Seasonality is one of those tools that have so many ways of distorting data or presenting results in a bad way. It starts with how many years of history you are looking at, e.g. are you considering the last 15 years in US stocks as being the new "normal" or are you choosing a longer timeframe? Next, the number of observations is very small, especially if you believe that market characteristics change over long periods of time. And there are huge outliers in single years like 2008 or 2020. Calculating seasonality based on the median of observations and not on the mean to filter out this effect leads to a completely different pattern. And the list goes on.
There's a small number of seasonalities that make sense. JPY strength after the end of the fiscal year on March 31 in Japan is one that comes to mind. But even that is not working every year (like right now) despite a sound reason behind it.
the problem with this article and the comments is that you don't really know anything about seasonality. Using seasonality profitably is relatively easy if you use the right tools. Seasonality works very well for individual tickers, indices and stocks in certain time frames. For example, earnings usually produce very stable recurring price patterns for stocks.
But you need the right tools to find these times of the year for each stock. I use Seasonality.Ai.
Agree, 2020 Covid ruins every chart- thanks for the insight!
Seasonality is one of those tools that have so many ways of distorting data or presenting results in a bad way. It starts with how many years of history you are looking at, e.g. are you considering the last 15 years in US stocks as being the new "normal" or are you choosing a longer timeframe? Next, the number of observations is very small, especially if you believe that market characteristics change over long periods of time. And there are huge outliers in single years like 2008 or 2020. Calculating seasonality based on the median of observations and not on the mean to filter out this effect leads to a completely different pattern. And the list goes on.
There's a small number of seasonalities that make sense. JPY strength after the end of the fiscal year on March 31 in Japan is one that comes to mind. But even that is not working every year (like right now) despite a sound reason behind it.
Very interesting! Dear Joachim, to what indexes you refer when you put US, Uk and Europe market to the chart ? Thank you in Advance,
S&P500 in the US, FTSE 100 in the UK and Stoxx Europe
the problem with this article and the comments is that you don't really know anything about seasonality. Using seasonality profitably is relatively easy if you use the right tools. Seasonality works very well for individual tickers, indices and stocks in certain time frames. For example, earnings usually produce very stable recurring price patterns for stocks.
But you need the right tools to find these times of the year for each stock. I use Seasonality.Ai.