Seasonality & Charts in TLT, SPY, GDX & BTC

I’ve been thinking more about seasonality lately as we come into 4th quarter. First, it is important to think about what seasonality is telling you. At a surface glance, it’s just telling you the number of times that the underlying stock ticker went up from beginning of month to end of month over the last 5 years. If it’s 4 to 5 times (80-100%), there is a good indication that there’s some seasonal reason you want to be long or at least not short. If it’s 0-1 times (0-20%) perhaps you want to be short or at least trim back longs. If it’s 2-3 times (40-60%) then look to other factors.

Why would seasonality actually work? There are a lot of financial forces that operate on a seasonal calender. The obvious ones are tax time, christmas shopping season, summer vacation season, weather-related affects (heating in winter, hurricanes in fall, etc). There are also reasons having to do with mutual fund activity such as end-of-quarter portfolio rebalancing or quarterly reporting.

Note that this is the site I’m using to get these seasonality charts. There are many other charts you can explore on this site as well:

Let’s start with long dated treasury bonds (TLT). Note that TLT rises in value as treasury yields drop, particularly in the 10-30 year range.

Why do US Treasury Bonds lose value ‘every’ August !? [Ep. 115a] – YouTube

In the link above, Jeff Snider and Emil Kalinowski explain how banks use the 4th quarter to adjust their balance sheets to hit whatever score they are aiming for in the Basel 3 regulatory system, so they often unwind derivatives and such in a way that causes bond yields to rise. This is certainly consistent with the typical drops in September & October.

The chart below helps you see the underlying trend and spot the anomalies. The anomalies above are very hard to spot, like the one in October 2017 when TLT bottomed at the end of September but barely rose in October – so the idea that these are weak months for treasuries based on repeating seasonal patterns is solid.

For the trend, TLT is barely above a declining 200-day moving average (above you see the 50-week moving average which is still rising). The current direction is still down as the TLT peaked with a major risk-off event in March 2020 and has been falling back toward its previous long-term rising trend.

Note that the primary reason I am interested in treasuries is because they tend to spike in a risk-off event. If all goes well, interest rates will steadily rise, but there are fragilities in the financial system which make periodic risk-off events more likely. The article below does a good job describing them, though it is a bit complex for the casual reader.

The Volatility Squeeze – by TheLastBearStanding – The Last Bear Standing (

That being said, I should not be adding to my long TLT position until at least the end of October. I added a bit last week, which I wouldn’t have done if I considered seasonality first, but I’ll hold it through because it is still my primary hedge in case of a significant risk-off event (like a 20% drop in the S&P 500).

Now onto the S&P 500, using it’s big ETF proxy SPY:

You can see on the seasonal charts, that the S&P 500 almost always goes up month-on-month, though it is particularly strong both in Summer and in the November run-up to the Christmas shopping season. Last year, the S&P 500 struggled in September & October and the same is happening again this year. A lot of people on twitter are calling a top right now, which makes be more inclined to be long, especially through November.

Note that sentiment is a lot more about positioning than anything else – and there certainly has been a lot of put buying in the indexes lately. This is important because these hedges tend to be sold into a drop forming the next layer of support before the market climbs higher. The S&P 500 has significant regular price-insensitive inflows from both corporate buybacks and regular 401k contributions, so you can run out of active-investor sellers who then start to buy back in as the index reverses creating violent upward corrections. Around mid October, I should have significantly more long exposure in the S&P500, which I will express through more call options in bullish charts I see here: Home – Dereck’s Trades (

I realize that the S&P 500 is very risky with it’s elevator-like drops and extreme valuations, which is why I am not positioned heavily here, why I have significant hedges in TLT, and why I think a relatively small number of call options might be the best way to get upside exposure.

Now let’s move on to my biggest position, precious metals miners, shown using the gold mining ETF GDX:

One thing I would like to immediately point out is the similarity between the 5-year charts GDX and TLT. Both were roughly flat with a minor rise in 2019, a spike in early 2020, and a consolidation lower ever since. In my opinion, this is because both are major hedges for local currencies that are held by central banks around the world. This gives them unique properties such as tending to well in risk-off scenarios.

As I have explained in previous posts, I follow Jeff Snider’s view that the world has been in a regime of top-tier collateral shortages and Eurodollar shortages since 2008. Central banks use “forward guidance” as their primary policy, with quantitative easing to give them the illusion of great power even though it does little, and their goal is to stoke the “animal spirits” of the market to make everyone more bullish with an inflationary bias. The result has been a series of compressed runs and panics in the real economy which has ever-lower employment and trend growth. Meanwhile financial asset declines are truncated by routinely scaring the “bears” and cheering the “bulls” while price-insensitive passive flows continue to drive valuations higher. As such, I believe that bonds and gold are both showing the same thing, which is a higher valuation related to asset scarcity and deflationary tendencies.

Precious metals miners are still my favorite holdings because miners in general tend to shut down production and consolidate after years of price declines (see 2011-2019), while ramping up production to meet higher prices and demand takes a lot of time and money, leading to long busts followed by long runs – and I believe that we are in the early stages of a long run through 2030. Add to that the increasing relevance of gold in the international system as many countries such as China attempt to move past the aging petro-dollar system and their reliance on US dollars for international trade. China looks to Russia and Iran as examples of what they want to avoid in the event that the US turns the weapon of sanctions their way.

