fabijo's account talk

[fabijo]

I feel much better now. I'm gonna do my best to follow what my spreadsheet tells me to do. For tomorrow, it says to go 100% C, so I'm sticking with it.

 

[fabijo]

Not that anybody really cares, but I now have a "system." Let's see what it does for this coming year. I created 5 spreadsheets. Each one tells you which fund to go into 100%, compared to other funds. These are the five different spreadsheets:

Sh1 All Funds - out of the G,F,S,C,I it tells you which one to go to.

Sh2 G C S I - out of these four choices, tells you which one to go to.

Sh3 C S I - out of these three, which one

Sh4 G S I - out of these three, which one

Sh5 S I - out of these three, which one

I ran the test on all of these sheets, giving an MACD of 19, 39, 9 as the parameters from June 2, 2003 to December 1, 2006. The worst performer was the one that allowed me to go into the F Fund. Here are the returns that each sheet would have given during those times:

Sh1 - 70.65% (choice between all funds)

Sh2 - 132.22% (choice between G S C I)

Sh3 - 128.18% (choice between S C I)

Sh4 - 133.37% (choice between G S I)

Sh5 - 143.14% (choice between S I)

Here are the returns for each fund during that time:

G - 16.7%
F - 12.64%
C - 53.84%
S - 86.07%
I - 111.8%
20% each - 56.21%

Add: I will be using the G S C I sheet. I believe the C Fund is going to start outperforming the small caps this coming year.

 

[fabijo]

For anyone who wants to play with it, here is a link to the spreadsheet (15.3 MB):

http://mircats.com/fabio/TrackerWithPrices2.xls

 

[ayla]

Thanks very much for posting all your data. I'm also working on a definite strategy so was very interested in what you have produced.

I was trying to get the total return for just 2006, i.e. January thru end of November and added the % numbers in column U of your first spreadsheet for each month, January thru November of 2006. I came up with a total of 13.83%.

Am I using the right columns and entries to come up with this number? Is this what you have for just the year 2006 so far?

 

[ChemEng]

Another spreadsheet fan! We definately need more people like us in the world. Ill give yours a look over, but it seems very counterintuitive that returns would decrease as the number of options increases. For instance, I would think that since SI is a subset of SCI, the return for SCI should be at least what the return for SI is. Let me think about it and Ill get back with you...

 

[fabijo]

Another spreadsheet fan! We definately need more people like us in the world. Ill give yours a look over, but it seems very counterintuitive that returns would decrease as the number of options increases. For instance, I would think that since SI is a subset of SCI, the return for SCI should be at least what the return for SI is. Let me think about it and Ill get back with you...

I was wondering the same thing. The way the spreadsheet works is that it calculates the percentage change of the MACD from one day to the next. Then it puts 100% into the fund that had the highest percentage change the previous day. Maybe the times that the MACD change more rapidly for the C Fund than the rest, being 100% C kept us from the gains that the I fund may gotten by following the U.S. markets.

Also, maybe the F Fund moves too slow to be using the same EMA as the other funds. I might have to update the spreadsheet to give the F Fund a different set of EMAs.

 

[fabijo]

I was trying to get the total return for just 2006, i.e. January thru end of November and added the % numbers in column U of your first spreadsheet for each month, January thru November of 2006. I came up with a total of 13.83%.

Am I using the right columns and entries to come up with this number? Is this what you have for just the year 2006 so far?

Those are the right numbers, but you can't just add them up, but you could if you were just trying to generalize (you'll be a couple percent off). Also, there are five sheets. The one you are looking at might be the All funds sheet, which has the lowest return.

 

[VirginiaBob]

or maybe just have the F fund as an alternative to the G, when your other indicators show that you should be out of stocks completely. Maybe something as simple as if the F fund is down, it's time to buy instead of G. Also, consider incorporating the G-fund penny into this decision making process.

 

[ChemEng]

Another option would be to tie in each fund's variance within a given time period in the decision rule? IE. change the decision rule from greatest MACD%del to MACD%del*var or somesuch. Ill keep twiddling with it.

 

[fabijo]

I'm too much of a geek, because this has kept me up all night. I was highly caffeinated all night and only went to bed at 4am and woke up late for work.

 

[fabijo]

I'm on a quick lunch break and I just added three more sheets. Here they are and here are the results:

G I - 119.86%
G S - 79.42%
F I - 49.59%

So, playing just the G and the I fund would have been the best option out of those three scenarios for the past 3 1/2 years. Hopefully, I'll be done checking the S&P and Wilshire for bear periods (like 1999 to 2003) by tonight. I'd like to see what this does during those down years.

 

[Birchtree]

Can I use those sheets as bingo cards - I need some better edges.

 

[fabijo]

There was a Birchtree, had a dog. Thrify was his name-o.

G-F-C-S-I
G-F-C-S-I
G-F-C-S-I

and Thrifty was his name-o

 

[ayla]

If you keep what you've got but add an add'l criteria of being in a C/S or I fund only when the VIX index is less than its 200 day average, I GUARANTEE you will be amazed.

 

[fabijo]

Thanks for the advice, ayla. I'll have to try it out.

For some reason, every time I read one of your posts, I think about following the white rabbit and taking a red pill.

 

[ebbnflow]

...I ran the test on all of these sheets, giving an MACD of 19, 39, 9 as the parameters from June 2, 2003 to December 1, 2006. The worst performer was the one that allowed me to go into the F Fund...

Somehow I find that very disturbing (about the F-fund).

Congratulations on finding your own system. Hope you stick with it. It looks very promising.

 

[fabijo]

I knew you would find the F Fund results disturbing! I'm gonna modify the algorithms more, because I think the bond market moves slower than stocks.

 

[Birchtree]

The bond market moves slower than your lawn grows.

 

[fabijo]

Okay, I just ran some tests on data from December 1, 1998 to December 31, 2003. I only used the S&P 500. For the G Fund, I just kept it the same price from December 1, 1998 to June 2, 2003. I found something interesting. If I used EMA 19 and EMA 39 during that bear timeframe, I come up with a whopping loss of -25.18%. But if I adjusted the EMAs to react faster, I get better results. Using MACD parameters, 5, 10, 7, the Monkey gets a +27.65% return as he bounces from C to G.

This tells me that I need to come up with some formulas that check for rates of price changes in negative and positive directions. I could use the formula to check data from previous months. Maybe a 50 day and 100 day moving average could be a trigger that tells the spreadsheet to use faster EMAs for its decision making.

 

[fabijo]

I just added today's prices to the Monkey Tracker. The darts landed on 100% I fund, so that's where I'll be cob tomorrow.