Dec 31...the last day of the year saw some interesting effects.  For one, stocks of large capitalization best resisted losses...the 1st, 2nd, and 3rd deciles of this group all made our list of today's top gainers.  While our own non-weighted grouping of all stocks was down by .8% today, most major indices (the Dow, Nasdaq, etc.) were roughly unchanged for the day, masking this split between small caps and big caps.  Cheap stocks and stocks with losses over the last three months were relatively strong.

On the losing side, stocks with a large number of institutional holders were weak.  Perhaps that shouldn't be surprising...these are the stocks that are subject to window dressing and tax-related selling.

Yearlong losers were a tad stronger than the market, while yearlong gainers were weak.  Oils were weak.  

**********************

We've crunched the data for the month of December.  The average stock was up about 2%.

Oil stocks led the pack.  They were up about 11% for the month.  Buying cheaply-priced oils would have pumped the gains up over 15%.  The gains were particularly significant when adjusted for risk.  Not far behind, mining stocks quietly racked up some nice gains as well.

Cheap stocks and yearlong losers were strong, while yearlong gainers were weak.  This is definitely counter-seasonal behavior and leads one to wonder if these groups might be stealing some thunder from the "January effect".  

Non-volatile stocks slightly outperformed the general market, meaning you could have simultaneously beaten the market and minimized risk this month (as with quite a few other Decembers).

Semiconductors were weak  Historically, January has been a good month for this group.

Data from the second half of the month mostly paralleled that from the full month.  The only surprise to speak of was that certain retail groups actually fared pretty well on a risk-adjusted basis in the second half.  Banks and REIT's were weak during this positive 2 week period (up about 4%).

****************

For the quarter...metals and mining stocks were up around 30% (the quarter itself gained around 15%).  Cheap stocks in general performed well, as well as stocks that suffered in the closing days of September.

On the losing side, non-volatile stocks were weak.  Biotechs and entertainment-related stocks were weak.  Expensive stocks performed poorly.

Dec 30...looking at our daily data tables, it's interesting to note that stocks that fared well on this day in the previous 3 years also fared well today.  Not only that, out of nearly 3000 possible days that could have been chosen as matches, our computer decided that 12/29/95, 12/31/96, and 12/31/97 were the best matches to today's session (as well as 6/26/95 and 4/4/00).  Hopefully, January will unfold equally mechanically and we can take advantage of these parallels.

Stocks trading well below their 20 day averages were strong today.  Cheap stocks fared well.  Stocks that got burned on Monday made a comeback today...it's possible that these stocks are destined for nice January gains, assuming the reversal is related to the end of tax-related selling.

Losers were nearly a mirror image of winners...stocks trading well over their recent moving averages got burned.  Yesterday's gainers were today's losers...another interesting reversal to note.

Oils were just a tad weak, though mining and metals were significantly weak.  Yearlong losers outperformed the market, but the group wasn't strong enough to make our tables.  Yearlong gainers paralleled the market.

Tomorrow, we've got a lot of work to do, updating our daily, bi-monthly, monthly, quarterly, and yearly tables.  We look forward to it, though...it'll be interesting to see what emerges.

Dec 29...Momentum was the word of the day...buying short-term or long-term gainers would have resulted in continued gains.  Volatile stocks in general fared well, while non-volatiles were weak...these conditions often signal further upward action.

Metals and mining stocks were strong...2.6% gains for this normally staid group.  Oils performed in line with the general market.

Dec 26...though the gains weren't spectacular, we did see some very significant trends.  Stocks with a high high-close differential on Wednesday scored well above the market averages.  Mining and raw materials stocks performed strongly.  Cheap stocks in general outperformed.

On the short side, the trends aren't so well defined.  Large caps performed weakly, as well as non-volatile stocks in general.

Though not found in our tables, both yearlong gainers and losers outperformed the market.  Oil stocks were just a tad stronger than the market as a whole.

Dec 24...despite the low market volume, there were some trends of note.  In general, the market reacted in opposition to yesterday's session...oils were up, yesterday's losers gained, yesterday's gainers lost.  

Retails were weak...pity the fools who invest in this group in December.

Though the trend is not particularly significant, yearlong losers outperformed the market, while yearlong gainers paralleled the market.

Merry Christmas!

Dec 23...though market volume was predictably light, there were some trends worth mentioning.  Most prominently, oils reversed and underperformed the general market in a significant way.  On the positive side 3 month and yearlong losers gained nicely, particularly those with the most extreme losses.

Stocks of low standard deviation performed poorly.  REITs were weak.  Tech stocks and volatile issues in general were strong.

Despite the gains in yearlong losers, yearlong gainers outperformed the market, though not in a major way.

