RELATED TOPICS --> | Sell in May and go away | Coupling Seasonality with Market Timing |
| Login to see this months' SATC data |
This section will explain how to interpret the PowerStocks Seasonality Actuarial tables and charts (SATC) that we maintain and update on a monthly basis. These keep track of the JSE ALSH index growth characteristics for months of the year, inter-month, days of the week and turn of the year, stretching back 25 years to 1985.
The purpose of SATC is to enable our subscribers to observe and possibly capitalize on seasonal anomalies in their investment strategies. Contrary to the "Efficient Market" hypothesis, strong, proven and undeniable seasonal anomalies exits in all major stocks markets around the world and the JSE is no exception.
1. INTRODUCTION
We have already shown in our "Sell in May and go away" research note than the JSE exhibits strong monthly growth seasonality traits that are slightly different to the US markets. We showed how one could implement a market timing investment strategy to exploit this seasonality and slightly beat a JSE buy-and-hold strategy whilst only being in the JSE for 68% of the time.
We then introduced an "Adaptive Seasonality Module" to our flagship SUPERModel Long Term Investing Market Timing signal, and launched SUPERModel-II which exploited seasonality effects to raise returns from 110,000% to over 200,000% over a 32 year investment period. We documented this in our "Coupling Seasonality with Market Timing" research note.
Since seasonal anomalies are powerful enough to be exploited in various ways, we have launched our SATC product for our BASIC SUBSCRIPTION subscribers which regularly updates actuarial (statistical) tables of seasonality in finer detail than merely looking at average monthly returns. The first iteration of this service is to provide greater granularity into individual monthly seasonal traits, by examining inter-month as well as month of the year seasonality, whilst the 2nd iteration will look at days of the week and weeks of the year seasonal anomalies.
SATC is published in the SATC Dashboard (available from the green link at top of page) for paying subscribers only, and are updated every month to reflect the effects of more recent data into the actuarial tables. This caters for seasonal trends that change, wane and strengthen over time.
2. MONTHS OF THE YEAR TABLES
The first iteration of SATC still includes the now familiar monthly average returns chart published weekly in WJP for subscribers as shown below:
3. INTER-MONTH TABLES
Our service has now been enhanced to include an additional table published each month (for the current month) as shown in the more recent chart for DECEMBER below:
We can see the new SATC tables include finer granularity on individual months in questions. From the first chart we can see December is traditionally a strong month for the JSE, and the 2nd chart shows us that it is a consistently good month with lower than normal standard deviation and the highest Sharpe ratio (average returns divided by deviation of returns) of all the months. It shows how the December month has performed for each of the last 25 years and tells us it is exhibiting a 68% probability of producing a positive return (excluding brokerage costs.)
3.1 INTERPRETING THE CHART
We will now go over each aspect of the above chart in detail.
The first trend chart on the left of the table shows the specific returns (growth) of the ALSH index (ALSI) for the month in question. The growths correspond to the left axis on the chart. It measures the growth from the first trading day of the month to the last. This provides useful visual cues to the volatility of the returns (consistency) as well as trend.
The green overlay line depicts the rolling 10-year average of returns for the month whilst the yellow overlay line depicts the rolling 5-year average of returns. This helps discern if the medium term trend is deviating from the longer term trend. We note that Decembers' medium and long term trends are close together and that Decembers' powerful returns are waning ever so slightly over the last 7 years, from 5% to the current 2.83%. We note that despite Decembers' strong showing, negative Christmas periods are not uncommon.
The beige line depicts the ALSH index's' close on the last trading day of the month in question, and corresponds to the right hand axis on the chart. This is useful to observe the behavior of the month in relation to what was going on in the JSE at that period.
The last bar in the chart, (for December) in the above example reflects data up until the day the chart has been published, in this instance it reflects December up and until 10 Dec 2009. Obviously the current reading for 2009 will change as December marches on.
Now look at the actuarial table on the bottom right hand corner of the picture. The values displayed have the following meanings:
NPER : The number of years (periods) depicted in the analysis
%NEG : The % of years the month produced a negative result (a loss).
%POS : The % of years the month produced a positive result (profit)
%>1 : The % of years the month produced more than 1% return (enough to cover brokerage)
%>2 : The % of years the month produced more than 2% return
%>5 : The % of years the month produced more than 5% return
AVG CHG : Average % return over NPER years
STDEV CHG : Standard deviation of each return over NPER years (volatility of readings)
HI CHG : Highest return recorded over NPER years
LO CHG : Lowest return recorded over NPER years
SHARPE : AVG CHG divided by STDEV CHG - an indication of confidence in readings
The bar chart on top of the actuarial table shows some of the above data in graphical form (a sort of Yield Curve if you like). The last two bars on the right represent HI CHG and LO CHG multiplied by 2 for easier viewing on the chart scale.