Back to the seasonality chart though. The big declines are in February are likely centered around the Chinese New Year, as the strongest buyers of physical gold are off the market. As for the weakness in autumn, particularly September, my guess is that it has some similarities with the decline in bonds during that period. Also, Thanksgiving (end of November) is thinly traded week in the US with an important expiration in Futures that is always preceded by large amounts of contract selling. This creates an annual event that is always so oversold that it sets up December to be an easy gain.

I’ll finish this off with Crypto. Although I am currently betting on Etherium, I am using the Bitcoin seasonality chart as a proxy because they tend to trade in tandem, and Bitcoin was a more mature market in 2017:

To be totally honest, I should be using a log chart for this one but I can’t figure out how to make log charts in Yahoo Finance, and the log chart in won’t let me set the date range. The reason for a log chart is simply because the a standard chart artificially shrinks the price fluctuations of the past. On a log chart, a 2x return is the same length whether it is $1,000 to $2,000 or $15,000 to $30,000 – and this makes sense because you experience a double the same way as an investor. If you aren’t into crypto, think of it as viewing a 50-year chart of the Dow Jones Industrial Average – a log chart shows the extreme volatility of the 1970’s on par with the moves today, whereas a standard scale would make it look like nothing really happened in the stock market until the early 1990’s.

As far as the chart goes, it’s easy to see the similarities between the 1st and second peaks of 12/2017 and 2/2108 as similar to the peaks in 4/2021 and 8/2021, concluding that we passed a blowoff top and we should avoid this asset class for another year. I get that, which is why my position is small. However, there is also an argument to be made that bitcoin is retracing its paths in 2013 and 2017 which included significant corrections and volatility mid-year before peaking in December.

As for seasonality, October has been a winning month for crypto for whatever reason. November has been iffy, but it has been a fantastic month for stocks and I consider Crypto a highly risk-on asset, so I will hold it through October at least. Perhaps I should then sell when it hits significant resistance on the chart (like if bitcoin just manages to hit it’s $60k peak at the end of October), then buy back in on a correction just to see if we get that parabolic year-end rise, but I don’t currently plan on holding it past year-end in case of a 4-year cycle repeat.

I’ll keep watching crypto though, the weird thing about tradeable asset classes such as stocks, commodities and cryptocurrencies is that they are always changing. There are plenty of reasons to expect different behavior now than during 2013 or 2017 because we have a large financial presence complete with futures markets and ETF’s that simply didn’t exist in those cycles. The bitcoin halving cycle was a primary driver of the movements back then, and there are many reasons to see this decrease in relevance as other forces take over – though for now I lean toward viewing these new forces as amplifiers of the current cycle, as more momentum chasers push us toward a final year-end pump.

Here’s where my portfolio landed:

  • HEDGES (20.0%)
    • 18.5% TLT Calls
    • 1.5% EEM Puts
    • 9.1% AG (Silver), mainly shares & some calls
    • 5.3% SAND (Gold, Silver & others), all calls
    • 4.8% EQX (Gold), mainly calls & some shares
    • 4.5% LGDTF (Gold)
    • 4.0% SILV (Silver)
    • 3.9% SILVRF (Silver)
    • 4.0% MTA (Gold & Silver)
    • 3.1% MGMLF (Gold)
    • 1.9% RSNVF (Silver)
    • 2.1% SSVFF (Silver)
    • 1.6% LWDEF (Gold)
    • 0.9% WPM (Gold, Copper & Silver), all calls
    • 0.7% GOLD (Gold, Copper), all calls
  • URANIUM (17.6%)
    • 9.3% CCJ, covered calls sold on all of it
    • 7.1% UUUU, covered calls sold on all of it
    • 1.2% BQSSF
  • COPPER & NICKEL (5.6%)
    • 5.6% NOVRF
  • CANNABIS (10.0%)
    • 1.7% CRLBF
    • 1.8% GTBIF
    • 1.9% TRSSF
    • 1.5% CCHWF
    • 1.7% AYRWF
    • 1.5% TCNNF
  • CRYPTO (2.2%): All ETH
  • CASH (-2.8%)

I did some buying last week as miners were hit hard. This includes small adds to precious metals miners, another Jan 2023 TLT call, a little more Cannabis to even out my holdings, and a touch more in Dereck’s trades (sticking with the bullish ones).

I also bought significantly more CCJ and UUUU. I’ve been worried about completely losing my stake in both as everything I held in these had covered calls expiring in 2 weeks with trading very near the strike prices. If the calls expire worthless and I keep my stake, I’ll be happy to keep a higher allocation given the success of the Sprott Uranium fund as a major catalyst in the space. If my shares get called, I managed to pick up an initial new stake significantly below the strike prices they were called at.

That buying has pushed me into margin a little bit (hence the negative cash), so I should add sparingly in coming weeks. Good luck and happy trading!

About johnonstocks

I've been trading stocks since 2003, active on Motley Fool's discussion boards and using first Hidden Gems, then Global Gains. I no longer have the newsletters, but I keep up on the WSJ and read David Rosenberg everyday at Education: CFA level 2 candidate MBA-focus in Finance, Marshall, University of Southern California - expected Dec 2010. BS Mechanical Engineering, UC San Diego, June 2002
This entry was posted in Uncategorized. Bookmark the permalink.

1 Response to Seasonality & Charts in TLT, SPY, GDX & BTC

  1. This has been a pleasure to read, many thanks.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s