What's the Risk?

[fordest]

Did some tweaking for a little better 7σ with 16198 - Lowest return: 25.99

Oh nice work! I like this one.

[Octjan2]

This may be the best one yet. However, there is a lot of time in F.

[ksmoly04]

This may be the best one yet. However, there is a lot of time in F.

Approximately 3 months total; I think that's something I could personally tolerate, and if the F-fund gets too dicey, there's always the G-Fund to swap into. For the limits that we're seeking, this is a rockstar strategy.

[onerepmax]

there are several great strategies with less than 2 months (15% time) in F

http://tspcalc.com/seasonal.php?ID=1619 ... =5&Fmax=15

HINT: Don't use 2017 in your sort yet. I think the incomplete yearly data skews it enough to miss out on some nice numbers.

I'll be transitioning to #16013 soon, either at the first of the year or next IFT (only 1 day difference)

[fordest]

In case you missed it in the other thread, I found a sigma of 10.1!

16250

We need updated charts 12X12...

And I don't get the math but I love reading the explanations. What does that mean for likelihood of a negative return?

[JoeMurphy]

there are several great strategies with less than 2 months (15% time) in F

http://tspcalc.com/seasonal.php?ID=1619 ... =5&Fmax=15

HINT: Don't use 2017 in your sort yet. I think the incomplete yearly data skews it enough to miss out on some nice numbers.

I'll be transitioning to #16013 soon, either at the first of the year or next IFT (only 1 day difference)

Good observation.

[shadogree]

In case you missed it in the other thread, I found a sigma of 10.1!

16250

We need updated charts 12X12...

And I don't get the math but I love reading the explanations. What does that mean for likelihood of a negative return?

I can't decide, because you keep finding the ones I like. Stop it. Good find.

[12squared]

In case you missed it in the other thread, I found a sigma of 10.1!

16250

We need updated charts 12X12...

And I don't get the math but I love reading the explanations. What does that mean for likelihood of a negative return?

The Charts that were made for σ0 also work with σx. The only difference is that they describe the chance of X% return instead of a 0%.

[RubberNeck]

Huh???

[evilanne]

Here is the referenced chart

The term "sigma" is used in the world of Quality Control as a metric of reliability. It is akin to the use of the decibel (dB) for sound and radio signal levels, in which ratios span great ranges, e.g. from micro to giga. Reliability is usually expressed as a 1 in a______ chance of defects. In the context of annual returns, I defined the defect limit as the 0% return, because anything less results in "loss of capital".


The relative reliability of sigmas was calculated by dividing their respective Cumulative Normal Distributions obtained from Excel's NORMDIST function.

[robika]

In case you missed it in the other thread, I found a sigma of 10.1!

16250

We need updated charts 12X12...

And I don't get the math but I love reading the explanations. What does that mean for likelihood of a negative return?

Hi everyone! Been on here for a little while but have never posted.

I believe I just came up with one that has a sigma of 12.08:
17823

This is a great tool, thanks mjedlin66 for all your work!

[dougellen1]

is it possible for someone to extrapolate the tspcalc daily data strategies back to 1989? Would give us a more MACRO look at all of this. I know this is huge.....just asking. I would do it but can't figure it out.

[Sad Al]

Whoever came up with strategy #24823, you are my hero. Mean 30.12, standard deviation 1.78%. Ratio 16.92.

[fordest]

Whoever came up with strategy #24823, you are my hero. Mean 30.12, standard deviation 1.78%. Ratio 16.92.

Wow. Mine too...... and the lowest year: 26.93% Thanks for pointing this one out.

[Sad Al]

So using your example, if the Mean/StdDev is 2.0, it is 99% likely that no more than 1.98 years out of 20 (20*0.099) will be negative, but at least (>0%) 0.456 years out of 20 (20*.0228) will be negative. Am I looking at this the correct way?

Yes you are. In whole numbers, 1 year out of 40, the same as the the mean-2SD value.

I've been reading and re-reading these posts and calculations, including the confidence intervals. What I don't see is how you arrived at 20 years or 40 years, instead of using 100 or 1000 years, assuming we are looking at percentages expressed in years.

If the sigma is 2.0, it is 99% likely that no more than 9.9% of the years will be negative, which is 9.9 years of 100, or closer to 1 in 10, but a negative year could be expected at least 2.28% of the time, or about 23 times in 1000 years. Where does 20 and 40 come from?