Speculation is a grand game of probabilities. When a speculator chooses to make a trade, he never has a 100% chance of winning or losing. Any trade unfolding over the inherently uncertain future is always drifting somewhere out in the great gray area between certainties.

In this purely probabilistic environment, it is always in a speculator's best interest to gain a deeper understanding of the underlying market in which he is trading. Literally everything that one can learn about a market ought to contribute to a more accurate perception of the true probabilities governing that market.

The reason deep knowledge of a market helps is because probabilities are always changing in real time. Launching a long trade today might have a 40% chance of success while launching this same long trade a month from now might have a 60% chance of winning. Obviously speculators want to have some idea of where these probabilities stand at any given moment to help generate superior timing for their trades.

It is in this spirit that I have been becoming more and more interested in seasonals lately. Seasonals attempt to quantify seasonal tendencies for a particular price to act in certain ways at certain times of the year. While seasonals generally don't get much use in stock trading, they are widely followed in futures trading. And in futures, especially crops that depend on weather seasons, is where the classic seasonal examples lie.

My best friend from high school is a wheat farmer in South Dakota. He generally harvests his spring wheat crop in late summer, at least when droughts aren't cooking his fields brown. Every other wheat farmer is harvesting wheat in late summer as well though, so during harvest wheat supplies coming onto the markets soar. This tends to drive wheat prices lower this time of year.

But a few months later as temperatures drop and winter arrives, much of the harvested wheat has been absorbed into the markets. Supplies shrink which generally drives prices higher in December. So my friend, and many farmers, game this seasonal tendency. Rather than selling all their wheat in August at low harvest prices they store as much as they can in their own bins and don't bring it to market until the winter to sell at better prices.

While the seasonal tendencies of crops dependent on weather cycles are fairly obvious, there are probably much more subtle seasonals that aren't widely understood. A knowledge of these could give a speculator a slight edge, increasing his understanding of the markets and hence improving the timing of his trades. Although I would never use seasonals as a primary trading tool, they could be very useful to fine-tune entry and exit timing.

There are some excellent information services that create beautiful seasonal charts for futures, but I have a couple issues with the usual presentation. First, I am a stock and stock-options speculator so I am much more interested in how seasonals may influence stock prices than futures. Second, the standard approach to building seasonal charts is to lump decades of price data together to define seasonal probabilities.

In all kinds of past studies I have done on the markets, one of the core observations underlying all markets is that bull markets behave differently than bear markets. Prices rising on balance in a secular bull act very differently from prices falling on balance in a secular bear. If seasonal charts are created across bulls and bears, I suspect the secular trends will kind of cancel out. This is good in the sense that it ought to distill purer true seasonal influence onto a chart, but I don't want to trade partially based on bear data when I am in a bull.

So I am interested in shorter seasonal analysis, considering only the years of the current bull markets in commodities and commodities stocks. While this will dilute pure seasonal influences by mixing in secular trends, I suspect it is more relevant to trade the current bulls based on their bull-to-date performances. Time will tell whether my thesis is correct, but hopefully you can understand my reasoning on bull-only seasonals.

I am interested in all kinds of bull-only seasonals and hope to delve deeper into this line of inquiry. I am interested in charting and understanding bull-only seasonals in gold, silver, the HUI, oil, the XOI, natural gas, the XNG, the base metals, and other areas. There may be subtle seasonal tendencies within each bull that aren't yet widely understood.

For this initial foray though, I chose to look at crude oil bull-only seasonals for two reasons. First, oil has more potential to be exciting this time of year than most other sectors. Second, we have heavy exposure in oil-stock call options today in our Zeal Speculator alert service, and oil prices drive oil stocks. So I want to better understand how seasonals may affect the probabilities governing the sell timing for realizing our big gains. Is there a particular week or month when the odds of the next oil interim top materializing are the highest?

Our current oil bull started in late 1998, so I analyzed the years 1998 to 2006 to build these seasonal charts. While 2001 was a down year for oil, it is included because it was ultimately just a long bull-market correction and not the start of a new secular bear. I used the raw oil price data to build two charts, one indexed annually and one indexed monthly.

The first annually indexed chart considers every year from 1998 to 2006 individually. The level where oil starts the year in each case is considered 100. Then the oil prices for the rest of each year are indexed off these initial 100 levels. Once each year is individually indexed and the dates are matched between years, all the self-contained year indexes are averaged. The resulting average of these individually indexed years is charted below.