Finally, you will note that the values in the Actuarial Table have colour coded bars representing how the number you are looking at compares to all the other months in the calendar. % NEG and LO CHG (with their associated negative connotations) are shown in orange whilst the rest is shown in green. If the coloured bars are small, it means the number is the lowest among all the months. If the bar extends all the way to the right of the cell it means the number is the highest of all the months. If the bar stops midway it means the number you are looking it is the average of all the other months.
Using this system you can quickly see that December has the lowest LO CHG reading of all the months and the highest %POS reading. Also note that December has the highest SHARPE reading of all the other months, which means its the most "persistent" of all the months with regards to its seasonality traits.
3.2 A TALE OF TWO HALVES
As with trends in global markets, the JSE exhibits very powerful, consistent and important seasonal traits between the first and second half of each month, although they differ somewhat to overseas markets.
For this reason SATC charts are also displayed for each months' two halves. A sample is shown below for the first half in December:
This is identical to the chart that covers the whole month, except that it runs from the first trading day in the month to whatever point in the month closest depicts the exact half-way mark to ensure each half has an identical number of trading days. This point differs for each month and indeed for each month in each year due to public holidays, weekdays versus weekends etc. and is not a simple 1st to the 15th assumption!
When comparing this chart to the chart for the entire month we note that the first half of December is statistically much weaker than the second half, and in fact displays far less confidence (much more volatility in the sample) than for the whole year (less than half the Sharpe ratio.) The first half of December is also subject to poor win/loss ratios as shown by a %POS of only 44%.
As with the data shown for the whole year, the actuarial chart also includes colored intensity bars to represent how a value compares to the other months. The only difference is that this time we are comparing each value to all the other months' first half figures. You can see from the size of the %NEG yellow intensity bar that the first half of December is just about the poorest 1st-half performer of all the months of the year. It also has the largest draw-down of -5.69%
To make up the gap between this poor performance and that for the whole month, you can gather that the 2nd half of December is going to have to require a fantastic win ratio with some pretty impressive confidence levels.
We note from the green and yellow rolling averages that this poor performance of the first half of December is relatively consistent over medium and long periods of time even though the blue bars seem to swing between positive and negative on a random basis for the last 9 years.
SATC will also display an identical chart for December representing the 2nd half of the year.
3. SATC UPDATES
The SATC charts and tables will be updated on a monthly basis to ensure they always reflect the most recent months' data as possible. However as the chart is published at the beginning of the month (as this is the month you will be interested in) you will only see the new updated chart for the previous month in 11 months time (when it counts.)
RELATED TOPICS --> | Sell in May and go away | Coupling Seasonality with Market Timing |
| Login to see this months' SATC data |
<---- GO BACK TO HOW-TO MAIN MENU
| Login to see this months' SATC data |
This section will explain how to interpret the PowerStocks Seasonality Actuarial tables and charts (SATC) that we maintain and update on a monthly basis. These keep track of the JSE ALSH index growth characteristics for months of the year, inter-month, days of the week and turn of the year, stretching back 25 years to 1985.
The purpose of SATC is to enable our subscribers to observe and possibly capitalize on seasonal anomalies in their investment strategies. Contrary to the "Efficient Market" hypothesis, strong, proven and undeniable seasonal anomalies exits in all major stocks markets around the world and the JSE is no exception.
1. INTRODUCTION
We have already shown in our "Sell in May and go away" research note than the JSE exhibits strong monthly growth seasonality traits that are slightly different to the US markets. We showed how one could implement a market timing investment strategy to exploit this seasonality and slightly beat a JSE buy-and-hold strategy whilst only being in the JSE for 68% of the time.
We then introduced an "Adaptive Seasonality Module" to our flagship SUPERModel Long Term Investing Market Timing signal, and launched SUPERModel-II which exploited seasonality effects to raise returns from 110,000% to over 200,000% over a 32 year investment period. We documented this in our "Coupling Seasonality with Market Timing" research note.
Since seasonal anomalies are powerful enough to be exploited in various ways, we have launched our SATC product for our BASIC SUBSCRIPTION subscribers which regularly updates actuarial (statistical) tables of seasonality in finer detail than merely looking at average monthly returns. The first iteration of this service is to provide greater granularity into individual monthly seasonal traits, by examining inter-month as well as month of the year seasonality, whilst the 2nd iteration will look at days of the week and weeks of the year seasonal anomalies.
SATC is published in the SATC Dashboard (available from the green link at top of page) for paying subscribers only, and are updated every month to reflect the effects of more recent data into the actuarial tables. This caters for seasonal trends that change, wane and strengthen over time.
2. MONTHS OF THE YEAR TABLES
The first iteration of SATC still includes the now familiar monthly average returns chart published weekly in WJP for subscribers as shown below:
3. INTER-MONTH TABLES
Our service has now been enhanced to include an additional table published each month (for the current month) as shown in the more recent chart for DECEMBER below:
We can see the new SATC tables include finer granularity on individual months in questions. From the first chart we can see December is traditionally a strong month for the JSE, and the 2nd chart shows us that it is a consistently good month with lower than normal standard deviation and the highest Sharpe ratio (average returns divided by deviation of returns) of all the months. It shows how the December month has performed for each of the last 25 years and tells us it is exhibiting a 68% probability of producing a positive return (excluding brokerage costs.)
