Friday, August 18, 2017

Friday, August 11, 2017

Tuesday, August 8, 2017

Sunday, August 6, 2017

8/6/2017 End of July Metrics

Sector Model
XLU
0.66%
Full Model
Date
Return
Days
BT
8/11/2015
-39.30%
726
TMK
11/23/2015
32.65%
622
NVR
12/16/2015
61.30%
599
CMP
2/19/2016
3.43%
534
NVR
2/22/2016
68.10%
531
AMWD
3/17/2016
41.77%
507
CASY
5/12/2016
-3.37%
451
AEM
6/7/2016
-10.10%
425
ESRX
6/13/2016
-17.63%
419
AMED
6/16/2016
-3.61%
416
FRO
6/27/2016
-21.93%
405
ASTE
7/12/2016
-17.09%
390
MFC
9/1/2016
53.76%
339
SFM
9/8/2016
20.99%
332
CFFN
9/12/2016
4.67%
328
FIG
12/6/2016
60.88%
243
FOSL
5/11/2017
-19.39%
87
HIBB
7/25/2017
14.50%
12
FOSL
7/27/2017
5.43%
10
HZO
8/1/2017
4.71%
5
(Since 5/31/2011)
S&P
Annualized
10.35%
Sector Model
Annualized
16.63%
Full Model
Annualized
13.65%
S&P
Total
83.92%
Sector Model
Total
158.92%
Full Model
Total
120.59%
Sector Model
Advantage
6.27%
Full Model
Advantage
3.29%
Previous
2017
S&P
66.43%
10.51%
Sector Model
120.54%
17.40%
Full Model
91.27%
15.33%


In the current metrics, the Full Model has still failed to find a maximum holding period, even after over 6 years of live testing:




Note on the “Taxed” line that the maximum annualized return rate continues to be the maximum recorded holding period.  Although the factor of taxes does indeed favor a longer holding period because of compound interest, I suspect that another factor is the effect of High Frequency Trading Algorithms eating away the potential profitability of short term trades.  Just as humans cannot compete with these black boxes, neither can my model.  But that’s the point of my model: I’m not trying to compete with the HFTs, but rather to get out of their way.

That said, I still ended up with three trades in the past few weeks, even though a 6 year holding period over 20 positions should have taken most of a year to trade that often.  There were two contributing factors: 1) I need to tighten up my gap threshold, and 2) some of my trades were forced by my companies being bought out for a nice profit.  Holding through the acquisition wasn’t even an option, and I needed to trade.

The Sector Model, on the other hand, has held the XLU position for an unusually long period of time, and the performance graph has become disturbingly quiet:



Note that such close hugging of a trend-line doesn’t always end well for the broad market, and in fact the sector ratios continue to favor defensive industries:





One can never time these things, but it remains a puzzle.  On the one hand, the recent slashing of regulations by the Trump administration typically helps the market for a few years, but on the other hand the failure of any sign of bipartisan healthcare reform bodes ill.

The dollar has been falling lately, which might explain why retail stocks are falling into my model.  But those are hardly bearish moves.

Will be interesting to watch.  As always, I’m happy I don’t have to worry about timing.

Tim


Wednesday, August 2, 2017

8/2/2017 Unscheduled Trade

Because PMC is being taken private, the model sold PMC and bought HZO with a 15% favorable gap.