Given today's positivity, with strength in volatile issues and weakness in non-volatile ones, and the seasonal tendencies toward gains in this time period AND before holidays, we'd expect a positive market tomorrow.

Dec 22...market volume was quite low.  One would assume this fact is somehow tied to the fact that there was little separation between our best and worst performing groups (less than 2%) today, and we observe no particularly significant trends to speak of.

Biotechs were relatively strong.  Yearlong losers were strong, and yearlong gainers rather weak.  Oils were also weak, just barely finishing with gains for the day.

Dec 19...instead of the continuing positivity we guessed about below, we got a market in need of some Viagra.  The separation between the most positive and negative groups was less than 2%, with no trends of major significance to report.  Of course, the Christmas vacation is coming up so volume is low, but today was options expiration day, a time when things are supposed to heat up.  As with last month's triple witching hour, things didn't.

Stocks with increasing volume of late performed well.  For whatever reason, we've noticed that volume statistics seem to play a more important role in December than other months.  Just a casual observation, though.  Semiconductors were weak.

Yearlong gainers were a bit weaker than the general market.  Oils performed in line with the market.

Dec 18...an archetypal positive market...volatile stocks performed best, non-volatiles worst.  Bottom fishing was also featured...stocks with low short-term moving averages prospered.  Positive industries were semiconductors and oils (again!).

Though the group didn't make our tables, yearlong gainers performed well (up about 2.4%).  Losers performed in-line with the market.

Banking stocks and REIT's were weak.

Given today's conditions, we'd say there's a high probability of a positive market tomorrow.  We'd also expect continuing bottom-fishing to produce the strongest gains.

Dec 17...Yesterday's losers and gainers reversed today.  Stocks with increasing short term volume did well.  Retails performed well, reversing their horrid recent performance.  In general, it was a good day for bottom fishing and a poor one for momentum plays.  

Though they didn't make our tables, oils continued to outperform.

Yearlong gainers and losers did not show any special tendency to deviate from the general market.

Cheap stocks performed poorly.

Dec 16...oils came back from a poor showing yesterday.  Thus far into the month, this group has absolutely been the place to put your money.  Stocks with large yearlong gains were again weak, continuing this anti-seasonal behavior.  Computer stocks and retails were again weak.

Dec 15...textbook negativity.  Volatile stocks dropped, non-volatiles resisted losses.  This kind of behavior often predicts further negativity.  Computer stocks were weak. Yearlong gainers underperformed, though not horribly so.

Dec 13 (Saturday)...We've crunched the data for the first half of December.

Oils stood out from the pack with nearly a 10% gain in the 10 days.  The gains were particularly strong when you adjust these stocks for risk.  Mining and raw materials didn't do poorly either.  

To toot our own horn just a bit, scroll down to Dec 4...we were pointing out the [burgeoning] move in oils while others were ignorant of it.

Further down the list of winning groups were cheap stocks in general...one might wonder whether this action is stealing the January effect from January.  We note that the first half of December 1996 and 2001 favored cheap stocks, and the January effect in 1997 and 2002 was not particularly pronounced.

On the losing side, stocks trading well over their predominant resistance levels and longterm gainers got hurt to the tune of 5%.  This is rather anti-seasonal behavior, but hardly unheard of...check out the December first half performance of stocks in 1996 and 2000.  Semiconductors got hurt.  As usual, retail stocks were a lousy investment, particularly after risk-adjustment.

In terms of matching up this period with historic periods, you might want to look at the first half of December 1995 and 1996.  In both these years, stocks performed more in line with the usual December pattern in the second half (yearlong gainers performed well, semiconductors were weak).

Below, we (and others) have speculated on the idea that yearlong gainers with longer term losses might be more vulnerable to losses at this time of the year than those with longer term gains.  An informal, non-scientific look at the charts of yearlong winner/long term-losers and yearlong winners/long-term winners didn't seem to predict which of the two groups would perform the strongest this first half of December.  If anything, we'd say that the stocks with longer term losses tended to outperform those with nice upward plots over the last three years.

Dec 12...oils still aren't ready to give up the ghost.  Today was the sixth day out of seven in which oils have outperformed the general market.  Though tech stocks were mediocre, volatile stocks in general fared well.  They didn't make our tables,  but yearlong gainers outperformed the general market.

On the losing side, big cap stocks got burned, as well as high-priced stocks in general.  The trend towards losses in big caps was quite broad, with the top 3 deciles of market capitalization finding their ways into our "biggest losers" tables. Retails were weak...not surprising at all in this month.

Dec 11...a classic up day...volatile stocks prospered, non-volatile positions suffered.  Often, such conditions predict further bullishness.  Stocks that have taken a beating of late, including yearlong gainers, rebounded.  Oils lagged, breaking a five day streak of outperforming the market.