3.1 INTERPRETING THE CHART
We will now go over each aspect of the above chart in detail.
The first trend chart on the left of the table shows the specific returns (growth) of the ALSH index (ALSI) for the month in question. The growths correspond to the left axis on the chart. It measures the growth from the first trading day of the month to the last. This provides useful visual cues to the volatility of the returns (consistency) as well as trend.
The green overlay line depicts the rolling 10-year average of returns for the month whilst the yellow overlay line depicts the rolling 5-year average of returns. This helps discern if the medium term trend is deviating from the longer term trend. We note that Decembers' medium and long term trends are close together and that Decembers' powerful returns are waning ever so slightly over the last 7 years, from 5% to the current 2.83%. We note that despite Decembers' strong showing, negative Christmas periods are not uncommon.
The beige line depicts the ALSH index's' close on the last trading day of the month in question, and corresponds to the right hand axis on the chart. This is useful to observe the behavior of the month in relation to what was going on in the JSE at that period.
The last bar in the chart, (for December) in the above example reflects data up until the day the chart has been published, in this instance it reflects December up and until 10 Dec 2009. Obviously the current reading for 2009 will change as December marches on.
Now look at the actuarial table on the bottom right hand corner of the picture. The values displayed have the following meanings:
NPER : The number of years (periods) depicted in the analysis
%NEG : The % of years the month produced a negative result (a loss).
%POS : The % of years the month produced a positive result (profit)
%>1 : The % of years the month produced more than 1% return (enough to cover brokerage)
%>2 : The % of years the month produced more than 2% return
%>5 : The % of years the month produced more than 5% return
AVG CHG : Average % return over NPER years
STDEV CHG : Standard deviation of each return over NPER years (volatility of readings)
HI CHG : Highest return recorded over NPER years
LO CHG : Lowest return recorded over NPER years
SHARPE : AVG CHG divided by STDEV CHG - an indication of confidence in readings
The bar chart on top of the actuarial table shows some of the above data in graphical form (a sort of Yield Curve if you like). The last two bars on the right represent HI CHG and LO CHG multiplied by 2 for easier viewing on the chart scale.
Finally, you will note that the values in the Actuarial Table have colour coded bars representing how the number you are looking at compares to all the other months in the calendar. % NEG and LO CHG (with their associated negative connotations) are shown in orange whilst the rest is shown in green. If the coloured bars are small, it means the number is the lowest among all the months. If the bar extends all the way to the right of the cell it means the number is the highest of all the months. If the bar stops midway it means the number you are looking it is the average of all the other months.
Using this system you can quickly see that December has the lowest LO CHG reading of all the months and the highest %POS reading. Also note that December has the highest SHARPE reading of all the other months, which means its the most "persistent" of all the months with regards to its seasonality traits.
3.2 A TALE OF TWO HALVES
As with trends in global markets, the JSE exhibits very powerful, consistent and important seasonal traits between the first and second half of each month, although they differ somewhat to overseas markets.
For this reason SATC charts are also displayed for each months' two halves. A sample is shown below for the first half in December:
This is identical to the chart that covers the whole month, except that it runs from the first trading day in the month to whatever point in the month closest depicts the exact half-way mark to ensure each half has an identical number of trading days. This point differs for each month and indeed for each month in each year due to public holidays, weekdays versus weekends etc. and is not a simple 1st to the 15th assumption!
When comparing this chart to the chart for the entire month we note that the first half of December is statistically much weaker than the second half, and in fact displays far less confidence (much more volatility in the sample) than for the whole year (less than half the Sharpe ratio.) The first half of December is also subject to poor win/loss ratios as shown by a %POS of only 44%.
As with the data shown for the whole year, the actuarial chart also includes colored intensity bars to represent how a value compares to the other months. The only difference is that this time we are comparing each value to all the other months' first half figures. You can see from the size of the %NEG yellow intensity bar that the first half of December is just about the poorest 1st-half performer of all the months of the year. It also has the largest draw-down of -5.69%
To make up the gap between this poor performance and that for the whole month, you can gather that the 2nd half of December is going to have to require a fantastic win ratio with some pretty impressive confidence levels.
We note from the green and yellow rolling averages that this poor performance of the first half of December is relatively consistent over medium and long periods of time even though the blue bars seem to swing between positive and negative on a random basis for the last 9 years.
SATC will also display an identical chart for December representing the 2nd half of the year.
3. SATC UPDATES
The SATC charts and tables will be updated on a monthly basis to ensure they always reflect the most recent months' data as possible. However as the chart is published at the beginning of the month (as this is the month you will be interested in) you will only see the new updated chart for the previous month in 11 months time (when it counts.)
RELATED TOPICS --> | Sell in May and go away | Coupling Seasonality with Market Timing |
| Login to see this months' SATC data |
<---- GO BACK TO HOW-TO MAIN MENU