Hold off to add to equities until October? Some data...

[Lsbcal]

I want to buy more equities in the next few months. But I've heard that September isn't a great month historically. So I did a little looking at the September and October numbers. Below is the table for 1950 to present, 70 years of data. This is for the SP500 with dividends.

You can see that indeed September has not been all that good. For all Septembers the results are about 50/50 for gains versus losses. On the other hand Octobers have been better with 65% of those months showing gains.

Looking at just election years the numbers are more dramatic. Only 28% of September's showed gains whereas fully 71% of October's showed gains.

I don't know if this is just chance or what. But maybe I will hold off to invest until October.

Thoughts?

 

[RobbieB]

Well then you should buy right at the end of a losing month.

 

[ut2sua]

I want to buy more equities in the next few months. But I've heard that September isn't a great month historically. So I did a little looking at the September and October numbers. Below is the table for 1950 to present, 70 years of data. This is for the SP500 with dividends.

You can see that indeed September has not been all that good. For all Septembers the results are about 50/50 for gains versus losses. On the other hand Octobers have been better with 65% of those months showing gains.

Looking at just election years the numbers are more dramatic. Only 28% of September's showed gains whereas fully 71% of October's showed gains.

I don't know if this is just chance or what. But maybe I will hold off to invest until October.

Thoughts?

Past performance does not gurantee future results.
Since no one knows what will happen this Sept, you have 50% chance to be right. It is best to try to find your correct AA and stay the course. There will be winners and losers in every move (whether it is a Sept or Oct or Jan move), long term market gain is what we can all hope for (and even that is not guarantee, but it is the most likely scenario)

 

[atmsmshr]

Wasn't it Mark Twain who said, "Never buy stocks in any month ending with 'r', avoid buying stocks in a months ending in 'y', and be particularly concerned with any purchases for months ending with an 'e, l, t, or h.'?

 

[target2019]

I want to buy more equities in the next few months. But I've heard that September isn't a great month historically. So I did a little looking at the September and October numbers. Below is the table for 1950 to present, 70 years of data. This is for the SP500 with dividends.

You can see that indeed September has not been all that good. For all Septembers the results are about 50/50 for gains versus losses. On the other hand Octobers have been better with 65% of those months showing gains.

Looking at just election years the numbers are more dramatic. Only 28% of September's showed gains whereas fully 71% of October's showed gains.

I don't know if this is just chance or what. But maybe I will hold off to invest until October.

Thoughts?

I'm waiting for the next V to appear...
If you're just going S&P500 my crystal ball says a pullback is imminent.

Statistically you're correct. But I'm afraid this upcoming season is not gonna be an average result.

 

[cyber888]

Well then you should buy right at the end of a losing month.

True. If September is the bad month, load up and buy on that month if October is much better

 

[Toocold]

Wasn't Black Monday in October? I would avoid October as well. January is good because you have the Christmas effect. Also try to avoid the double, triple, and quadruple witching hour.

 

[Lsbcal]

Note to self, don't bother posting this kind of thing.

 

[MRG]

Wasn't Black Monday in October? I would avoid October as well. January is good because you have the Christmas effect. Also try to avoid the double, triple, and quadruple witching hour.

Black Monday, which one? 1929 or 1987? Both were in October.

 

[OldShooter]

Note to self, don't bother posting this kind of thing.

But it's such fun! You could try posting a few head-and-shoulders charts or maybe a bump-and-run-reversal !

 

[bolt]

Note to self, don't bother posting this kind of thing.

Interesting Hard Data Research by Lsbcal
I vote Lsbcal continues.
Thx, Good luck & Best wishes.

 

[Lsbcal]

But it's such fun! You could try posting a few head-and-shoulders charts or maybe a bump-and-run-reversal !

You seem to have a penchant for snarky comments. How about trying to be a respectful poster.

 

[Whisper66]

I believe you’d have a better chance at picking up good buys by watching for another dip due to the impact of Covid. Those with funds available to invest back in the March dip got some good run ups since then.

 

[Closet_Gamer]

Interesting data. Hard to tell whether its predictive or just randomness that appears to be a predictive pattern. When looking backwards across a complex data set there is always a dispersion of good/bad, up/down, etc.

I don't recall all my stats work, but I think if you really want to assess it, you have to look at smaller periods and see if they forecast future periods within the same larger data set. For example take the average of the first 20 years, and then calculate 30 sets of rolling 20 years averages. If the average remains similar across the rolling data sets, its probably predictive. Some quant jock could tell me to what % its predictive.

If its not similar, it's probably randomness appearing to be a predictive pattern when looked at in retrospect.

Or something like that.

I definitely feel agree that the market is over-bought. Of course, I've thought that since June. I played with some put options for a few months but ultimately decided that the Fed, Congress, the virus, and the news cycle were firmly in charge right now.

So I'm just rolling along with my AA and adding each month.

Good luck.

 

[target2019]

Note to self, don't bother posting this kind of thing.

There are members who read and respect your comments. Please keep that in mind.

 

[DrRoy]

Without clairvoyance, the question is unanswerable. When in doubt, dollar cost average over the 2 months in 3 or 4 batches.

 

[Sanstar]

This is a timely thread for me. I had taken a chunk of $ out of market last summer. In Feb decided to put it back to work only to watch it tank in March. It’s back in the black so I’m planning to take it out on Monday. Then DCA back in US equities over a period of months. With everything on the horizon, I’m hopping to benefit from a few dips along the way.

 

[Toocold]

So here is the intellectual discussion.