The second monthly indexed chart does this same process but at a more granular level. Individual months of each year from 1998 to 2006 are indexed and then averaged across all the same calendar months. So every January from 1998 to 2006 is individually indexed, all these Januaries are averaged together in indexed terms, and the resulting chart shows pure intra-month tendencies over the lifespan of this oil bull.

Each chart also includes an inset chart of the same indexed dataset with the addition of standard deviations. One standard deviation is charted above and below the indexed data from the large chart. I included these to show that these index averages are not usually tight and hence bull-only seasonals are not appropriate as primary trading tools. Nevertheless, they still ought to help us sharpen our understanding of oil probabilities.

On average since 1998, oil rose an astounding 30% or so from 100 indexed in January to over 130 indexed in early October! This sounds somewhat unlikely on the surface, but this oil bull really has been extraordinary. In December 1998 oil actually closed at $10.73 per barrel, lower than the early 1970s pre-Arab Oil Embargo prices in real terms. As of July 14th of this year, the oil bull had rocketed 618% higher!

So if we take our composite indexed oil behavior over each of the past nine years, oil starting at 100 indexed, running to just above 130 or so in early autumn, and then retreating back to 123ish does this make sense? Yes. This oil bull is about 8 years old. If 23% returns are compounded annually for 8 years, they yield ultimate gains of 424%. This is much lower than the 618% actual oil gain and proves the plausibility of such a steep oil seasonal track.

Interestingly in its bull to date, oil has tended to spend the first half of each year gradually climbing higher within a tight uptrend. Seasonally in this bull oil has tended to run about 20% higher between early January and mid-July. On January 1st this year oil closed at $63.10. By July 14th it had risen 22% to $77.03, an all-time nominal high. Thus so far in 2006 oil has remained in line with its bull seasonal precedent.

While oil's gains have been consistent in the first half of the years of this bull on average, this time of year is when the king of commodities really starts to shine. From late July until the end of September has been oil's strongest time seasonally in this bull to date. Check out the enormous August surge in the chart above! Today we are right at the early stages of where this surge has tended to occur on average.

Why does oil tend to surge in August? The futures guys attribute it to various factors. One is anticipation of the hurricane season in the Gulf and hence possible supply disruptions spawned by hurricanes. Another is the fact that August is often the biggest vacation month which drives very strong gasoline demand. August tends to be the highest-demand month for gasoline, exerting upwards pressure on oil prices. Whatever the true reasons, oil's seasonal strength tendency this time of year definitely exists.

If oil follows its pattern this year and heads to 131 indexed by the end of September, we are looking at $83 oil as the probable seasonal top. Since oil averaged $71 in June and $74 in July, I think it is totally reasonable to expect low $80s oil at the height of this oil season. Interestingly this dovetails nicely with the probable Relative Oil target for this upleg which should be running around $84 by the end of September.

And in reality these projections could be conservative, as they are based on pure technicals alone and assume no major hurricane damage nor major geopolitical crises. If we get a nasty hurricane that does some serious damage to oil infrastructure in the Gulf of Mexico or a real shooting war erupts in the Middle East that affects the oil-producing countries, then oil could go much higher.

But even without such exogenous shock events, oil has had a very clear tendency in this bull to date to rise sharply on average between late July and the end of September. Speculators who are long oil or elite oil stocks and oil-stock call options today as we are should take comfort in these seasonals. Oil seasonals suggest that the best time to close long oil-related speculations is typically the end of September.

Once oil peaks seasonally in late September or early October, it tends to spend most of Q4 in a well-defined seasonal downtrend. This is the weakest time seasonally for oil and hence not the time to be long for short-term trades. This downtrend is very valuable though as it grants an excellent entry point for new long trades ahead of the subsequent year. Seasonally oil has tended to bottom near 118 indexed in early December. Speculators should definitely be aware of this.

Now that you've seen how impressive oil seasonal tendencies look in the next couple months or so, a big caveat is in order. In the grand scheme of things there really aren't that many years between 1998 and 2006, so the sample size used to build this bull-only seasonal chart is small compared to conventional seasonal charts. Due to this small sample size, the standard deviations between individually indexed years are huge as shown by the yellow bands in the inset chart above.

By the end of September near seasonal highs, the standard deviation of annually indexed oil is running above 40! This means that there is theoretically a 68% chance that oil will end up in a massive indexed range between 90 on the low side to 170 on the high side. Translated into dollars, this standard deviation band will run between $57 to $107!

