In Part 1, I introduced the Know Trends Index (or KTI) that I wrote about in my book “Seasonal Stock Market Trends.  I also noted that stock market performance:

*Tends to be very good when the KTI is greater than or equal to +5.

*Tends to be very bad when the KTI is +1 or less

*Is mostly very good when the KTI is at +2, +3 or +4 – except of course for when it is horrifically bad (-38% drawdown in 1938 and -57% drawdown in 2008).

So the goal this time out is to create a trading strategy that takes all of this into account.

A Seasonal Stock Market Trends Trading Strategy

The KTI for this test will be calculated exactly as spelled out in “Seasonal Stock Market Trends” and has historically ranged from -1 to +8.  Because all of the time periods are know in advance (hence the “Known Trends” moniker) the KTI reading for each given day can be calculated ad infinitum into the future.

Trading Rules:

A) If KTI <= +1 – Hold cash (test assumes 1% of interest annually)

B) If KTI >= +5 – Hold Dow Industrials using leverage of 2-to-1

C) If KTI >=+2 and <=+4 AND the Dow Jones Industrials Average is above its 200-day moving average(*) then hold Dow Industrials using no leverage.

D) If KTI >=+2 and <=+4 AND the Dow Jones Industrials Average is below its 200-day moving average(*) then hold cash (test assumes 1% of interest annually).

(*) – There is a one-day lag in measuring Dow versus its 200-day moving average.  Allow me to explain:

-Let’s say that it is Tuesday and we know that the KTI reading for Wednesday will be +3.  If the Dow closes on Monday above its 200-day moving average (calculated as of the close on Monday) then we will buy (or continue to hold) at the close on Tuesday. If the Dow had closed on Monday below its 200-day moving average we would not have entered a new position on Tuesday (or if we were already in a position we would sell it on Tuesday because the Dow was below its 200-day moving average and the KTI was less than +5).

We will start our test on 12/31/1934.  When a long position is held we will assume that we gain or lose the daily percentage change for the Dow Jones Industrials Average

Random Notes regarding Results:

*This is a “long-term” method.  When I say “long-term” I am not referring to holding period as much as I am referring to the fact that an investor needs to let this method work over an extended period of time in order to derive the benefits.

*This method makes money by “grinding it out” – i.e., by making an above average return when using leverage, by tracking the market when long but not using leverage, and missing a lot of the “pain” by being out of the market during certain unfavorable periods.

*Year-to-year results should not be focus.  In other words, some years the ystem will outperform buy and hold and other years it will not.  What really matters is long-term rolling rates of return.

OK, with all that out of the way, let’s look t the actual results.

Results

The short version is this:

*The average annual return for the Seasonal System is +15.8%

*The average annual return for Buy-and-Hold is +8.0%

(these results reflect price only plus interest earned while in cash, no dividends)

Now a funny thing happens when you compound returns over a long period of time.  So for the record:

*$1,000 invested in the Dow using buy and hold since 12/31/1934 grew to $175,240

*$1,000 invested using the seasonal system since 12/31/1934 grew to $40,402,501

Like I said, compounding at a higher rate of return really adds up over long periods of time.

To better illustrate the results over time let’s look at two separate roughly 40 year periods. Figure 1 shows the growth of $1,000 for the Seasonal System versus Buy and Hold from 12/31/1934 through 12/1/1974.Figure 1 – Growth of $1,000 invested using Seasonal System(blue line) versus Buy-and-Hold (red line); 12/31/1933-12/31/1974

Figure 2 shows the growth of $1,000 for the Seasonal System versus Buy-and-Hold from 12/31/1974 through 5/22/2015.Figure 2 – Growth of $1,000 invested using Seasonal System(blue line) versus Buy-and-Hold (red line); 12/31/1974-5/22/2015

Figure 3 shows the rolling 10-year returns for the System versus buy and hold.Figure 3 – Rolling 10-Year returns for Seasonal System (blue bars) versus Buy-and-Hold

The bottom line is this: Rolling 10-year returns for the Seasonal System have outperformed Buy-and-hold 70 out of 71 times. In other words, starting at the end of 1944, if you looked back at the performance over the prior 10 years, the  Seasonal System outperformed Buy-and-Hold 70 times.  Buy-and-Hold outperformed the Seasonal System one time.