Dec 10...big, continuing losses of 3+% in the yearlong gainers group, despite much more modest losses in the general market.  Though they didn't make our tables, oils outperformed yet again.

Dec 9...a negative market.  Semiconductors were burned to the tune of a 3+% loss.  Volatile stocks in general suffered, while non-volatile issues best resisted losses, a condition that has been a decent predictor of a continued negative market.  We'll find out soon enough. 

Though the group finished with a slight loss, oils again found themselves amongst the best performers...four straight days in the "top gainer" column for this group.

Though you won't find the group in our tables, yearlong gainers continued to underperform the market, while yearlong losers slightly outperformed....anti-seasonal behavior.

Dec 8...again, oils were strong.  That's three straight days in which oils have topped our tables.  Stocks with good momentum continued to gain.  On the negative side, stocks of high volatility lost...a bit unusual given the fact that the market was up today, but in line with the fact that December tends to reward non-volatile stocks over volatile ones.  Biotechs were hurt.

Dec 6 (Saturday)...here's a link to an article at Forbes predicting a weak January effect.  The argument is that only 10% of stocks are down for the year (our own data shows 15%), therefore tax-loss selling and the subsequent January effect should be minimal.

Looking at the historical data, does the above make sense?  Take a look at the table below:

Yearlong gainers and stock prices were divided into deciles...the second and third columns show results for the stocks with the lowest 10% of yearlong gains and stock prices.

With only 10-15% of stocks down for the year, the best January comparison would be 1996, which did indeed have a weak January effect.  The place to be that month, for what it's worth, was biotechs.  Then you go to 1997 and 1998 and 1994, where stocks that lost big %'s in the closing days of December prospered, though not monstrously.

Obviously, our strongest "January effect" occured in January of 2001.  This was coming off a 2000 in which the majority (56%) of stocks actually gained.  Note also that although yearlong losers outnumbered gainers by a 2:1 ratio in 2002, January 2003 didn't show a strong January effect.  Whatever happens in January, it doesn't happen quite as mechanically as the Forbes article might suggest.

Our best guess:  the Forbes article is probably correct....the January effect won't amaze in 2004.  Nevertheless, we'll be putting our money in cheap stocks, which haven't disappointed yet.  Another promising approach, based on 1994, 96-98 data would be to put your money in stocks that suffered in late December.

Dec 5...again, oils were strong, up 1% while Nasdaq was down 1.5.  Further down the significance ladder, you could have realized small gains by staying with stocks with nice gains in the last couple of days.  

A very significant trend on the negative side was losses in semiconductors.  Again, stocks with high yearlong gains got burned...very anti-seasonal behavior.  We'd assume that this behavior would reverse soon.

Finally, we note one odd trend that reared its head again today...stocks that have performed poorly on this day (or thereabouts) last year also performed poorly today.  It's hard to believe that you could predict stock behavior by looking at the behavior of a stock 252 days ago, but that's what we've seen over the last few days, with strong significance. 

Dec 4...the dominant trend today was toward gains in oils.  You'd have gained around 3% with a basket of oil stocks, while the larger market was a tad negative (the major indices were slightly positive, but they'd become negative if big stocks weren't weighted more heavily than small).  Amazingly, one popular market report didn't even bother to mention the move in oils, focusing instead on tech stocks (Cisco was up $.75!!!).

Stocks trading significantly above predominant resistance levels got burned again today.

Dec 3...yearlong gainers were burned to the tune of 3%+ today.  The market favored non-volatile issues and punished volatile ones, a signal that there's a greater than average probability that tomorrow will be negative as well.

December, of course, normally favors yearlong gainers given the tax ramifications of selling these positions, so we wouldn't expect to see strong selling in the group to continue.  On the other hand, of course, tax selling considerations aren't limited to one year...if your yearlong gainers have actually suffered longer term losses, they could be subject to tax-selling.

Dec 2...though the market was entirely unspectacular, it's interesting to see that several December patterns were evident in our data table for today.  Yearlong gainers fared well in comparison to most other groups and retails fared poorly...just as one would expect.

Biotechs fared well.  Stocks with a minimum of short-term debt fared well, as well as dividend-paying stocks.

Dec1...Stocks that lost on Friday came roaring back today with very strong significance.  Although the November/December transition isn't normally thought of as window-dressing time, one has got to wonder if window dressing has a role in this.

Nov 28...the last trading day of the month.  We saw several interesting effects today.  Stocks trading well over their resistance levels gained well above the average stock, with very strong significance.  Yearlong gainers also fared well.  Historically, these groups fare well in December.  Volatile stocks in general fared well, despite the fact that the general market was rather dull today.

On the negative side, regional banks were weak.  Interestingly, stocks with a high number of different institutional holders were also weak on this final trading day of the month.

