Similar to last month, first 15 trading days and last 5.

SPY – a bit mixed for the first trading days, some strength and then mixed again. Somewhat strange is that the average is a loss of .3%, but the median is a gain of .98%. 3 times out of 21 the low of the month is above the July close (and this has not happened since 2000). 2 out of 20 years has the high being lower than the July close. After the first couple of days of August being positive, half the time the end of month is higher and half its lower (than the close of the 2nd trading day). Overall, nearly half the time the low of the month is in the first 5 trading days (9/20), with 8/20 times the high is in the last 5 days of the month. Median low of the month is down 3%, median high for the month is 2.6%.

For IWM, the August on average and median is positive, with median bottoming around day 8.

## August Daily Performance 2013 edition

August 3, 2013## July Daily Performance 2013 edition

June 28, 2013Similar to last month, first 15 trading days and the last 5.

SPY:

The low of the month is doesn’t seem to have much of a pattern, but half of the SPY highs have been in the last 5 trading days of the month. About 1/4 of the time the low of the month is above June’s close.

July 2002 was a rough month.

## How often SPY open to high is greater than open to low by day of week, year

June 18, 2013I’m not sure if it isn’t just statistical noise, but this year SPY open to high is quite often larger than open to low. I decided to run two tests – one for the first six months of the year and the other for the last six months. I looked by day of the week for each year. For this year, about 70% of the time on Tuesdays and Thursday’s, the open to high is larger than open to low. Back in 1995 was the last time when 2 days of the week were near 70% for the first half of the year and 2004 for the second half. There seems to be some mean reverting as I don’t see any cases where the first half was greater than 60% and second half was also greater than 60% for the same year/day of the week. Over the years, Friday’s seem to have the fewest cases of open to high greater than open to low, particularly in the first half of the year. Wednesdays for the first half of the year normally fall in the 40-60%, but the second half of the year about one third of the time the open to high is smaller than open to low. I did not look at where the close fell. Generally speaking the second half of year seems to have more cases of open to high being smaller than open to low.

First the first six months – percent of days where the open to high is greater than open to low:

MON | TUE | WED | THU | FRI | |

2013 | 43% | 71% | 50% | 71% | 52% |

2012 | 55% | 46% | 50% | 42% | 44% |

2011 | 57% | 54% | 46% | 54% | 42% |

2010 | 39% | 54% | 54% | 44% | 33% |

2009 | 57% | 54% | 40% | 44% | 44% |

2008 | 48% | 44% | 50% | 65% | 24% |

2007 | 45% | 60% | 65% | 58% | 44% |

2006 | 50% | 38% | 77% | 58% | 36% |

2005 | 57% | 58% | 42% | 46% | 25% |

2004 | 52% | 62% | 54% | 40% | 42% |

2003 | 39% | 56% | 52% | 50% | 56% |

2002 | 45% | 44% | 58% | 54% | 44% |

2001 | 45% | 50% | 42% | 50% | 28% |

2000 | 43% | 54% | 46% | 38% | 40% |

1999 | 39% | 35% | 46% | 48% | 29% |

1998 | 43% | 62% | 52% | 48% | 36% |

1997 | 50% | 64% | 40% | 42% | 52% |

1996 | 65% | 46% | 46% | 54% | 48% |

1995 | 70% | 69% | 46% | 50% | 44% |

1994 | 58% | 42% | 44% | 38% | 50% |

for the last six months:

MON | TUE | WED | THU | FRI | |

2013 | |||||

2012 | 48% | 42% | 24% | 60% | 38% |

2011 | 35% | 65% | 35% | 44% | 52% |

2010 | 42% | 50% | 62% | 42% | 58% |

2009 | 64% | 38% | 56% | 50% | 50% |

2008 | 32% | 44% | 44% | 38% | 64% |

2007 | 38% | 52% | 44% | 32% | 46% |

2006 | 71% | 60% | 58% | 40% | 31% |

2005 | 35% | 46% | 50% | 52% | 41% |

2004 | 42% | 58% | 73% | 42% | 62% |

2003 | 60% | 52% | 33% | 58% | 52% |

2002 | 42% | 48% | 52% | 50% | 62% |

2001 | 50% | 42% | 38% | 50% | 48% |

2000 | 67% | 44% | 31% | 52% | 42% |

1999 | 46% | 23% | 50% | 38% | 50% |

1998 | 44% | 50% | 33% | 46% | 42% |

1997 | 44% | 52% | 56% | 38% | 36% |

1996 | 42% | 33% | 48% | 54% | 54% |

1995 | 54% | 40% | 46% | 48% | 46% |

1994 | 48% | 46% | 46% | 36% | 52% |

## June Daily Performance (2013 edition)

May 31, 2013Similar to last month, first 15 trading days and the last 5.