As I am wont to say, ratios like 70-to-1 are something we quantitative analyst types refer to as “statistically significant.”

Finally, Figure 4 shows the year-by-comparative results. 

Figure 4 – Year-by-Year Results for Seasonal System versus Buy-and-Hold

Summary

See Part 3 to follow….

Jay Kaeppel

See also Good News Bonds, Bad News Bonds

In my book “Seasonal Stock Market Trends” I pulled a variety of well known (and not so well known) seasonal stock market trends (what else?) into a composite index referred to as the Known Trends Index”, or KTI for short.  Sure, if I was a marketing guy I would’ve opted for something more along the lines of “Jay’s Hidden Order in the Stock Market Revealed” Index.  Alas, the marketing gene apparently skips a generation in the Kaeppel Family.

No matter. The gist is this:

The KTI is comprised of 13 indicators (including The Election Cycle, Trading Days of The Month, Sell In May and Go Away, the Summer Rally, the 40-Week Cycle, the 212-Week Cycle, Trading around Holidays and 6 others, each with specific bullish, bearish or neutral criteria) and has historically ranged from a high reading of +8 to a low reading of -1.

Using the KTI to Trade

The more seasonal trends that are favorable at a given point in time the higher the KTI and vice versa.  The theory is that higher KTI readings suggest more favorable conditions for the stock market than do lower KTI readings.

Does this theory hold true in reality?  Take a look at Figures 1 and 2 and decide for yourself.

Figure 1 displays the growth of $1,000 invested in the Dow Jones Industrials Average (price only, no dividends included) only on those days when the KTI reads +5 or higher, starting on December 1^st, 1933.Figure 1 – Growth of $1,000 invested in Dow Industrials only when KTI >= +5; 12/1/1933 through 5/22/2015

Compare and contrast the results displayed in Figure 1 to the results that appear in Figure 2.  Figure 2 displays the growth of $1,000 invested in the Dow Jones Industrials Average (price only, no dividends included) only on those days when the KTI reads +1 or less, starting on December 1^st, 1933.Figure 2 – Growth of $1,000 invested in Dow Industrials only when KTI <= +1; 12/1/1933 through 5/22/2015

For the record:

*$1,000 invested in the Dow only when the KTI >= +5 grew +5,470%

*$1,000 invested in the Dow only when the KTI <= +1 declined (-96.1%)

This dichotomy in performance is what we quantitative analyst types refer to in our highly technical terms as “statistically significant.”

Under the category of “What have you done for me lately”, Figure 3 displays the results generated from holding the Dow Industrials when KTI >= +5 versus KTI <= +1 since 12/31/1994.Figure 4 – Growth of $1,000 invested in Dow  Industrials when KTI>= +5 (red line) versus $1,000 invested in Dow Industrials when KTI<= +1 (blue line); 12/31/1994-5/22/2015

One problem with all of this information (going back to 1934) is that:

*The days for which the KTI >=5 represents only 14.65% of all trading days

*The days for which the KTI <=1 represents only 21.56% of all trading days

So what about all the other days?  Figure 3 displays the performance of the Dow during all days since 12/1/1933 when the KTI was between +2 and +4.Figure 3 – Growth of $1,000 invested in Dow Industrials only when KTI = +2 or +3 or +4; 12/1/1933 through 5/22/2015

*The good news is that the net gain was +8,500%. Which is what we quantitative analyst types refer to in our highly technical terms as “not too shabby.”

*The bad news is that investing only on days when the KTI read +2, +3 or +4 witnessed a drawdown of -38% in 1938 and a massive drawdown of -57% in 2008 (i.e., “Shabby”).

So how to put it altogether and actually trade?

Simple. Stay tuned for Part 2.

Jay Kaeppel

See also Crude versus Bond Yields (from One of My Favorite Sites)

First the (potential) good news.  The (potential) good news is that one trend in bonds that I wrote about a while back here and here may finally (potentially) be playing out the “right way.”  Although, as there are still three more trading days left in the month of May, it is clearly a little early to declare “victory.”