****************************

We've got our monthlong and second half of November results.

As with many of the individual daily trading sessions during the month, this November came in without any great spread between the best and worst performing groups, and without any groups that stood out in terms of performance.  The best gains were gotten by focusing on healthcare stocks, cheap stocks, and those with low profit margins.  On a risk-adjusted basis, REIT's fared well, bucking the usual seasonal negativity in the group.  Non-volatile stocks in general fared well.

On the short side, nothing of great significance emerged.  It would have been best to avoid stocks that finished October with a flourish.  At the same time, those stocks with longer term losses also fared poorly.  On a risk-adjusted basis, semiconductors were weak.  Stocks that performed well in the same period over the last 3 years were weak.

In the second half of the month, the only trend worth mentioning was that of stocks with low institutional ownership prospering.  Not horribly significant, though.

In general, seasonality was weak...a number of typical November trends failed to materialize, though given the uninspired nature of the month, losing or gaining a great deal of money would have been a difficult task.

Nov 26...a thoroughly tedious pre-holiday market.  The only trend even worth mentioning would be that of yearlong losers continuing to lose.

Nov 25...looks like a good day for the technical guys...stocks that traded well over their resistance levels gained in a very significant way.  Historically, these groups fare well in December, so maybe we're seeing some early activity in the group.  Other gainers included volatile stocks in general, and those with large book values relative to stock price.  Specific industry groups were not in focus today.

On the negative side, nothing of significance emerges.  Large-caps were weak.

Nov 24...The best place to be in this very positive market was volatile stocks.  The worst...non-volatile stocks.  We take this as an indication, based on historical analysis, that tomorrow will probably be a positive market again.  Other very positive places to be were biotechs, semiconductors, and, interestingly, stocks that performed well 252 days ago.  That last group came in with good statistical significance, meaning it's probably no fluke.

Oil stocks performed poorly.

*******************

We're beginning to wonder if there's any serious correlation between the expectation of macroeconomic announcements, triple witching hour (etc.) and market volatility.  The standard wisdom says there is, yet we've seen any number of volatile, strongly up, or strongly down markets of late with no underlying news, and tedious markets with news.  Today was without any major news, yet we saw a good spread between best and worst performing groups (3.5%) and a very positive day in general.

Nov 21...again, very little distance between the best and worst performing groups in our tables.  So much for the "political considerations" theory below (Nov 20).  In fact, there was news on the macroeconomic front, with the Fed making statements regarding future interest rates.  What's more, today was "triple witching day", a time that is supposedly much more volatile than usual.

The strongest trend was a tendency for stocks that opened on the 20th at levels well below the close on the 19th to perform well...whatever caused these stocks to lose ground in the afterhours on the 19th was the impetus for gains today.

Bottom fishing was also featured today, with stocks that had lost over the last month performing well today.

Nov 20...very little "separation" between the best and worst performing groups in our tables (about a 2% difference between the two).  Perhaps this is the case because the market was driven by political considerations (the bombings in Turkey), rather than the sorts of economic data that would favor certain groups over others.

The most significant observation from our tables is simply that stocks of high volatility lost the most ground today.

Nov 19...yesterday's minor trend toward gains in small caps and losses in large caps came into greater focus today.  Our data shows that there's a good chance this behavior will repeat tomorrow.  Recent losers performed well.

Nov 18...though the market was down again, we broke out of the "textbook" negativity...the biggest losers were not the most volatile issues.  Rather, they were stocks that got burned last Friday, software stocks, and big-caps.  On the positive side, mining stocks did well, as well as those held by a large number of institutions.  There seemed to be a slight theme of favoring issues with good fundamentals.

Given the nature of today's market, we'd say there's a better than average chance that tomorrow will break out of the negativity.

Nov 17...more textbook negative market conditions.  Historically, we've found that days like this, where the most positive stocks are non-volatile and the most negative are volatile, are usually followed by more negativity.  So, longs must hold out for a market that is something of an exception.

Stocks with large gains over the last 3 months were creamed today, losing 3% of their value versus a 1.2% loss for the general market.  Friday's big losers continued to lose.  Nothing of major significance was found on the positive side...financial stocks did OK, as well as stocks with large dividends in general.

Nov 14...a textbook negative market...volatile stocks lost big money, non-volatile stocks resisted losses.  Oils actually gained close to 1%, while REIT's stayed above the water.  Computer stocks of every variety got burned.  Though neither group made our tables, it should be noted that yearlong gainers got burned while yearlong losers resisted losses fairly well.

*******************

Well, we've crunched the data for the first two weeks of the month.  The strategy of choosing stocks with a high high-close differential at the end of October resulted in gains around 4%.  Going with the group with the largest 4% of losses over the last three months resulted in gains of better than 5%...this particular strategy has worked quite nicely on a historical basis as well.

