Gold Bull Seasonals 2
Adam Hamilton July 13, 2007 3002 Words
Investment and speculation are ultimately the world’s greatest probabilities games. Traders exist in a realm of constant uncertainty, where capital must be deployed today well before the unknown future arrives. To increase the odds of success for any trade, traders should only deploy when probabilities swing way into their favor.
So it is not surprising that virtually all financial-market analysis is designed to help deepen our understanding of the underlying probabilities governing the markets at any time. The whole gamut of research approaches from fundamental to technical to sentimental ultimately boils down to gaming whether or not today is a high-probability-for-success opportunity to buy or sell.
Outside of the futures world, one particular technical tool used to help hone our understanding of probabilities in play is widely ignored. It is known as seasonality. Seasonality is the tendency for a price to be stronger or weaker at different times during the calendar year. This concept naturally emerged in the commodities world, where seasonal effects can make big differences in tactical price trends.
The classic seasonal examples are the agricultural commodities such as wheat. Summer, of course, is the prime growing season for wheat. So most new wheat that hits the markets floods in around harvest at the end of summer. With wheat supplies usually hitting their peak just after harvest, prices tend to retreat leading into harvest in anticipation of the temporary supply glut. Wheat traders know this and trade accordingly.
But seasonality is not limited to agricultural commodities rigidly governed by celestial mechanics. It exists, usually in more subtle forms, all over the place. For a little-considered example, think of general stock trading. It usually tends to taper off in the summer as traders go on vacation, so the summer stock markets are generally considerably weaker and softer than the winter markets when everyone is paying attention.
Seasonality also exists in gold. There are times of the calendar year when gold tends to do well and other times when it does not. Although there are many varying reasons for this phenomenon around the world, the most famous example of gold seasonality has to be the Indian wedding season.
Indians have a deep cultural affinity for gold, so in the autumn India’s farmers tend to invest their profits from harvest in gold. But even more gold is bought for the Indian weddings that happen late in the year during festivals, mainly in October and November. Something like 40% of India’s annual gold demand occurs in this short period of time. Wedding gold is often in the form of intricate 22-karat jewelry that the bride’s parents give her to secure her financial future and financial independence within her husband’s family.
Just as wheat traders use wheat’s seasonality to help them make trading decisions, gold and even gold-stock traders can use gold’s seasonality. With gold having definite seasonal tendencies at different times during the calendar year, investors and speculators can study it to better understand when seasonality helps or hinders their probabilities for success in launching new trades.
For my own seasonality analysis, I prefer a different approach to the classical way seasonality is analyzed in the futures world. Very long time horizons, often 15 to 40 years of price data, are crunched to build a typical futures seasonality chart. This approach is sound and has a great benefit. By considering seasonality across bulls and bears alike, secular trends are largely filtered out. This yields the most purely-distilled form of long-term seasonality that persists across all market conditions.
But this classical approach also has limitations. Prices behave very differently during secular bulls compared to how they behave in secular bears. Since I happen to be trading gold and gold stocks during today’s secular bull, I am most interested in how seasonality has affected gold in this bull only. While such a relatively myopic perspective dilutes seasonality by crossing it with the secular trend in force, these hybrid seasonals ought to be much more representative of what we can expect seasonally in this bull.
Hence the title of this essay, gold bull seasonals as opposed to the usual ultra-long-term gold seasonals. Our current gold bull was born from deep secular lows back in late 2000 and early 2001. So this hybrid take on gold seasonals includes every trading day in gold from January 2000 to June 2007, the period of time this gold bull has encompassed. It truly reveals the unique seasonal tendencies of gold in this bull to date.
The calculation methodology behind this first chart is simple in concept. For each calendar year, the daily gold price is indexed off the first trading day of that year which is assigned a value of 100. Then these individual annually-indexed results for all calendar years are averaged on a day-by-day basis. The result shows gold’s relative performance tendencies in an average bull-market year. It is pretty interesting.
In addition to these gold bull seasonals drawn in blue, I also rendered one standard deviation above and below them in yellow. Since these standard-deviation bands largely fall outside of the bounds of the main chart, an inset chart is included with the same vertical-axis intervals to illustrate the full range of these bands. The distance between these bands is important for interpreting the seasonals, as I’ll get into a little later.
On balance, gold has tended to rise 12% annually in this gold bull, from an indexed level of 100 in early January to 112 in late December. These are very impressive average gains which have already earned fortunes for contrarian investors and speculators.
But the most fascinating attribute of gold bull seasonals is they have formed a beautiful uptrend channel since 2000! There are actually parallel well-defined seasonal-support and seasonal-resistance lines which combine to make a seasonal uptrend channel. These simple technicals help traders understand when gold has the highest probability of being strong during the calendar year.
