Attn: Sigma Hunters

[bamafamily]

I have been away for a little while..I decided to go Sigma hunting
I was looking for the lowest sigma as my ulcer % needs to stay very low....
IMHO I don't think the "F" fund is all that it once was so my other requirement was to replace all the "F" fund trades with "G" fund trades....(so my "F"=0)

My new guideline strategy (17842):
CAGR: 28.33
Std Dev: 2.45
Sigma: 11.56
CSI - 55.2%

[alphaalpha]

So what is the highest sigma strategy that does not use F? Or no F and no I? ( mainly no F)

[bamafamily]

So what is the highest sigma strategy that does not use F? Or no F and no I? ( mainly no F)

The one in my signature line is not too shabby.
It does have “I” in it but it has no F.

[TopComm]

So what is the highest sigma strategy that does not use F? Or no F and no I? ( mainly no F)

The best I've found with no F and no I is 19656
29.01 Average (2004 - 2016)
4.64σ

[onerepmax]

http://tspcalc.com/seasonal.php?ID=1998 ... tby=stddev

19983

Mean 27.59
Std Dev: 1.90 (sigma =14.52)
CAGR: 27.57
G F C S I Total Compound Return: 2,924.84
43 2 16 34 4 Time In Stocks (C,S,I): 54.80

[Jones44]

Questioning the hunt for super high sigma. Would you not rather have a strategy with a mean of 32 and a std dev of 5 or 6? It would have a much higher upside and the down side would be just as high, if not higher than the means of all the super high sigma strats. It seems like a win/win. Same floor but way more potential. As I'm typing this, I have a feeling that I'm not thinking about it right. Is my thinking wrong?

[onerepmax]

Personal choice I guess. Lower SD, higher sigma is a smoother ride I guess. I can't decide. I just like sorting and looking at the choices.

[onerepmax]

Current best SD strats for the mean...
35/12+
34/10+
33/~8
32/5.25
31/~4
30/2.85
29/2.52
28/2.19
27/1.9

These are with no fund restrictions or filters. The 27/1.9 is very low F and I BTW.
Oddly, there are really no results like this for means of 26 and below.

Is ~25% 10 out of 10 years better than say

~25% 8 of 10 years with one year at ~20% and one year ~30%

[bamafamily]

Agreed....you have to draw the line somewhere....
Is 33/7 better than 30/4 or than 27/2...
I guess it all depends on your level of comfort with possible swings in the market....
Me personally, I think a lower std Deviation shows more robustness across different types of market situations....60% 1 year followed by a 15% year tells me that the strategy was super tuned for the first year but did not fare near as well in the second year.....

given the short timeframe I have until retirement, I want the lowest std Dev which has the better chance of returning the ~~same average amount over the next 4-6 years....(I don't need a 2% year from some wild strategy that Swings a lot and then have to wait 3 years for the 60% return....I just don't have that amount of time to deal with that...)
My 2 cents....

Personal choice I guess. Lower SD, higher sigma is a smoother ride I guess. I can't decide. I just like sorting and looking at the choices.

[onerepmax]

Same, but then I say I have high tolerance for risk...

[mindofmush]

Questioning the hunt for super high sigma. Would you not rather have a strategy with a mean of 32 and a std dev of 5 or 6? It would have a much higher upside and the down side would be just as high, if not higher than the means of all the super high sigma strats. It seems like a win/win. Same floor but way more potential. As I'm typing this, I have a feeling that I'm not thinking about it right. Is my thinking wrong?

You are right to question this but you need to determine which strategy fits your risk tolerance.

High sigma strategies give up maximum return for safety (and will under perform a straight C fund buy&hold strategy during the "good" years).

The strategies with the highest returns generally greatly out perform the stock market (C fund B&H) during the good years as well as the bad years but can have returns as low as 14.91% for #16870.

2007 C fund = +5.55% #16870 = +14.91%
2008 C fund = -36.99% #16870 = +93.86%
2009 C fund = +26.67% #16870 = +110.0%
2015 C fund = +1.46% #16870 = +15.97%

The high sigma strategies will beat the returns of the high mean strategies on some years but will never give the possibility of a 110% return.

Even middle of the road strategies like #7980 have 78% years with very few years under 25%.

So, pick what's best for you:
buy & hold C fund (11% average over last 25 years--better than 80% of mutual funds)
extreme safety (with a very, very good return approx 25%)
maximum return (with better than average lows and astronomical highs)
middle ground (best of all?)

All of them are better than the 5-7% that the financial advisers say you'll get.

[Rothwell]

I included all complete years in my search (2017 excluded).
I calculated the theoretical dollar amounts using the TSP’s “How Much Will My Savings Grow?” calculator.

StdDev no greater than 2% = 20094
Mean: 28.04
StdDev: 1.97
Sigma: 14.23σ
TSPHMWMSG: $5,756,502.25

StdDev no greater than 3% = 17619
Mean: 30.34
StdDev: 2.98
Sigma: 10.18σ
TSPHMWMSG: $7,797,972.89

StdDev no greater than 4% = 19627
Mean: 31.69
StdDev: 3.98
Sigma: 7.96σ
TSPHMWMSG: $9,312,532.71

StdDev no greater than 5% = 17835
Mean: 32.23
StdDev: 4.82
Sigma: 6.68σ
TSPHMWMSG: $9,996,120.39

StdDev no greater than 6% = 19046
Mean: 32.71
StdDev: 5.34
Sigma: 6.12σ
TSPHMWMSG: $10,644,888.30

StdDev no greater than 7% drops below “Six Sigma” so is excluded.

[Rothwell]

Further analysis:

Going from 20094 (2%) to 17619 (3%) there is a 35.46% increase in theoretical savings with a 39.78% decrease in Sigma.

Going from 17619 (3%) to 19627 (4%) there is a 19.42% increase in theoretical savings with a 27.88% decrease in Sigma.

Going from 19627 (4%) to 17835 (5%) there is a 7.34% increase in theoretical savings with a 19.16% decrease in Sigma.

Going from 17835 (5%) to 19046 (6%) there is a 6.49% increase in theoretical savings with a 9.15% decrease in Sigma.

Going from NGT* 2% to 3% appears to have a roughly equal trade off between increase in theoretical savings to loss of Sigma (11.48% difference between savings to Sigma).

Going from NGT 3% to 4% has a still respectable increase in theoretical savings but a larger loss of Sigma (35.77 difference between savings to Sigma).

Going from NGT 4% to 5% has a small increase in theoretical savings especially when compared to the big drop in Sigma (89.20% difference between savings to Sigma).

Going from NGT 5% to 6% has a small increase in theoretical savings and roughly equal loss of Sigma (34.01% difference between savings to Sigma).

*No greater than

[bamafamily]

Further analysis:
Going from 20094 (2%) to 17619 (3%) there is a 35.46% increase in theoretical savings with a 39.78% decrease in Sigma.......

Nice calculations Rothwell...I assume you mean increase in Sigma?? (as in it gets worse???) tomato, tomaato...I know)

[Rothwell]

Yes, the Sigma gets worse as the StdDev goes up.

My thoughts are these:

If playing it safe is your primary concern then the leader in not greater than 2% (20094) has an amazingly high mean to low StdDev, but swinging up to 3% (17619) will perform even better over time with a still really good mean to StdDev.

However, my money is on 4% (19627). The significant increase in theoretical savings when compared to 3% or 2% is worth the risk. 19627's worse year was 2007 with 24.60% return and is an outlier.

The loss of Sigma to the low increase in theoretical savings doesn't make 5% (17835) or 6% (19046) that attractive, unless perhaps someone has a short window of time to retirement...