Yearlong gainers got hurt.  The group had some good days during the period, concealing the fact that over the two week haul the group lost a significant amount of ground in comparison to the average stock.  Yearlong losers basically paralleled the market...interesting when you look at the nice gains available by going with 3 month losers, a group that one would expect to overlap a good deal with 1 year losers.

Retails were generally weak.  Semiconductors, a group which has historically performed well in the period, basically paralleled the market, though the ride has been a bit bumpy in the last two weeks.

Looking at risk-adjusted data, REIT's performed better than would ordinarily be expected at this time of year.  Dividend-paying stocks in general did well.  Stocks of low-volatility also gained to a degree greater than what would ordinarily be expected.

Looking to the future, we'd expect yearlong losers to prosper in the next couple of weeks, alongside cheap stocks in general.  Big cap stocks that have performed poorly of late should be avoided.

Nov 13...a good day for bottom-fishing.  Stocks with recent losses gained upwards of 2% today despite an essentially flat general market.  Oils did well.  On the negative side, nothing of great significance emerged, though it's interesting to note that yearlong gainers dropped.  The group has really been topsy-turvy over the last 6 weeks or so.

We've noted that, in general, a good time to go bottom fishing is the day after a strongly positive market (like yesterday's).  It definitely worked today.

Nov 12...an archetypal up market...volatile stocks prospered, non-volatile stocks lagged.  With over 1000 groups examined, the worst-performing group (utilities) was still up about .5%, giving an indication of the broadness of today's advance.  As with yesterday, the market rewarded stocks that have performed well over the last few months (the highest decile of yearlong gainers was up better than 3%).  Semiconductors were up as much as 4%.

It might be worth remembering that one pre-market forecast predicted a "range-bound" market, given a lack of important macroeconomic data to be released.  Wrong!  Hindsight is 20-20, of course, but we believe that a lot of these pre-market predictions are essentially worthless if you "backtest" them.

Nov 11...yearlong winners and volatile stocks continued to lose today.  Some retail groups did nicely in comparison to the market in general.

******************

Here's a link to an article on the "newly discovered" tendency of yearlong winners to continue to gain in December.  I don't quite follow some of the logic regarding the January effect...the idea is that the January effect is "true", but casual investors can't capture it because they would need to buy a huge basket of stocks to realize gains.  To me, the large basket merely means you're less likely to get unlucky (or, for that matter, very lucky) and buy January effect candidates that turn out to be duds (or, January effect candidates that greatly exceed your expectations).  The large basket means you "hone in" on the "true" extent of the January effect, but the large basket shouldn't gain more or less than the average small basket, at least as far as I'm concerned.

Nov 10...a textbook negative market...volatile stocks got burned, non-volatile stocks resisted losses.  Utilities held up well.  Long term winners (those that have gained nicely over 3 months or more) got burned.  Stocks with low profit margins got hurt...a possible reversal of last week's tendency for "junky" stocks to prosper.

Nov 7th...continued upward action in the "junky" stock department, with fundamentally sound low PE ratio stocks getting burned.  A rather murky, directionless day in general, though.

Nov 6th...it looks like some trends reversed today.  We had three month gainers in the loss column, biotechs as winners, and some hints of bottom-fishing.  Our fundamental indicators seem to show a tendency toward buying "junky" stocks...those with negative PE's and low profit margins.

Historically, it's a good time for bottom-fishing...this could be the beginning of a strong period for the group.

Nov 5th...conditions were similar to yesterday's...a mediocre market in general, rewarding volatile stocks most.  Still seeing gains in stocks with long term gains.

Nov 4th...despite the negative tone in the market, you would have done best sticking with volatile stocks.  Tech stocks, therefore, fared well...one market report I saw had them underperforming...that's not what I see.

Nov 3rd..."momentum" is the word for the day.  Semiconductors scored big, biotechs were weak.  Stocks that gained over the last month continued to gain.

Oct 31st...Yearlong gainers got hurt in a significant way today, the last day of the month.  On the long side, nothing of note to report.

**********************

It's the end of the month, so we've got a lot of data to crunch.  In brief, here's the verdict: the general market was up about 7%.  The best place to be this October was semiconductors.  You'd walk away with about an 18% gain that way.  Stocks that got burned on the last day of September were also strong (12%).  Yearlong gainers were strong (10%)...this is rather odd behavior given past October history (the group tends to lose), though perhaps it's not so odd when you consider the fact that we're in an unusual market...one that has racked up overall losses over the last several years, with a new set of tax selling ramifications.  If that sounds fuzzy, it's because we're not pretending to understand what went on with great clarity.  Any thoughts?