In any event, the trend I am referring to is the historical propensity for t-bonds to advance in price during the last five trading days of the month.  Figure 1 displays the growth of equity achieved by holding a long position in t-bond futures during the last five trading days of every month since 12/31/1983 versus the performance for t-bond futures during all other trading days.Figure 1 – Gain from holding long t-bond futures last 5 trading days of every month (blue line) versus gain(loss) from holding long t-bond futures all other trading days of the month (red line); 12/31/1983-present

The long-term trend is fairly clear.  Ah, there’s the rub.  While the long-term is all well and good, the short-term can be a real pain in the you know what.  T-bonds lost -$3,800 during the last 5 trading days of March and April combined.  Of course, I did point out in a subsequent article that both of these months have been something of “weak sisters” compared to the majority of other months.  So what’s the latest buzz?  Well on 5/22 – the 5th to last trading day of the month of May – t-bonds futures lost another -$625.  However, on 5/26 bonds soared in value by $2,219 per contract, so the “last five trading days of the month” period has something of a bullish leg up at the moment.  We will just have to wait and see what transpires between now and the close on Friday.

In any event, this constitutes “the Good News” presently in t-bonds.  Which leaves us with – what else  – the “Bad News”.

The Japanese Stock Market and T-bonds

As I wrote about here the U.S. t-bond market has demonstrated a strong tendency to perform inversely to the Japanese stock market.  Now for the record, I have heard theories from market analysts whom I consider to be smarter and/or more enlightened than I when it comes to explaining relationships between various – seemingly unrelated – markets regarding “why” this Japanese stocks/U.S. t-bond relationship has been so persistent.  However, as a proud graduate of the “School of Whatever Works” I will be candid – I personally don’t care about “why” as much as I do “for how much longer.”

Figure 2 displays ticker TLT – an ETF that tracks the 20+ Yr. U.S. t-bond in the top clip and ticker EWJ – an ETF that tracks the MSCI Japan Index – in the bottom clip.  Using very simplistic analysis:

*The trend is considered bullish for bonds when the 5-week average for EWJ is below the 30-week average for EWJ, and;

*The trend is considered bearish for bonds when the 5-week average for EWJ is above the 30-week average for EWJ

Figure 2 – T-bonds (ticker TLT) tend to trade inversely to Japanese Stocks (ticker EWJ)

Figure 3 displays the gain achieved by holding a long position in t-bond futures ($1,000 per point) since December 1997 as follows:

*Holding a long position in t-bond futures when the trend is bullish (red line) and;

*Holding a long position in t-bond futures when the trend is bearish (blue line).Figure 3 – Amount gained or lost being long t-bond futures when EWJ is bullish for bonds (red line) versus bearish for bonds (blue line)

The difference in performance pretty well speaks for itself.  So the “bad news” for bonds is that the 5-week average for EWJ remains well above the 30-week average, i.e., this indicator is still “bearish.”

 Summary

From a trading perspective I am looking to scratch out some sort of gain from a long position in bonds through the end of May.  But from there, the bearish configuration for EWJ as it relates to bonds suggests that a switch either to the short side of bonds, or to simply hold no position at all, might be the way to go.

As always these are not “recommendations”, merely “observations.”

Jay Kaeppel

Another great post from one of my favorite sites (McClellan Financial Publications) regarding the use of crude oil as a leading indicator for bond yields.  Definitely worth knowing about…Click here to read Crude Oil Leads Bond Yields

Jay Kaeppel

See Jay’s recent post: Will Natural Gas Break Wind in June?

OK, my last article (Will Natural Gas Break Wind in June?) did sound a little apocalyptic regarding the prospects for natural gas in June. But maybe that did not present the full picture. While that previous article did detail a bearish seasonal period for natural gas starting at the close of trading on June Trading Day #11, very often there is a brief “pop” prior to things taking a turn for the worse.

The Brief Respite

The brief favorable period we will look at now:

*Starts at the end of June Trading Day #8 and;

*Lasts for 3 trading days before things flip to “unfavorable”.