When trading a secular bull market, the greatest opportunities for profit are on the long side. So both investors and speculators seek to buy when prices are relatively low. The bullish tailwinds of gold seasonality are the strongest when gold approaches its seasonal support. Over the course of a typical calendar year, gold tends to hit its support three times and approach it on a fourth. These are the highest-probability-for-success times in seasonal terms to add new long gold and gold-stock positions.
The first support approach occurs in early January and the second in mid-March. It is provocative that these first two low points arose on schedule in calendar 2007. Gold hit an interim low in early January and another in early March. These two points proved to be the best times to add gold and gold-stock positions this year. Although it is true that H1’07 data is already factored into these seasonals, this is only one year out of eight so 2007’s influence on the chart line is relatively moderate. Check out last year’s pre-2007 chart which showed the same tendencies.
It is actually off of the mid-March seasonal low when gold’s first big seasonal rally launches. It tends to run until late May before a pullback into the summer. This year gold started its first big seasonal rally on schedule in mid-March, but it failed early in late April due to abnormally heavy central-bank gold sales. Still though, as we have seen in our Zeal trades this year, it was very profitable to buy gold and gold stocks near the March seasonal low.
After the usual May top, gold’s seasonally weakest time of the year comes into play. From May to late July, gold tends to grind sideways to lower. We certainly saw gold mirror these weak seasonal tendencies this year, as it has been fairly weak on balance since late April. While gold usually isn’t all that exciting in the summer doldrums, this necessary consolidation leads up to the biggest seasonal gains of the year.
In late July, in the next couple weeks here, gold tends to bottom and then start powering higher as autumn gold demand builds worldwide. The second big seasonal rally in gold occurs between late July and late September. This is similar in magnitude to the first one in the spring. The spring rally tends to take gold from 100 to 105 indexed, while the late summer one tends to run from 103 to 108 indexed. Both rallies are nice and often very beneficial for gold stocks, which are primarily driven by the gold price.
With August and September typically weighing in strong in seasonal terms, the obvious implication here is to get long gold, silver, and precious-metals stocks now if you are not already deployed. If late summer buying drives gold higher as usual, the more-speculative silver and precious-metals stocks will follow it up. Investors and speculators alike should take advantage of the seasonally weak summer to add positions ahead of the big seasonal rallies expected in the second half of the year.
After a brief seasonal pullback in early October, gold starts its third and greatest seasonal rally. Incidentally this happened last year too, as gold carved a major interim low in October way down near $562 an ounce! After this low, gold exhibits great seasonal strength in November and December. This third major rally continues into January and early February. Since this one takes gold from 105 indexed up to effectively 115 indexed if the January/February portion is added, this third rally is about twice as big as either of the first two.
Thus gold’s bull-market seasonals greatly increase the probability for success for long positions in the second half of the year starting in late July. From August to early February, traders have the opportunity to ride two of the three big seasonal rallies including the biggest by far that starts in October. While it remains to be seen if gold will reasonably mirror its established pattern for the rest of this year, it sure has been mirroring it fairly well since last October. I sure wouldn’t bet against these seasonal tendencies today.
While these seasonals are certainly exciting and encouraging since we are heading into the best part of the year for gold, they shouldn’t be considered apart from their yellow standard-deviation bands. The standard deviation is a dispersion measure of how far apart the underlying annually-indexed numbers are that are averaged on any particular day to produce the blue seasonal line.
You can get the same average of 50 with underlying data of 45 and 55 or 10 and 90, but the average is certainly more representative of the first set. Obviously the greater the dispersion in the underlying data, the less representative is its average and the less reliable it is as an indicator.
The yellow SD bands in these charts, particularly in the inset charts, offer a visual proxy of how dispersed the underlying annually-indexed gold data is. Narrow SD bands, such as from July to October, show a relatively tight cluster of data from individual bull-market years. This means gold’s seasonal tendencies during these times are more representative of the actual underlying data. Conversely the huge SD spread in May shows that gold’s seasonal tendencies during that month are less representative. Please keep this in mind as you digest these charts.
Overall, I think the SD bands for these seasonals seem reasonable. They never come close to diverging far enough to suggest no meaningful clustering of the underlying individual-year data. If there were extreme spikes the seasonality’s usefulness would be suspect, but thankfully there are not.
But one problem with applying SD bands to indexed data is they are always closest near where the indexing starts. Since the first trading day of every year is indexed at 100 for gold, the average of the first day across all years is always 100. The deeper into a year you get, the more each individual year’s index diverges and the greater the SD dispersion grows. To attempt to gain an alternative seasonal perspective less affected by this tendency, I also indexed gold on a monthly basis.