Yearlong losers pretty much paralleled the performance of the general market.

On a risk-adjusted basis, banking stocks performed well, particularly in the second half of the month.

On the losing side, stocks with large dividends were the most significant losers.  That's right...we just started incorporating some fundamental data into our analysis.  For the time being, though, we don't have historical fundamental data, so we can't ascertain whether avoiding dividend paying stocks is a good strategy for October in general.  

As expected, REIT's were lousy performers.  There's a good deal of overlap between REIT's and dividend-paying stocks.  This is a subject to investigate further...if you exclude REIT's with low dividends, how would the data look?

Other mediocre groups were biotechs, oils, and utilities.  The latter two groups, of course, are also associated with dividends.  

On the whole, you'd probably have to agree that stocks did a pretty nice job of performing in line with historical seasonal trends.  Check out the data for yourself...this October's performance is tabulated at the top of the page, while the most general historical results are at the bottom.

It will be interesting to see how biotechs perform in November.  We've noted a give/take relationship between semiconductors and biotechs in the past.  

Oct 30th...Though they didn't make our tables, the highest 10% of yearlong gainers underperformed (-.4%) today.  We point this out simply because we've been focusing on this group a lot in recent days.  On the whole, though, the day was bland...no groups stood out from the pack in terms of big gains or losses.

Oct 29th...Yearlong gainers were strong yet again.  This has really been the group to watch this month, not any particular industry.  The group performed well earlier this month, got burned last week, and has looked great this week.  We're guessing on another reversal next week...a mini January effect at the beginning of November.  

Oct 28th...Again, yearlong gainers were strong, with three month and one month winners just behind.  Nice gains in the semiconductors (as much as 6%) as well.  Yearlong losers performed in line with the market.  No big surprises in the losing columns, except that on such a positive day, REIT's still would have lost you money.  This is in line with the expected seasonal weakness of the group.

********************

We've been toying with some of our daily data.  The question we're asking is, "if group X is today's strongest group, what group is likely to be tomorrow's strongest group?"  After crunching the data, we came up with some interesting results.

We've never been fans of naive "momentum" strategies.  The data just doesn't show that simply buying stocks that are going up tends to produce benefits.  What we're finding though, is buying the right group that is going up could definitely be of benefit.  If cheaply-priced stocks prospered yesterday, they're likely to do so again today.  If stocks with low standard deviation prospered, they're likely to do so again.

If you're an experienced momentum guy and are reading this, you're probably saying "duh", but it's nice to see this sort of thing validated by the data.

We'll be publishing some of the results in a few days.  Thinking ahead, I have a feeling that while the strategy of simply repeating the purchase of today's best group probably  produces profits above the general market, it can reverse in a hurry.  This is where stop-losses might come in (another popular strategy that, taken in its naive form, we don't see much purpose to).

Oct 27th...The behavior of yearlong gainers reversed again today...they performed quite well.  Yearlong losers and large caps performed poorly.

Oct 25th...Below (the 22nd), I wrote about how seasonality and charting could relate.  Again, I'm not much of a technical guy.  On the other hand, to the extent that tax-loss selling and window-dressing influence the markets, the charts should be quite revealing on the amount of tax loss selling one might expect, and whether these losses are short term or long term or both.

That gets one wondering if a historical chart might resemble the current one.  The two charts would have to be several years long...short enough so that the spikes aren't blurred out, but long enough to capture long term gains and losses that could be relevant to taxation.

You can view a chart of the Dow that goes back to 1930 at Yahoo.  Problem is, a big gain in the 1930's might be 2 or 3 points in a day.  Nowadays, the Dow can move 300 or more points in a day, so peaks and dips that were actually quite significant 20 years ago get blurred out.  So we charted the Dow in three year periods from 1970 to the current date on the following page.  In the long term, if not the short term, the Dow actually does a decent job of paralleling all the other important indices.  The advantage of using Dow data is simply that there's a lot of it.

If you're looking to find a mirror resemblance to the current market, you might look at 1973-1975.  It's a bit of a stretch, and we won't claim there are any special odds that 2004 will resemble 1976.

Looking at all the charts, one is struck by the dearth of losing three year periods since 1970.  There are only a few.  We're in one now.

I threw in some historical notes as well.  Sometimes sophisticated market analysis blurs the simple fact that the 80's and 90's were relatively optimistic times (freeing of Iran hostages on the first day of Reagan's presidency, end of communism and fear of nuclear war), while the 70's (Vietnam, Watergate, the Cold War, OPEC, the Iranian hostage crisis) and current times (terrorism, North Korea, the Iraq situation) feel a bit gloomy.  Thus the markets.