In general with any form of market analysis, the greater the standard deviation the less tightly the data is clustered. The less tightly it is clustered, the less alike it is and hence the more the final chart is a product of the smoothing inherent in averaging rather than true market tendencies. With these big standard deviations in oil's case, it suggests caution is in order and we shouldn't try to read more into this chart than is actually here.

My next chart is the monthly indexed one I described above. I built this chart because we have one opportunity at the very beginning of each month to trade oil stocks and oil-stock options in our monthly Zeal Intelligence newsletter. So I wanted to understand seasonally within this bull which calendar months tend to be the strongest, great times to be long, and which tend to be the weakest, the times not to be long. While I once again wouldn't use this as a primary trading tool, I appreciate the secondary probability perspective it offers.

Unfortunately due to limitations in our charting software, there are some graphical artifacts in this chart that are meaningless and should be ignored. Since each month is indexed individually, each month should be considered totally self-contained. Both January and February start out at 100. But January ends at 104 or so. The sharp drop rendered below between the end of January and early February is not in the data, it is just the software connecting the dots between months and not allowing the concept of totally discrete months.

So as you examine this chart, consider each month in isolation and ignore the final sharp move between the end of one month and the beginning of the next month. These sharp moves back to index 100 to start a new month shouldn't be here and therefore need to be ignored for analysis purposes. I apologize that I couldn't resolve this graphically and had to run with this flawed chart.

If a speculator is trading oil-related plays in pure calendar-month terms, there are definite seasonal tendencies that should increase his probabilities of winning. In both January and August since this bull began, oil has risen about 6% on average before pulling back 1% or so. If primary indicators concur, I definitely want to be heavily long oil stocks and oil-stock options heading into these months since the oil stocks usually follow oil very tightly.

With such big intra-month moves in oil probable in January and August, this is also when major interim tops in oil are most likely to occur. Indeed this was the case in the past year. Oil challenged $70 for the first time ever in nominal terms late last August on the hurricanes. And it didn't approach these levels again until this past January. Last August's highs weren't exceeded until April, so a major interim top happened in a surge month.

Tying these monthly seasonals together with the annual seasonals above, oil does have a tendency to rise in September but at a much slower pace to only about half the height of its typical August surge. As such, even though August is a surge month and oil topped last August, I still think traders ought to look for a September top in oil in pure seasonal terms. Of course any storm or geopolitical surprises could change this.

If oil goes up 5% or so in August and another 3% or so in September, roughly in line with its seasonals, then we have a potential monthly-indexed oil target above $80 for the next major interim top in a couple months. This is right in line with the low $80s targets mentioned above based on both annually indexed oil seasonals and more importantly Relativity. Multiple technical approaches yielding similar price projections increase confidence in the forecast.

In monthly terms really the only bad month for oil is October, when it tends to fall about 5%. For speculators, this means it is probably best to close out one's short-term oil-related longs by late September and not look to redeploy into new long positions until early December or so. We will certainly try to exploit these seasonal tendencies with our own oil stock and oil-stock options trading in our newsletters this year if our primary indicators agree with these seasonals.

If you are a Zeal Speculator alert service subscriber and have been riding our very profitable oil-stock call options campaign with us this year to realized gains in the hundreds of percent, we will probably be holding our current positions until late September based on these seasonals to let our current unrealized gains multiply. Then we'll sell out to realize our profits and wait. If oil follows its seasonal tendency to correct in October and November, we will be ready to redeploy in early December for the next upleg.

If you don't subscribe to one of our newsletters, you ought to! We are actively and successfully trading the unfolding commodities bull in multiple sectors including metals and energy. All of our extensive research, including this foray into seasonals, is ultimately designed to lead to high-probability-for-success trading opportunities for our subscribers. Please subscribe to our monthly Zeal Intelligence newsletter today to get prepared for the next big buying opportunity in oil stocks and oil-stock call options!

The bottom line is bull-only seasonal analysis offers a secondary perspective on when prices are most likely to be strong or weak. Prudent speculators can use this knowledge to help fine-tune the timing of their entries and exits on trades. When both primary indicators flash buy signals and seasonals look favorable, the odds for success are likely quite high for entering new long positions.

In oil's case, we are just entering the seasonally strongest part of the year for this crucial commodity. If you have existing oil-related longs, you'll probably be well served by holding them and watching them continue to appreciate over the next couple months. I'm interested to see how this type of bull-only seasonal analysis works in other commodities and commodities-related stock indexes as well.