Figure 1 displays the results generated by holding a long position in natural gas futures during this 3-day period since the inception of trading in 1990.Figure 1 – Long Natural Gas futures June Trading Days #9, 10 and 11

For the record:

*There have been 17 up years and 8 down. 5 of the first 6 years showed a loss. Since then 16 out of 19 years and 10 out of the last 11 have seen a gain registered during this three day trading period.

*The average gain was $2,914

*The average loss was -$1,413

*The largest gain was over $10,000 (2006)

*The largest loss was over -$5,000 (2003)

*13 of the 25 years witnessed a 3-day gain of at least $1,800.

So clearly not for the “faint of heart” nor the “small of trading account.”

Figure 2 displays the year-by-year results

Figure 2 – Year-by-Year Results long natural gas futures during “favorable” June period

Combining Bullish and Bearish June Periods for Natural Gas

Now let’s assume a trader held a:

*Long position in natural gas futurse during June Trading Days 9, 10 and 11

*Short position in natural gas futures from the close of June Trading Day #11 through the close of the 14th trading day of July

The cumulative gain for this “long then short” strategy appears in Figure 3.Figure 3 – Long and Short Periods for Natural Gas combined

The year-by-year results appear in Figure 4.

Figure 4 – Yearly results for Long and Short periods for Natural Gas combined

For the record:

*21 years showed a gain (84%)

*4 years showed a loss (16%)

*The average gain was over $7,576

*The average loss was -$1,573

*The largest gain was over $27,000 (2008)

*The largest loss was -$3,790 (2012)

Also for the record, these results include no deductions for slippage, commissions, taxes, etc. and use futures data that I have had sitting on my computer which are assumed to be accurate.

Note too that these results do not take into account the sheer terror that can be associated with holding any position in natural gas futures – even if things are going your way at the moment (trust me on this one) – because this market can be incredibly volatile and can turn on a dime.

Summary

A few notes:

*There clearly seems to be “something anomalous going on” in natural gas in the month of June.

*There is (sadly) no guarantee that these anomolies will play out in June 2015 as they have in the past.

*The last time I wrote about natural gas tendencies in the month of June it ended up being the worst year ever to follow these trends (Still think I am just paranoid when I say “Murphy hate me”?).  So I think we all know what that means.

*Trading natural gas futures is absolutely, positively “not for everyone.”

Thanks to the advent of ETFs and options on ETFs there are ways to play these trends without ever trading a futures contract.

I’ll plan on revisiting this topic as the June seasonal time frame draws closer.

Jay Kaeppel

See this link from McClellan Financial Publications

Apparently, positive thinking does help…..

Courtesy of one of my favorite sites..http://jayonthemarkets.com/?p=1718

See Jay’s latest post: One More Plunge for Crude Oil?

It’s been (alas, correctly) brought to my attention that my description of the “Simple Pattern” I wrote about was, um, not so simple.  To wit, the author writes:

“Not every Day 1 signal generates an entry signal. If the 2nd trading day after the “close below previous low” (Day 3) does not trade above the high of the day after the “close below previous low” day (Day 2), then no further entry signal exists and the pattern is voided.”

Huh?

OK, so under the category of “A picture is worth a whole bunch of words”, please consider the examples below.

Example #1

Day 1 – the close is below the previous day’s low

Day 2 – the day after Trading Day 1

Day 3 – price breaks above Day 2 high triggering a long trade

1^st profitable close occurs on day after Day 3

Example #2

Day 1 (1^st red arrow) – the close is below the previous day’s low

Day 2 (1^st blue arrow) – the day after Trading Day 1

Day 3 (not marked) – price DOES NOT exceed Day 2 high, so signal is voided

Day 1 (2^nd red arrow) – the close is below the previous day’s low

Day 2 (2nd blue arrow) – the day after Trading Day 1

Day 3 (green arrow) – price DOES exceed Day 2 high, trade is entered above Day 2 high

1^st profitable close – Price closes above entry price on Day 3 so trade closed there

Hopefully these examples help.

Jay Kaeppel