Successful Back-Tested PP Tweak

[moda0306]

I took the liberty of combining a few "tweaks" we've suggested... 1) No cash, 2) Relative Strength, and 3) Investing in PP after weak PP years (<10% return the year before)... I calculated the returns from the two best-performing non-cash assets the year after a <10% PP year.  This is what I came up with:

1976: 21.75%
1978: 17.75%
1982: 30.6%
1984: 10%
1985: 31.6%
1988:  1.15%
1989: 23.4%
1991:  3.5%
1993: 13.7%
1995: 14.95%
1997: 22.45%
1998: 18.2%
2000: -7.95%
2001:  3.35%
2002: 20.55%
2003: 11.15%
2005: 6.3%
2006: 12.35%
2009:  1.5%
2010: 21.5%



This averages 13.89%.  Not too shabby... but not amazing.  When it gets interesting is if you are in your second year of following <10% PP returns... Look at the follwing years:  1985, 1989, 1998, 2001, 2002, 2003, 2006, and 2010.

All those years were following the second year of sub-10% PP performance years, and had an average return of 17.76%.

We're a ways out from weak PP performance to trampoline off of, but take from it what you will.

[Lone Wolf]

I took the liberty of combining a few "tweaks" we've suggested... 1) No cash, 2) Relative Strength, and 3) Investing in PP after weak PP years (<10% return the year before)... I calculated the returns from the two best-performing non-cash assets the year after a <10% PP year.  This is what I came up with:

Just to make sure I've got the right idea, does this mean you take the two best-performing of gold, stocks, and bonds from the previous year and go all-in on them 50-50 the next year?

I was interested in how volatile this would be but it's a little hard to tell with so many years missing from the series.  How'd you pick which ones to include?  Just how horrific was 1981 in a no-cash, relative strength PP? 

[moda0306]

LW,

That's exactly correct.

I ONLY chose years with weaker than average (<10% returns) performance.  I chose those years only because I wanted to highlight the affect of doing tweaking after some weak years, hoping for a bounce up.

As a compliment to this, in years following high-performing years, maybe bringing cash more heavily back into the picture would be the appropriate response.

I was simply trying to show a  move that would only be able to be taken after low-performing years, hoping for a reversion to a mean.

[moda0306]

LW,

Your point made me think... maybe some kind of 40/30/20/10 PP, with the volatile assets being ordered to the degree they had a positive year the year before, and the cash allocation always being at 10% unless the previous year was a >10% PP year, then you'd put cash at 20%.

If you had 2 years of >10% PP years in a row, cash could go to 20%, with the other assets ordered as we've already described.

So for example, after 2011's performance, since the PP has had 2 positive years in a row, the allocation would be 40% bonds, 30% gold, 20% cash, and 10% stocks.

I'd probably insist on always tweaking it away from gold and towards stocks... for instance, stocks could never be the 10% asset, and gold could never be the 40% asset... since there are volatility and long-term growth considerations involved... but that'd probably be optional.

I'm going to work on this... I'll get back to you!

[moda0306]

LW,

Ok here we go... 40/30/20/10 PP... cash always at 10%, unless the year before the 4x25 PP had a >10% year... then cash goes to 20%... 30% after 2 10%+ years... 40% after 3 10%+ years.

The other 3 volatile assets fit in at whatever order their prior-year performance order would indicate.

I got a CAGR of 12%, but oddly the Standard Deviation was quite large at 11.52... though I think this is misleading.

The worst and only negative years were 1994 & 2000, with a -3.12% and -1.01% loss, respectively.  2008 delivered a 9.55% gain. 1981 delivered .08% gain.

It appears that these strategies can be combined for decent affect.  I'm willing to bet if cash had been muted a bit more we'd be seeing pretty impressive results, even with the increased volatility.

[AdamA]

I calculated the returns from the two best-performing non-cash assets the year after a <10% PP year.  This is what I came up with:

How did it do buying just the one non-cash asset with the highest RS?

[MediumTex]

It's a shame that we can't "forward-test" various PP tweaks.

[AdamA]

It's a shame that we can't "forward-test" various PP tweaks.

Even if we could, I think some of us would still find it difficult to resist the urge tinker.

[clacy]

It's a shame that we can't "forward-test" various PP tweaks.

Many would say the same about the PP

[MediumTex]

It's a shame that we can't "forward-test" various PP tweaks.

Many would say the same about the PP

Or about anything in life.

How cool would it be to be able to "forward-test" a marriage or a business.

"Honey, I know we really seem to love each other right now, but my forward-testing software indicates that we are going to be at each other's throats in 5.7 years."

[cabronjames]

3) Investing in PP after weak PP years (<10% return the year before)... I calculated the returns from the two best-performing non-cash assets the year after a <10% PP year.

I'm guessing that it would be better to use a X (6?%) real PP year (using CPI inflation index) than a 10% nominal PP year.  iirc in the CPI itself was over 10% at some time (early 80s?) in the 1973-2011 PP history, whereas on the other extreme it was ~0% in 2008.

Out of curiosity, what's your take on using real return vs nominal return?