Oct 24,2003...Again, yearlong winners took a beating without losing groups showing a tendency to move outside the general market.  Looking at the data, a sub-theme of large-caps performing well may be developing.  Others have noted a slight seasonal tendency for large-caps to outperform around this time of year, and this trend does seem to be manifesting.

*****************

Here's an article on the likelihood of market "bubbles" at Forbes.  Basically, the Nobel laureate professor in question, Vernon Smith, says that another 1999-2000 style market bubble is unlikely at this time because the memory of that bubble is simply too close.  We'll have to wait a while for another generation to enter the market.  The article, naturally, is a bit shallow...I'd like to see what this guy is saying and what his methods are in a bit more depth.  Maybe I'll dig around.

Oct 23, 2003...Again, yearlong winners took a beating, but losing groups performed in line with the general market.

Oct 22, 2003...Yearlong winners got hurt today, but yearlong losers performed in line with the general market.  We're continuing to wait for some sign that it's time to go bottom-fishing.

*****************************

I dug an interesting article up in a search on the "January effect".  Actually, all I was trying to do was see if this web site pops up when "January effect" is entered on Yahoo (it doesn't).  

Here's the idea put forth in the article (dated September 29):  we may see something of a January effect in October for the usual reasons:  tax selling (many mutual funds wrap up their fiscal years September or October) and window dressing.  There's a twist, though:  while some stocks have made very nice gains over the last year or more, many of these gainers still aren't anywhere near their highs of 2000 and 2001.  That means these funds could sell these sorts of stocks before the end of your fiscal year...nice gains over the course of the last year, but still losers from the time of purchase...take a tax loss, and then buy these positions back.  This offers a possible explanation as to why we are seeing the current gains in yearlong gainers.  Given this logic, you'd still expect the usual yearlong losers to prosper...we'll take a look at the end of the month.

The strategy is confounded just a tad by the "wash sell rule", which states that you can't buy back your tax losers immediately...you must wait 30 days.  But this isn't such a big deal...you buy similar companies that you didn't own.  Or simply wait 30 days.

If you buy this argument (sounds reasonable, no?), it's interesting to note that the charts of these stocks should have a well-defined pattern:  1)  Going back a couple years, you see lofty prices, 2)   a bit later, a big drop,  3)  within the last year, a nice upward slope and,  4)  near the end of September, a little pullback.  Now, of course, these stocks are engaged in step 5...more upward action.  In the cited article, Lucent (lu) is given as an example, and this is precisely what you see in the chart, right down to an end-of-September pullback (on a volume spike, I might add).  I'm not much of a chart watcher, but one might imagine a chartist going gaga over the beauty of this formation (for all I know, the technical guys would consider this ugly, but that's not my point).

At this point, Lucent is up nearly 25% from the beginning of October.  There's a miniscule bid/ask on the company, so almost all that 25% is profit.

Now imagine the same chart, but step 4 is completed in, say, the middle of February.  Same chart, but step 5) could be very different given the fact that nobody does tax selling or window-dressing in the middle of February.  It points to the possibility of using charts in conjunction with seasonal considerations. Or, more strongly, it suggests that seasonality can be primary over charting considerations.

As always, things are probably a bit murkier than the above would have it.  After all, we're talking about mutual funds whose fiscal years end in either September or October (supposedly, one third of all mutual funds).  So you've got simultaneous tax-loss selling and buying.  I'll try to take a look at precisely which stocks gained in October at the end of the month, and see what makes sense and what doesn't.

Oct 21, 2003...We're seeing a continued trend toward gains in yearlong winners.  Today, you could have made 2.5% by grabbing the stocks with the highest 4% of yearlong gains, while the market was only up about 0.5%.  

The trend started at the beginning of the month and hasn't quit.  This is a bit unusual...normally, one would be seeing losers, especially September losers, getting a bigger piece of the action.  Note that the trend toward gains in yearlong gainers from June to September never really manifested, so maybe we're seeing a delayed reaction.  Also, it's been a positive year for the market in general, so it's not like there are hoards of beaten-down stocks to be picked.

It'll be interesting to see when this trend reverses.  My own guess...soon.  It's not so unusual to see yearlong gainers prosper in early October, but now it's the 21st.  You have to go back to 1995 to find anything similar.  November 1995, for what it's worth, did reward yearlong losers.

*******************

Do a search under "seasonal investing" on Google and you'll get a Charles Schwab article called "Does seasonal investing really work?".  Some research guys at Schwab tested the "sell in May" thesis with data going back a thousand years, and concluded that the strategy doesn't work.  Therefore, SEASONAL INVESTING DOESN'T WORK.  There you have it.

The article begins with "At Schwab, we believe in..."

Oh well, give Schwab credit for discouraging churning, at least in this particular article.  There's also a nice bar chart on the page showing monthly market performance since 1926.

