Golden Capital Securities

Investment Outlook

October 2000

Have you ever wondered if there were particular months that have performed better historically than others?  Some of you may have heard of the January Effect (the performance of the month of January indicates the performance for the rest of the year), Black Octobers (the crash in 1987 and worries of a repeat in 1997) and even the Friday Afternoon Effect (all the brokers go home early so liquidity vanishes from the market).  There is a great deal of debate as to whether these market imperfections actually occur, but what really matters is whether we as traders and investors can take advantage of these supposed market anomalies for fun and profit!

Some the traders here at Golden Capital have mentioned that they have a gut feeling that September tends to be a soft month for returns in the markets.  This contradicts a popular feeling that there is a tendency for markets to improve as investors return from their summer holidays.  We thought that it would be interesting to put intuition to a statistical test and see if either of these gut feelings match up with real world statistics.

Empirically, there is some indication that these occurrences do happen with some significance. 

First, the month of September seems to be the month with the lowest average return at 0.04% and a geometric average -0.02% (which takes into consideration the effect of compounding) over the last 25 years.  This means that if you invested money every September 1st and withdrew it September 30th for the last 25 years, you would actually have about the same amount of money as you originally deposited.

By comparison if you did the same investment plan with January, the month with the highest average return, you would have doubled your money.

However, averages only give a partial picture as -25% and 25% make an average of zero, so we also looked at the standard deviation.  Standard Deviation indicates the level of dispersion of a sample; a high standard deviation indicates widespread observations, with the converse being true for a small standard deviation and is a good measure of volatility, or risk.  The tightest level of dispersion actually occurs in June (the start of vacation season?), with the greatest level being in October.  Part of the high dispersion for October is as a result of the market crash in 1987, remove that one month and October ends up right near the other monthly average.

So it does look like September is a soft month: the lowest average return, the lowest geometric return, and a relatively tight distribution at 3.6%.  But wait! There's more. 

During the 1990's there has been a strong correlation between the monthly returns from September and November.

If we look at the chart to the left you can see a remarkable correlation between the two months.  For you quant jocks, the correlation is 0.53 with an R-Squared of 0.28.  For the non-statistically inclined, this means that during that decade of the Nineties, 28% of the returns in each November could be explained by the returns in the same year's September and how well they move together (0.5 is pretty good for financial figures).  However, if we remove 1990, as there was a large divergence, the correlation jumps to 0.86, with an R² of 0.73, or explaining 73% of the movement in November with the return in September. WOW!  And this carries back to the Eighties as well, but shifting forward by one month to October. 

Although the correlation for September to November in the 1980's is almost meaningless, October (see 

below) has a very strong correlation (0.71) and R-squared (0.5) to November.  

You may notice that the years 1980 and 1990 had movements that differed from the remainder of their respective decade's. Plus, the month that held the high level of correlation to November moved by one month depending on the decade, from October in the Eighties to September in the Nineties.  It is not so much the pure science behind these observations as it is the fact that it did occur which is interesting.

In September of 2000 the market is down approximately 6%.  Therefore, what's a trader to do?  

There are a few things that could happen:

1) The market could repeat what has happened over the last nine years and repeat the performance of September in November.  If this is the case increase your cash now and start to purchase back in as we approach the end of November.

2) Or, like in previous decades, the year ending in zero moves opposite to the market, in which case be fully invested for the beginning of November.

3) Or, the markets are truly random and that they will move in a direction completely independent from previous periods; the famous Random Walk theory holds.

In other words, the market could go up, down, or randomly.  What insight.

Traders and Investors are always looking for that extra level of knowledge, the "Golden Nugget".  Therefore another possible source of superior market knowledge was discussed in a recent article by the Wall Street Journal that listed Insider transactions as another widely used, leading indicator.  

The use of sales and purchases by Insiders has long been a topic of discussion; who would know their business better than the people intimately involved in it?  The article mentions that the website InsiderTrader.com has followed the ratio of insider buyers to sellers for several years.  By using a ratio of sales to purchases based on a weighted average of the number of shares and the volume of shares traded, InsiderTrader.com points out that this ratio reached a relative peak for the week ending September 20th of 2.65 to 1.  In other words, Insiders sold 2.65 times as much stock as they purchased. 

The article also points out that there was an interesting divergence between Old and New Economy stocks, and not in the way you would normally think.  Again for the week ending September 20th, Insiders who's stock traded on the NYSE had a sell : buy ratio of 3.31 : 1, while Insiders on the NASDAQ, the exchange of predominantly high-tech companies, only had a 1.68 : 1.  The blue chip Insiders were selling at a rate double their high tech counterparts.

As mentioned before, the Insiders know their companies and the industry better than most investors do.  The chart below really shows the activity by Insiders over the last 5 years.  The major point to take away from this chart is the increased level of buying that Insiders conducted in mid-98.  As the market dipped, Insiders swallowed up huge amounts of shares.  Since then, the Insiders have been a very good leading measure for the Russell 2000 (a broad measure of the top 2000 US companies).  However, the chart also shows that during this year, the Insider effect has been almost neutral.  Even the Insiders have little feeling for where the market is headed.

Source: InsiderTrader.com

Finally, the dollar has hit its lowest point versus the US Dollar since late spring of this year.  The recent downdraft of the dollar can largely be attributed to the possibility that the Canadian Federal government could call an election for the very near future.  Generally speaking, the financial markets do not like uncertainty and the close race between the Federal Liberal and the Canadian Alliance Parties has resulted in a high level of uncertainty.  Although we are by no means close to the all time low for the Canadian Dollar, that occurred back in 1998 at around 63 cents, we certainly are trading under the psychological level of $1.50 US Dollar per Canadian. 

Source: Bridge

Politically, international financial markets would likely prefer a conservative type government that will reduce our huge debt and keep a tight fiscal and monetary policy.  The traditional method of running deficits in good times and bad cannot continue.  Almost all of the political parties in Canada agree with balancing the budget, although to varying degrees.

Therefore, barring interference from the Bank of Canada, the Canadian Dollar's reaction to the strength of the respective political parties will likely be:

 

PARTY	GENERAL MARKET RESPONSE	EFFECT ON DOLLAR
Liberal	Status Quo, more of the same	Likely return to 66-70 cents
Canadian Alliance	New, fiscally conservative	Improve over current range
NDP	Heavy social spending	Likely dip even lower
Conservatives	Fiscally conservative, unlikely to lead	Not an issue

It will be interesting to watch the impression that the international money traders have to the polls leading up to the election and the results.  We will keep you posted.

Until next month.