11/03/2012 new trade, and backtest

Small Portfolio	XLF & IAU	16.16%
Position	Date	Return	Days
DECK	6/15/2012	-36.88%	141
RIMM	7/16/2012	20.14%	110
OKE	9/25/2012	-3.77%	39
SEAC	9/25/2012	8.20%	39
CAJ	9/25/2012	-7.88%	39
DDAIF	9/25/2012	-8.67%	39
SSD	9/25/2012	1.14%	39
NSC	10/8/2012	-7.90%	26
WMK	10/22/2012	0.59%	12
CFI	10/31/2012	0.00%	3
S&P	Annualized	3.59%
Small Portfolio	Annualized	11.31%
Large Portfolio	Annualized	15.80%

 

Scheduled rotation: selling SSD; buying CGX.

The new purchase, CGX, is the first based entirely on the new adaptive fundamental feature.  Quite honestly, when I look at this stock, it looks horrific.  Scares me about as much as RIMM did when I bought it.

Since my gut instinct is about as bad as it gets, that may be a good sign…

I feel like I’m George Castanza trying to succeed by intentionally doing the opposite of what I want to do.

In the same vein, SSD (in the building industry) is something that I would expect to benefit from Hurricane Sandy.  We do, after all, have a lot of rebuilding to do up here.

This is one of those trades that I have to throw up my hands and hold my breath to see what happens.

NEXT SUBJECT:

Just to clarify my earlier note – the “Annualized” return percentages are calculated from the time that the model went live with real money, 5/31/2012.  The total returns from that time are:

S&P	Return	5.13%
Small Portfolio	Return	16.16%
Large Portfolio	Return	22.58%

 

So, the 22.58% return from 5/31/2012 translates into an effective annual return rate of 15.80%.

This corresponds to the return rates that were back-tested for the Sector rotation model from 2/07/2000 through 10/22/2012.  The following graph shows the S&P 500 index compared to the sector model returns for a regular taxed account (in green) and an IRA account (in red):

Tim

 

11/02/2012 We survived!

Small Portfolio	XLF & IAU	17.80%
Position	Date	Return	Days
DECK	6/15/2012	-38.03%	139
RIMM	7/16/2012	20.00%	108
OKE	9/25/2012	-2.18%	37
SEAC	9/25/2012	9.79%	37
CAJ	9/25/2012	-5.38%	37
DDAIF	9/25/2012	-8.61%	37
SSD	9/25/2012	3.11%	37
NSC	10/8/2012	-6.93%	24
WMK	10/22/2012	0.02%	10
CFI	10/31/2012	1.38%	1
S&P	Annualized	4.30%
Small Portfolio	Annualized	12.50%
Large Portfolio	Annualized	16.52%

 

Enter the super-storm.

We were a direct hit and were down for the count until tonight.  Before the storm I had entered a conditional market order to buy CFI if the AM sell price was hit.  As a result, the trade occurred without any need for me to be online.

Thankfully, our house was spared, and after the mandatory evacuation we came back to a place that was without power or cell service, but was not permanently harmed from the wall of water that decimated the south shore and buried fancy homes just a few miles from us.  Even buildings two blocks from us were flooded, but we didn’t get a drop.

I hear another nor’easter is set to hit on election day.

I’ll be happy when it’s over.  After two years of political wrangling I feel like I’ve been watching back to back marathons of the voice, the x-factor, American Idol, and so you think you can dance. 

In any case, no trades tonight, but I did want to comment on the model itself.  I mentioned recently that I had created a self-adaptive fundamental selection filter.  There was a reason for that.  The dirty little secret inside the Small Portfolio is the fact that IAU (gold) is just a secular toss-off to keep all those eggs from being in the same basket.  Gold isn’t really a trade, as such.  It’s, if anything, a kind of hedge to balance the model in case something screwy happens inside the guts of the sector ETF.  While unleveraged ETFs are pretty reliable, I don’t like to put all my assets in ANY single vehicle.  Accordingly, I plan to update the Small Portfolio to rotate through another group of ETFs that the model can analyze, but which are asymmetric to the sector ETFs (currently in XLF).  That will LIKELY be a regional rotation (Europe, Asia, the U.S., Emerging Markets, etc.), but I have some more research to do.

In any case, if you look at the guts of the Small Portfolio from 5/31/2011 to today, IAU has gained 11.13% and the nine sector ETFs have rotated into a 24.47% total gain (the average of those two is the 17.80% listed on the Small Portfolio line at the top of the page).

The Large Portfolio, on the other hand, has a total gain of only 23.52% since 5/31/2011.

While those are both far better than the 6.12% total gain of the S&P 500, it’s not exactly how the model was designed to work.  The Large Portfolio is supposed to work BETTER than the simple sector ETFs.  The fact that it has not outperformed is because its gains have been in SPITE of the fundamental selection filter, rather than partially BECAUSE of it.

