the market as a partial cause. There’s also the possibility of dividend reinvestment and any bond interest payments being put to work. you might be able to attribute quarterly rebalancing by mutual funds and hedge funds 4 times a year. The market tends to get a bit choppy on the last day of every quarter, with funds taking profits or wanting to “show” better performing stocks in their holdings. Good catch.

    • Squirrelers says

      JP – good points on dividend reinvestment/bond intererest payments. With 401k/IRA money a potential contributor, as well as these other factors, it’s interesting how there appears to be some market inefficiency here.

  7. Nick says

    In looking at your monthly and daily data, I asked why? I have been doing this for years.

    Institutional investors are now a big player in the stock market. They are largely investing the retirement contributions of individuals and have to invest them according to their investors’ wishes – based on their fund selections.

    I think you are absolutely right that the rise at the beginning of the month is due to the need for institutionally investors (401k, pension fund managers) to invest individual contributions. Much of this money goes towards buying stocks. The money flows in a the beginning of the month and the institutions will then invest it (buy stocks and bonds). This increases demand for stocks in general and thus causes market prices to spike. Also, there may be a need for many investors to sell at the end of the month to pay bills such as a payroll. This may explain the decline in prices later in the month.

    Have you investigated the bond market as well? My sense is that you may find that average bond prices also go up at the beginning of the month and perhaps perform poorly towards the end of the month.

    The monthly data is harder to explain. The January thru April rise may be due to individual investors IRA contributions which they often make between the beginning of the year and April 15. Holiday bonuses also may contribute to the demand this time of the year.

    Why is September such a horrible month for stocks? And why are the summer months (June through August) also not so good for stocks?

  8. Chuck says

    I like your stat work! You clearly did a lot of research into this trend. That being said, I never use these types of historical calendar trends to influence my trading decisions.

    I guess if you trade index ETFs or futures, there’s value in knowing calendar and seasonal trends, but personally, I prefer to make buy and decisions based on fundamental and technical analysis of each individual security.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>