In my search for probabilistic gains (methodology here), I have already written articles about seasonal strategies on stock indexes (sample here). In this one I would like to show that seasons also have an influence on metals.
The following table summarizes the historical monthly trends for the price of Futures contracts on 8 metals. For every month and every metal, a number represents the magnitude of the trend. For example, 1 means than the monthly average return is roughly between 1% and 2%; -3 means that it is between -3% and -4%. An empty cell means no significant trend (between -1% and 1% on average).
The "Y" column gives the number of years in historical data, which is not the same for all metals. The "ETF" column gives the most liquid ETF on each metal.
| Metal | Y | ETF | Jan | Feb | Mar | Apr | May | Jun | Jul | Aug | Sep | Oct | Nov | Dec |
| Gold | 40 | GLD | 1 | 1 | 1 | 2 | 1 | 1 | ||||||
| Silver | 37 | SLV | 2 | 1 | -2 | 2 | 3 | -1 | 1 | |||||
| Plat. | 32 | PPLT | 2 | 1 | -1 | 2 | -1 | 1 | 1 | |||||
| Pall. | 30 | PALL | 5 | -1 | -1 | |||||||||
| Copper | 37 | JJC | 2 | 3 | -2 | -1 | 2 | 2 | 1 | |||||
| Alu. | 17 | JJU | 2 | -1 | -1 | -2 | -1 | -1 | 1 | 1 | ||||
| Lead | 13 | LD | 2 | 3 | -2 | -2 | 4 | 1 | 3 | |||||
| Nickel | 15 | JJN | 4 | 3 | -3 | -1 | 2 | 3 |
You may draw as many conclusions and hypotheses as you wish from this table; I will just focus on what appears the most important to me:
- There is a bullish seasonal trend on metals from November to February.
- Gold has no calendar month below -1% on average for 40 years.
- The highest historical probability of being whipsawed is in platinum, silver and copper.
- The highest historical gain of an hypothetic seasonal swing trade is in lead and nickel. Caution: they're also the metals with the shortest price history.
I would like to insist on 2 important facts:
- It is not a strategy description: the moves against the historical trend may be strong in metals.
- The numbers do not have the probabilistic value of a backtest: the data availability period is different from a metal to another.
However, these historical facts might be used as a filter or a confirmation to improve one's timing. For example, I would avoid selling short palladium and nickel in January, or buying lead and aluminum in March, except when all other indicators strongly confirm the trade. On the other hand, a weak technical "buy" signal on palladium in January would be reinforced by the seasonal trend. As a long term investor in gold and silver, I can also confirm that it's usually better to accumulate after the summer.
Additional disclosure: I am long physical metal funds in gold and silver.