I therefore now have two self-adapting features of the model.  I have a self-adapting holding period feature and a self-adapting fundamental feature.  Having started with Greenblatt’s simple formula (which didn’t work), I tried a Graham filter – much more complex – which helped, but not enough.  Greenblatt’s approach is a GARP formula (i.e. Growth At Reasonable Price).  It’s kind of a hybrid Growth-Value process.  Graham is a pure value investment process.

But the big question is… what the heck works best with my model?  What am I supposed to do, trial and error?

Well, no.  Took me a while, but I worked out a way to track what fundamentals actually work WITH the holding periods and technical environments selected by the model.

The first selections will not be perfect, and it will take a while for the model to fine tune.  As the fundamentals adapt, that may shift the holding periods.  As the holding periods adapt, that may shift the fundamentals.

And as the High Frequency Trading algorithms out there mutate, so will my own home made Slow Frequency Trading algorithm.

Kind of sick to think of quarter trades as “Slow Frequency”, when that phrase should properly apply to two year holding periods.

But times change, and so will we.

There will be a new trade on Sunday.  Until then, I hope everyone is safe.  Have a splendid weekend.

Tim

 

10/28/2012 rotation: selling AM; buying CFI

Small Portfolio	XLF & IAU	16.99%
Position	Date	Return	Days
DECK	6/15/2012	-38.57%	134
RIMM	7/16/2012	4.41%	103
OKE	9/25/2012	-1.40%	32
SEAC	9/25/2012	10.04%	32
CAJ	9/25/2012	-6.16%	32
DDAIF	9/25/2012	-8.40%	32
SSD	9/25/2012	-0.30%	32
AM	9/25/2012	16.51%	32
NSC	10/8/2012	-6.17%	19
WMK	10/22/2012	0.17%	5
S&P	Annualized	3.52%
Small Portfolio	Annualized	12.05%
Large Portfolio	Annualized	15.13%

 

Scheduled rotation: selling AM; buying CFI.

AM is in the Publishing industry, and CFI is in the Furniture industry.  Although AM has been a profitable trade, and may become more profitable still, it’s also a great opportunity for profit taking and repositioning into a technically stronger industry.

The move into Furniture is related to the SSD position in the Building industry.  More indication that the efforts of Bernanke to back mortgage related securities are gaining traction.

The CFI position also represents the first stock selection using the new adaptive fundamental model, which will gradually fine tune the fundamentals that are most synergistic with the technical environment targeted by the Mousetrap.

The first adaptation is to add an examination of long term Growth Persistence.  A stock with seemingly good fundamentals has less of a chance of being a value trap if it has persistent earnings growth.  That’s not fool-proof, of course, since of all the stocks on the NYSE, DECK is considered the best value by the model.  That MAY be true NOW, but it was certainly something of a value trap when I entered the position five months ago.

In any case, the adaptive model will slowly update the fundamental selection process until it reaches a (hopefully more profitable) equilibrium.  Although the model is outperforming the S&P by an 11-12% annualized rate, I think it can do even better with these steady adjustments.

Time will tell.

Tim

 

10/21/2012 selling AF; buying WMK

Small Portfolio	XLF & IAU	18.53%
Position	Date	Return	Days
DECK	6/15/2012	-22.73%	128
RIMM	7/16/2012	7.03%	97
OKE	9/25/2012	-0.40%	26
SEAC	9/25/2012	2.33%	26
CAJ	9/25/2012	-5.15%	26
DDAIF	9/25/2012	-2.31%	26
SSD	9/25/2012	0.00%	26
AF	9/25/2012	-4.11%	26
AM	9/25/2012	15.42%	26
NSC	10/8/2012	-1.88%	13
S&P	Annualized	4.69%
Small Portfolio	Annualized	13.30%
Large Portfolio	Annualized	17.12%

 

Scheduled rotation: selling AF; buying WMK.

AF is in the Thrift industry, which is losing breadth and money-flow relative to other industries.  WMK is in the Grocery industry, which is showing technical strength.

This is a net negative for the broad market, since Grocery is a defensive industry.

In any case, WMK is the last selection to be made on a pure Benjamin Graham style fundamental filter.  The model now has enough data to begin a self-adjusting fundamental selection process.  Fundamentals will progressively fine tune to best perform in the specific technical environment selected by the model.  Rotation periods and fundamentals will each self-adjust to the other until they reach long term equilibrium.

Don’t really have much to say about the broad market.  It’s under obvious pressure, but the most pressure is on defensive sectors like utilities.  So far it’s a healthy pause.  We’ll see how long that continues.

As always, a negative gap between AF and WMK will prevent the trade.

Tim