Looking at when the high and low occur for the month for SPY, there is a tendency for the high to be at the start of the month, and lows toward the end, although looking at the average line in the graph, it is W shaped (double bottom?). SPY:

IWM:

## May Daily Performance (2013 version)

April 25, 2013Similar to last month, first 15 trading days and the last 5.

SPY – a bit bipolar – high / low of the month at the ends of the month. Overall, it does favor a high at the end of the month, although the last 3 years it was in the first 5 days of the month. Trading days 3 – 5 seem to have a particular downward tilt.

## April Daily Performance (2013 version)

April 14, 2013SPY

Slightly different graph this month, this is trading days 1-15, and the last 5 trading days (noted as 16-20) This picks up the option expiration week.

Other statistics about SPY over the years for April

- low of the month occurs in the first 5 trading days 40%, last 5 20%.
- 35% of the Aprils never have a daily close below March’s close.
- the high of the month occurs in the last 5 trading days 60% of the time, 35% in the first five days, and practically never in the middle of the month.
- it seems that after a flat start, around mid month the trend is up.

## Parabolic SAR

March 10, 2013I was interested in trying to calculate Wilder’s Parabolic SAR. The description in Stock Charts Chart School is a bit abbreviated, and I don’t have a copy of his book. I have created a spreadsheet that seems to calculate correctly for S&P 500 index for the last year or so. The spreadsheet is: PSAR SPX . Columns A-G are from a download that includes the SAR value. Column H is whether the trend is up or down [cell H3 is =IF(I3<E3,1,-1) ]. Column I is the calculated SAR [cell I3 is =IF(H4=1,IF(MIN(D4:D5,L3+I4)>D3,MAX(OFFSET(C4,0,0,MATCH(-1,H4:H104,0),1)),MIN(D4:D5,L3+I4)),IF(MAX(C4:C5,L3+I4)<C3,MIN(OFFSET(D4,0,0,MATCH(1,H4:H104,0),1)),MAX(C4:C5,L3+I4))) ], Column J is the extended point value [ J3 is =IF(H4=1,MAX(C4,J4),MIN(D4,J4)) ], Column K is the acceleration factor [K3 is =MIN(IF(H4+H5=0,$K$2,IF(OR(AND(H4=1,J3=C4,J3<>J4),AND(H4 = -1, J3=D4,J3<>J4)),$K$2,0)+K4),$L$2) ], with K2 being the starting factor and L2 being the max factor value. Column L is the adjustment to the previous SAR to get the current SAR [ L3 is =(J3-I4)*K3 ]. Column M is just the difference between the downloaded value and calculated value.

Beware, I don’t have Wilder’s book to verify the formula, so the calculations might not be correct in all cases – use at your own risk!

## March daily performance (2013 version)

March 2, 2013SPY

Other statistics about SPY over the years for March –

- 10% of the years never have a daily close below February’s close.
- 40% of the time the low of the month is in the first 5 days, 30% in trading days 6-10, the low is in the last 5 days 15% of the time.
- the closing high is in the first 5 days 20% of the years, last 5 days 20%. trading days 11-15 had the high 45% of the time.
- It appears that first 10 trading days are flattish, positive option expiration week, with dip following.

## February Performance by Day

January 29, 2013Other statistics about SPY over the years for February –

- 15% of the years never have a daily close below January’s close.
- 50% of the time the low of the month is in the first 5 days, the low is in the last 5 days 30% of the time, only once in the last 6-10 trading days (so if we hit a low in those days, expect a lower low in the last 5 trading days).
- the closing high is in the first 5 days 30% of the years, last 5 days 35%.
- days 6-9 seem to have positive bias, and days 13-17 have more of a negative bias.