The better question is, what variables can be predictive of future returns, either by individual stocks, by sector, by geography (us vs international), or by the whole market. Then looking historically, or just hypothesizing, to find potential correlations and then applying them in the future and see if it's statistically significant.

So within that context, why do you believe that the predictive market variables are 1) months and 2) election cycles? Is there any formal research? Otherwise, you get all these weird variables like the Christmas effect, that people try to justify. Also, why aren't active fund managers using it to justify their expenses by producing a consistent alpha.

Obviously the issue is, if there is research such Fama/French's small value cap theory, once it is known, it may no longer be predictive.

 

[target2019]

The data on monthly returns has been studied. I guess nothing predicts the future, but there is plenty of past data. So the OP's idea is to focus on two variables whcih have past data, and place a wager.

Monthly returns data is covered not-so-deeply right here.
https://stockanalysis.com/average-monthly-stock-returns/

There are studies, with their theory and holes and so on. I'm sure of it.

 

[Toocold]

The data on monthly returns has been studied. I guess nothing predicts the future, but there is plenty of past data. So the OP's idea is to focus on two variables whcih have past data, and place a wager.

Monthly returns data is covered not-so-deeply right here.
https://stockanalysis.com/average-monthly-stock-returns/

There are studies, with their theory and holes and so on. I'm sure of it.

I don't think it has been studied. I think there have been calculations made. A study would take this result and then apply it (vs a market control) and short first of August and Sept, and buy more first of April and October, and then report out on it.

 

[target2019]

The data on monthly returns has been studied. I guess nothing predicts the future, but there is plenty of past data. So the OP's idea is to focus on two variables whcih have past data, and place a wager.

Monthly returns data is covered not-so-deeply right here.
https://stockanalysis.com/average-monthly-stock-returns/

There are studies, with their theory and holes and so on. I'm sure of it.

I don't think it has been studied. I think there have been calculations made. A study would take this result and then apply it (vs a market control) and short first of August and Sept, and buy more first of April and October, and then report out on it.

As I said...
https://scholar.google.com/scholar?hl=en&as_sdt=0,31&q=market+returns+monthly&btnG=

To be clear, I'm not interested in picking sides or supporting one view or another. I enjoy reading about stuff.

 

[OldShooter]

You seem to have a penchant for snarky comments. How about trying to be a respectful poster.

Sorry. I thought I was making a humorous post in the spirit of the previous ones, including yours.

... Obviously the issue is, if there is research such Fama/French's small value cap theory, once it is known, it may no longer be predictive.

... There are studies, with their theory and holes and so on. I'm sure of it.

I don't think it has been studied. I think there have been calculations made. A study would take this result and then apply it (vs a market control) and short first of August and Sept, and buy more first of April and October, and then report out on it.

I don't think there is any question that this idea has been studied, probably hundreds of times. With about 10,000 mutual funds, mostly stock pickers, and a constant flow of PhDs from top schools into the industry, there are probably no unturned rocks. Most fund prospectuses say that their plan is to exploit "market inefficiencies" and "mispricing," like that was some kind of innovative idea.

Regarding "report out," probably not. To the extent there is reporting it will be in academic papers but they are probably a small single-digit digit percentage of the number of proprietary studies done by stock pickers.

@Toocold makes an important point, though. Inefficiencies and mispricings are almost always self-healing. As researchers discover them, the errors are arbitraged away. So if there ever were opportunities like the OP is seeking, they have probably come and gone. (The old joke about the economists and the $20 applies: https://www.barrypopik.com/index.php/new_york_city/entry/if_it_were_a_real_20_bill)

Re Fama, again @Toocold's point is interesting. With the development of the Fama/French three factor model, which says that small cap and value are good investible characteristics, the cat is out of the bag. Fama argues that the advantages of small and value are not mispricings because they have gone on too long, hence they must be permanent. But I am with @Toocold -- why won't the market's self-healing mechanism arbitrage these away too? Witness the number of "factor based" stock picking funds that have appeared. I dunno. Fama has the Nobel.

 

[NW-Bound]

... For all Septembers the results are about 50/50 for gains versus losses. On the other hand Octobers have been better with 65% of those months showing gains.

Looking at just election years the numbers are more dramatic. Only 28% of September's showed gains whereas fully 71% of October's showed gains.

I don't know if this is just chance or what. But maybe I will hold off to invest until October.

Thanks for posting this. The data is interesting, but can it be explained away by mere chance? Can you show a histogram of Sep and Oct monthly returns, all years combined and just the election years?

What I am thinking is that there might be some particularly bad years that skewed the average for October, and these bad years were caused by some events that had nothing to do with the election.

 

[Fermion]

The market will keep going up until Tesla hits $2100. The reason for this is that after the 5:1 split it will be 420, which seems to be the answer to everything.

 

[Lsbcal]

Thanks for posting this. The data is interesting, but can it be explained away by mere chance? Can you show a histogram of Sep and Oct monthly returns, all years combined and just the election years?

What I am thinking is that there might be some particularly bad years that skewed the average for October, and these bad years were caused by some events that had nothing to do with the election.

Sorry for the delay as we are having fires near here in Northern California and I'm a little distracted.

Well I was thinking of doing the histogram too. So I started into this and discovered an error which greatly changed my election year results. So here is the revised data which includes the month of November too. Please ignore the OP data, sorry.

You can see that for election years the months of September, October, and November have similar winning and loosing number of months! Interesting that in election years since 1984 (last half of election year data) the September returns have a lot more gaining months then loosing months. Note the 2008 September and October results can affect the data as these months were a -24% loss.

The averages look like nothing special. No "signals" in this data to make a bet one way or the other, I think. This now is not really all that surprising to me but I just think the exercise is interesting anyway.