NASDAQ Seasonality 2000

NASDAQ Seasonality

Data file: c:\inv\NASDAQ Statistical Analysis.xls

Data used: NASDAQ monthly closing data, October, 1984, - July, 2000. Source: yahoo.

Objective: determine any seasonal patterns in the NASDAQ.

The following table shows the mean percentage change in the NASDAQ broken down by month, October, 1984 - July, 2000:

NASDAQ stats, 10/84 - 7/00:
Month:	Average:	Rank:	Z Value:
All:	1.65%
1	4.64%	1	1.67
2	3.52%	3	1.05
3	0.83%	8	(0.46)
4	-0.14%	11	(1.00)
5	1.82%	6	0.10
6	2.19%	4	0.30
7	0.30%	9	(0.75)
8	-0.12%	10	(0.98)
9	0.87%	7	(0.44)
10	-0.84%	12	(1.39)
11	2.15%	5	0.28
12	4.23%	2	1.44
Nov-Mar	3.07%

- The average month during the study period experienced a 1.65% rise in the NASDAQ.

- The best months were as follows:

- January with a 4.64% average gain;

- December with a 4.23% average gain;

- February with a 3.52% average gain;

- June with a 2.19% average gain;

- November with a 2.15% average gain

- May with a 1.82% average gain;

- The worst 3 months are:

- October (-0.84%);

- April (-0.14%);

- August (-0.12%).

- Notably, September is not a bad month for the NASDAQ, ranking 7^th with a respectable 0.87% gain.

- The best timing strategy using these data would be to go long the NASDAQ from November to February, then from May to June:

Seasonality: investing 100% in Nov-Feb, May-June, else 100% cash 10/84-7/00.
	Model:	B&H:	Diff:
10k =>	246,889	155,404	91,485
rr:	22.6%	19.0%	3.6%
Average DD:	1,092	2,281	(1,190)
Worst DD:	28,472	53,456	(24,984)
Worst %DD:	11.5%	34.4%	-22.9%

Using the median rather than the mean gives a similar picture:

NASDAQ stats, 10/84 - 7/00:
Month:	Average:	Rank:	Z Value:	Median:	Rank:
All:	1.65%			2.14%
1	4.64%	1	167.12%	3.83%	1
2	3.52%	3	104.67%	3.10%	2
3	0.83%	8	-45.67%	1.91%	6
4	-0.14%	11	-99.87%	0.86%	10
5	1.82%	6	9.55%	3.04%	4
6	2.19%	4	30.04%	1.59%	8
7	0.30%	9	-75.09%	0.91%	9
8	-0.12%	10	-98.34%	3.10%	3
9	0.87%	7	-43.52%	0.77%	11
10	-0.84%	12	-138.74%	1.73%	7
11	2.15%	5	27.88%	0.27%	12
12	4.23%	2	144.43%	2.82%	5
Nov-Mar	3.07%

Note that the top 6 months by median return are:

January 3.83%

February 3.10%

August 3.10%*

May 3.04%

December 2.82%

March 1.91%

These are similar to the top 6 by mean but not identical; note that August had a negative mean return but ranked third by top median return.

Typical Year:

Typical Year, 100 = prior December close, based on average monthly change:
Month:	Mean:	Index:
12/31:		100.00
1	4.64%	104.64
2	3.52%	108.33
3	0.83%	109.23
4	-0.14%	109.07
5	1.82%	111.05
6	2.19%	113.48
7	0.30%	113.82
8	-0.12%	113.69
9	0.87%	114.67
10	-0.84%	113.71
11	2.15%	116.15
12	4.23%	121.07

Presidential Year: Monthly percentage change broken down by presidential year versus month.

		Pres Yr:
	All:	Post	Mid	Pre	Ele
all:	1.65%	1.73%	0.58%	3.12%	1.15%
1	4.64%	6.94%	0.24%	9.48%	1.91%
2	3.52%	-1.81%	4.45%	3.55%	7.90%
3	0.83%	-0.95%	1.00%	4.54%	-1.28%
4	-0.14%	1.17%	-0.19%	1.06%	-2.60%
5	1.82%	6.24%	2.27%	0.92%	-2.17%
6	2.19%	0.72%	1.15%	3.17%	3.70%
7	0.30%	4.15%	-3.13%	3.34%	-3.16%
8	-0.12%	1.81%	-5.96%	3.76%	-0.06%
9	0.87%	0.95%	-1.31%	0.11%	4.67%
10	-0.84%	-0.65%	1.23%	-4.22%	0.66%
11	2.15%	1.18%	3.78%	1.40%	2.24%
12	4.23%	1.07%	3.45%	10.38%	2.05%

- note that the preelection-election year pattern holds, but less dramatically than for the S&P500; the average preelection month did best (3.12%), followed by the average postelection month (1.65%), then election (1.15%).

- December is a phenomenal month across all years, averaging 4.23% average across all years;

- The best 6 month span appears to be postelection October through midterm March, inclusive.

Unfortunately, using this information does not add much value at the simplest level; if you invested in every Presidential year except midterm years, your results would have been as follows:

Presidential: investing every year except midterm:
	Model:	B&H:	Diff:
10k =>	163,119	155,404	7,715
rr:	19.4%	19.0%	0.4%
Average DD:	2,139	2,281	(142)
Worst DD:	56,110	53,456	2,654
Worst %DD:	34.4%	34.4%	0.0%