Thread: Last Month's Best Fund Method Strategy

Originally Posted by malyla

Just tested this method from Jun 03 to present

I think this method was most impressive during 2001 and 2002.

In 2001, this method was up 5.1% while all 3 stock funds were down (C = -11.94 , S = -9.04 , I = -21.94 ).

In 2002, this method was down a little ( -1.2%) while all 3 stock fund were down a lot (C = -22.05, S = -18.14, I = -15.98 ).

Originally Posted by malyla

Just tested this method from Jun 03 to present (share prices inception). As a buy and holder with 85%C and 15%S starting in Jun 03 I would be about 54% gain on my initial investment on Jun 2003. As of today, I'm really only 44% gain on my initial investment of Jun 2003 due to some bad I fund/G fund timing with transfers. (I track both my buy&hold and trading strategies). The last month best fund method at first did not look like it was going to work as the 2003-2005 showed that I was missing massive gains by looking to the last month best fund, but after finishing the whole spreadsheet (using Tom's) with some adjustment to federal paydays as the start or end of a month to initiate a transfer, the following conclusion jumps out. I would be up 128% on my Jun 2003 investment if I had followed this 'last month best fund' method. And my current balance would be almost double what I have now.

Why has no one (TSP, employer, etc...) ever mentioned this before? The last 4 months would have been made mute as this method had us out of stocks during the big drop from Dec 07 - present. Also missed the drops in March-April 2004, June-July 2006, July 2007. Missed some gains as well (Oct 2003, May 2004, Aug-Sept 2004, Feb 2005, May 2005, Nov 2005, Mar 2006, Aug 2006, March 2007,and Aug 2007), but it's the drops that affect the rate of return the most.

Very interesting. I do appreciate all that this website has taught me and look forward to more illuminating ideas and methods.

Good luck all. Currently 100% G.

malyla

That is correct(from a quick check of 2000-2002). I used the method of picking the last month's best fund for the beginning of the new month. I used 100% transfers. I compared it to my buy and hold (85%C 15%S) and was amazed. B&H 54% since Jun 2003 compared to 'last month best fund' 128% since Jun 2003. I will continue to track my two methods plus this one for awhile before I make a choice.

Well I have two weeks to better understand the concept. If I understand it better I will give it a try starting in May. Of course I will keep 50% in "G" just to play it safe. I would like to retire in 4 years and can not afford to be to agressive. YET!

Originally Posted by RPM

I am sure if we look hard enough, we'll know why it works, and the reason will make it less "simple"

It's very simple why this works, it indicates how the different funds are trending.

There has to be a slightly better indicator than the end of a calendar month. I'm hoping someone here comes up with it. But without the odd-ball corrections that people add to the back-testing but never work in the future.

Originally Posted by RPM

Would appreciate getting the info if you do back test to the beginning of TSP.

I hope others will do their own independent numbers to verify my figuring. But I ran the numbers for 1988 thru 2000 when there was only the G, F, and C fund.

First the good news. During this 13 year period, this method would have returned 278% while the F-fund would have returned 172% and the G-fund would have returned 148%.

The bad news is that during those 13 years, the C-fund returned 608%. The C-fund had 20-30-something-% returns for at least half those years.

You can't get greedy with this method.

Originally Posted by nasa1974

Well I have two weeks to better understand the concept. If I understand it better I will give it a try starting in May. Of course I will keep 50% in "G" just to play it safe. I would like to retire in 4 years and can not afford to be to aggressive. YET!

This strategy is good only as a multi-year approach. It will be a "dog" some months and even some years. (Just like Ebb's method could be a good multi-year strategy but it just been down for the last 6 months and it's going roar back and finish the year at 30% or maybe next year).

I think this method can be improved.

I think this method is kinda like what Ebb was doing, but he was trying to indicate when to jump in/out from date to date (and had complicated rules that only worked during back-testing).

1. "The trend is your friend" is a very, very old maxim to live by.

2. Theorys abound. Greatly.

3. Past performance is no guarentee of future results, and

4. I'll look at setting up an autotracker account with that theory in mind, if someone wants to become the "owner" of that theory, and post it up, and maintain it- so that we all can watch for a while.

Anybody want to take responsibility for this? (Requires a LONG term commitment).

Rule #1, (For my self only)
The Market goes up, the market goes down,
or to put it in other terms,
the market goes up, AND the market goes down!

Will this work with any 30 day period?

say from the 15th to the 15th of the month?

Originally Posted by James48843

4. I'll look at setting up an autotracker account with that theory in mind, if someone wants to become the "owner" of that theory, and post it up, and maintain it- so that we all can watch for a while.

Anybody want to take responsibility for this? (Requires a LONG term commitment).

I'll volunteer (I'm only 15 minutes from the Tennessee state-line.)

There will be a 2 day gap using this method. I will calculate the fund % COB-last-day-of-the-month. Autotracker enter the IFT (if any) and it takes effect COB-fist-day-of-the-month.

Can we start it in the autotracker back to the beginning of the year?

I performed a more rigorous analysis ( and eliminated a mistake from my last post) and came up with the following:

Assuming an IFT made COB on the first of each month:

Jun -Dec 2003 results = 20.29% (assumes initial 100% C for Jun, then 100% I in July, 100% S in Aug-Sep, 100% F in Oct, 100% S in Nov, 100% I in Dec)
Jun - Dec 2003 fund results =(G fund = 2%; F = -1.3%; C = 11.6%; S = 23.7%; I = 22.3%)

Jan - Dec 2004 results = 13.3%
Jan - Dec 2004 fund results =(G fund = 4%; F = 3.72%; C = 9.13%; S = 14.4%; I = 15.47%)

Jan - Dec 2005 results = 7.39%
Jan - Dec 2005 fund results =(G fund = 3.93%; F = 1.34%; C = 7.03%; S = 13.72%; I = 10.9%)

Jan - Dec 2006 results = 19.2%
Jan - Dec 2006 fund results =(G fund = 4.57%; F = 5.05%; C = 13.07%; S = 14.72%; I = 18.93%)

Jan - Dec 2007 results = 17.64%
Jan - Dec 2007 fund results =(G fund = 4.44%; F = 6.71%; C = 5.11%; S = 4.69%; I = 11.76%)

Jan - Mar 2008 results = 0.87%
Jan - Mar 2008 fund results =(G fund = 0.9%; F = 0.92%; C = -4.96%; S = -5.27%; I = -5.77%)

Totals for the time period from Jun 2003- Mar 31, 2008:
last month best fund method = 78.69%
fund results =(G fund = 23.9%; F = 21.2%; C = 55.1%; S = 85%; I = 131.8%)

Conclusion: Method seems to beat the C,F,G funds, however, a buy and hold of S and I would have produced higher earnings. This is mainly due to this method failing in the years 2003-2005 to adequately pick the reoccuring best fund. However, the years from 2006-2008 have the last month best fund method performing better that all the funds (B&H strategy). So can one assume that this method does not work well in a strong bull market but during a weak bull / bear market, this method can lead to significant capital preservation.

Good Luck in building your third leg of your retirement stool. It's only 30-40% of your retirement ;-)

Originally Posted by malyla

Good Luck in building your third leg of your retirement stool. It's only 30-40% of your retirement ;-)

Thanks for your analysis, this has been quite interesting to see it unfold.
By the way, although only 30-40% of our retirement, it's the only part in
FERS that we have any control over. At least, until they strip us of that
ability too. Again, this is great stuff and I'm sure they'll be looking very
close at the tracker for everyones benefit.