No. of Recommendations: 10
While cleaning out files on my hard drive I came across this old (2010) "timing" post by Jim. Can't say I ever followed it but with the gold surge maybe it is of value, if it is still accurate...
I gave it a test, from mid 2010.
It didn't add value. In short, the suggested 145 day time-out was way too long. CAGR in bullish and bearish periods were very similar, and backwards. No joy.
But the general idea has held up, sort of. Most of my models for gold created somewhat later suggested much shorter timeouts. I say "sort of" because none of them used as a long/cash signal beat buy and hold, but they did distinguish pretty strongly between high-return and low-return eras, much like the 99 day rule for the S&P 500. I imagine it's difficult to watch your gold position sit way below its peak for a decade, as it did.
Other time-outs for this model:
Classic 99 day rule worked better. Bullish 74% of the time, CAGR on bullish days 11.5%, CAGR on bearish days 3.9%
Examples:
No new high in last 25 days: Bullish 38% of the time, CAGR on bullish days 18.8%, CAGR on bearish days 4.1%
No new high in last 45 days: Bullish 50% of the time, CAGR on bullish days 19.0%, CAGR on bearish days 0.8%
No new high in last 75 days: Bullish 62% of the time, CAGR on bullish days 15.4%, CAGR on bearish days 0.3%
I didn't calculate it, but the 45 day version must have had a nice ulcer index...it held onto most of the gains from the peak around 2012 rather than leaving you way below your portfolio peak until the new bull market started in 2019. It had an average of 1.3 round trips per year.
Jim