In this next chart, the same raw feedstock data since 2000 is indexed at 100 at the beginning of each calendar month instead of each calendar year. Then all the individual January indexes are averaged and plotted, then the February ones, and so on. This resulting chart shows the calendar-month tendencies of gold within this secular bull market. It helps illuminate gold seasonals from a different perspective than the annually-indexed approach, highlighting both similarities and differences.
Also, I didn’t want the resulting data to get lost within a line since each individual trading day is potentially important seasonally in this approach. So in both these charts I rendered the actual datapoints visible as dots and then connected them with thin lines. This atypical approach enables us to quickly see whether a sharp monthly move is merely a single connection between two widely-separated datapoints or a more statistically-meaningful solid connection running through multiple datapoints.
As you digest this chart, realize that each calendar month is a discrete individually-indexed unit. At the first trading day of every month the indexing starts anew so there are no inter-month connections in this perspective. January is totally independent of February and so on down the line here.
Interestingly this monthly seasonal approach confirms the four outstanding seasonal long points that the annual gold seasonality revealed. Gold tends to be weak seasonally in early January, mid-March, late July, and mid-October. These four points, as well as a fifth in early June, represent the times of the year when gold is most likely to be seasonally weak and hence the highest-probability-for-success times to add long positions.
In addition, seven months are classifiable as seasonally strong. If gold gained more than 1% in a given month on average, if it closed a month above 101 indexed, then it is a strong month. 1% monthly gains correspond well with the 12% annual gains seen above in the annual seasonal analysis. January, April, May, August, September, November, and December all weighed in as strong by this definition.
Encouragingly, the very best calendar months of the year seasonally are all clustered in the second half. We are heading into this typically exciting time of the year for gold. November was the top-ranked month seasonally, running up to 103 indexed on average. September was not far behind, nearly hitting 103. 3% monthly gains in gold are huge and often translate into excellent returns in silver and PM stocks too.
The third and fourth highest-performing months for gold have been August and December respectively. Both saw monthly gains just shy of 102 indexed. 2% monthly rallies are certainly nothing to scoff at either. The really great thing to realize this time of year is that we are now rapidly approaching these seasonally exceptional months of August, September, November, and December. If gold holds true to form in 2007, we should be in for a bullish and profitable second half.
Obviously the broad strategic message behind these gold bull seasonals is that probabilities favor getting long in the summer to ride the major seasonal gold rallies between August and January. Buy gold, silver, and PM stocks in the summer doldrums when they are out of favor and hold for the expected gains over the next six months.
But there is an important caveat to bear in mind regarding seasonal analysis. While seasonals do offer valuable insights into probabilities that are not readily apparent by other means, they are a secondary indicator at best. Sentiment is much more important than seasonal technicals on a short-term basis and it can easily override seasonals. So seasonals should only be used as a secondary confirmation of primary indicators and not as a primary trading tool by themselves.
A key example occurred last year. Oil’s bull-market seasonals a year ago showed a huge seasonal tendency for oil to soar in August and September. Many traders, including me, were heavily deployed in oil stocks and options in advance of this seasonal tendency. But when no hurricanes materialized in July and August, sentiment plunged and traders sold oil aggressively. Seasonal tendencies established over eight years failed to hold in the second half of 2006. Bad sentiment overwhelmed the positive seasonals and traders took big losses in oil-related positions.
Thankfully this year the gold seasonals do line up with the already bullish gold fundamentals and technicals. And sentiment in the PM realm has been pretty rotten in recent months, typical of bottoming periods before big uplegs launch. With the primary indicators suggesting a major gold, silver, and PM-stock upleg is due, the secondary confirmation provided by the gold seasonals is very welcome. It increases our odds for success in long PM positions.
At Zeal we have been anticipating the inevitable return of strong bull-market conditions to the precious metals and we have been trading accordingly. We have been aggressively buying elite gold and silver miners and explorers on weakness and working to get fully deployed in the PM sector. To mirror our trades and enjoy practical cutting-edge commodities-stock-trading analysis, please subscribe to our acclaimed monthly newsletter today!
The bottom line is gold has already established very definite seasonal tendencies in its bull to date. Gold tends to be weakest in the early summer leading up to the end of July. Then it tends to rally in increasing strength through the end of the year and into January. With primary indicators suggesting that such a rally is indeed due, the secondary confirmation provided by the gold seasonals is very welcome.
If you are bearish on gold today like the vast majority of traders, realize that contrarians must trade against the crowd. The time to least doubt any asset’s near-term prospects is when the most people doubt them the most. Gold typically tends to get beat up in the early summer months just like it did this year. But after the summer doldrums its big seasonal rallies make the wait well worth it.
Adam Hamilton, CPA July 13, 2007 Subscribe at www.zealllc.com/subscribe.htm