I'm not big on the "sell in May" philosophy myself, but I do believe that if the data were more recent (going back to 1950 or thereabouts, as opposed to 1926), then the case could be made that this sort of "seasonal investing" actually does work.  It's an easy statistic to manipulate, depending on how far back you go in history, which market you choose, where you put the money between May and November, etc.

I'm reminded of football commentary, where someone proclaims something like "the Detroit Lions haven't won in Dallas since 1973".  Are we supposed to believe that this fact has any bearing on the Lion's current chances of winning in Dallas?  I do believe the Lions have replaced some of the players that were on the roster in 1973.

Likewise, does data that goes back to Lindbergh's time really have any relevance for an investor in 2003?  Seasonality has underlying causes, however fuzzy they might be, and when these causes...tax laws, corporate laws, accounting trends, politics, saving rates, speed of information transfer, whatever...change, so should seasonality.  Schwab treats "seasonality" as something that is set in stone, and then conveniently proceeds to "debunk" it.

October 3, 2003...It may seem odd to the "random walk" and "market efficiency" guys that we're able to dissect just about any market, daily, monthly, quarterly, yearly, whatever, and arrive at trends that often go far beyond chance in terms of statistical significance.  

When looking at non-seasonal data (i.e. data that has not been compiled and clumped together over the same yearly period over a number of years), we don't think this significance should be so surprising.  The money to purchase stocks often comes out of the sale of other stocks.  Sales have the effect of driving stocks downward, and purchases have the opposite effect.  Simple enough, but this shows that stocks aren't really analogous to numbered balls in a lottery tumbler...the action of buying and selling creates greater "separation" between two stocks than would be expected if the stocks were allowed to simply wander up and down in a way reminiscent of Brownian motion.

So, you extend the "two stocks" thinking to two groups of stocks, and then to many groups of stocks, and you get one-time trends that can be very significant, whether measured over a day or a year.  Then you've got the momentum guys who continue to push stocks upward, the window-dressers and stop-loss guys who sell for no better reason than "it already went down", political and economic events on a national scale, and a million other factors that serve to increase, not decrease, the separation between groups of stocks..

More surprising is the significance of the various seasonal trends we follow.  A lot of thinking has gone into the subject with regard to the January effect.  The academics fight it out over the proportion of the effect that is due to year-end selling for tax purposes, and "window dressing" on the part of money managers, with a smattering of "the traders have gone on vacation" thrown in.  

Maybe people just go for bargains after Christmas.  We're not entirely facetious here.  Look at what happens for a couple weeks following April 15th... a big market run, on average, with money-losing stocks faring quite well.  Or maybe it's like this...after Christmas and April 15th, people simply can't quit the habit of sticking their hand in their pocket and pulling out a wad of money.

Whatever the underlying reasons, it seems that maximizing profits simply isn't the only concern that investors have.  And even when it is the prime concern, folks don't necessarily go about investing in a very rational way.  If they did, we might (might) have an efficient market.

***********************

I've been lurking on a message board where the subject of seasonal investing is sometimes broached (investorhub.com).  One guy who actually mentioned this site got a minor tongue lashing about what "seasonal investing" is and isn't (apparently, this site belongs in the "isn't" category).  Over there, it's defined as kind of a periodic momentum strategy...if a stock has done well in a certain period over the years, one should assume that it will continue to do well in the same upcoming period.

If he who dealt out the tongue-lashing actually spent more than a couple minutes perusing this site, he'd see that we do examine this strategy with our "prd_lastyear" and "prd_3years" indicators.  The interesting thing is, these two indicators don't often show up as strong positive indicators in our tables.  Sometimes (e.g. the month of February) they even appear to have a negative correlation...a strong performance last February predicts a weak performance this February.

The intriguing thing about this chat room is that one guy has set up a model portfolio based on this concept of seasonality, and has fared outrageously well, turning a hypothetical  $25,000 into something like $400,000,000 in five years.  He has twelve stocks, one for each month, and you simply buy the right stock for the right month (qlgc is the stock for every October).  Unfortunately, it does seem that the portfolio was set up after some of these gains already occurred, and there's a bit of hemming and hawing as to why that should be considered "fair".  Even if hindsight wasn't of benefit, it may be that this guy is making excellent investment decisions for the wrong reasons...without additional screening, this "periodic momentum" strategy just doesn't work very well as far as our data is concerned.

By the way, an excellent way to view how a stock has performed in various months over the years is to go to thomsonfn.com and click on their "seasonality" feature.  The subject of whether it's at all possible to make money with this information isn't addressed, but it's definitely interesting to see graphically (a bar chart) how your favorite stocks have performed in various months.

Copyright © 2010 MarketSynopsis.com. All rights reserved.