No. of Recommendations: 5
The best historical timing was generally clustered around 9 - 12 month trailing averages when Meb Faber wrote his paper.I ran 10, 12, and 15 month.
10 because that's about 200 days, which is common. Also, GTT timing works best at 43 weeks.
12 because that's simple and explainable. YOY is a very common measurement interval.
15 because that's what musselmant likes.
Anything in that general range works okay. None of the periods or indexes are a hill to die on.
If the two important risk metrics, MaxDD and Sortino, are about the same, less trading and fewer whiplashes are better. The thing here is to have good overall returns without scaring you out of the strategy.
You and musselmant are arguing for longer period averages (64 week, 325 day, etc.).Not me. Him.
10 month -- 43 wk SMA for 1 wks -4% below IXIC
9/1/2006 12/1/2025 1/2/1986 9/1/2006
CAGR 13.6% CAGR 19.4%
Stdev 21.3% Stdev 28.1%
MaxDD(12) -30% MaxDD(12) -41%
Sortino 1.09 Sortino 1.37
Grth $10K $116,507 Grth $10K $388,405
# Sells 12 # Sells 15
Mx mo out 12 Mx mo out 15
whiplash 6 whiplash 8
12 month -- 52 wk SMA for 1 wks -4% below IXIC
9/1/2006 12/1/2025 1/2/1986 9/1/2006
CAGR 16.4% CAGR 20.6%
Stdev 21.1% Stdev 28.4%
MaxDD(12) -23% MaxDD(12) -40%
Sortino 1.34 Sortino 1.44
Grth $10K $185,662 Grth $10K $480,034
# Sells 8 # Sells 12
Mx mo out 17 Mx mo out 18
whiplash 2 whiplash 5
15 month -- 65 wk SMA for 1 wks -4% below IXIC
9/1/2006 12/1/2025 1/2/1986 9/1/2006
CAGR 15.4% CAGR 21.7%
Stdev 21.3% Stdev 29.1%
MaxDD(12) -24% MaxDD(12) -40%
Sortino 1.25 Sortino 1.50
Grth $10K $156,652 Grth $10K $578,190
# Sells 9 # Sells 9
Mx mo out 13 Mx mo out 31
whiplash 5 whiplash 4
