12/31/2016 Happy New Year!

Mouse	XLB	1.79%
Rabbit	Date	Return	Days
NVR	12/16/2015	0.62%	381
CASY	5/12/2016	4.25%	233
AVB	5/24/2016	0.58%	221
AEM	6/7/2016	-16.08%	207
AMED	6/16/2016	-15.38%	198
FRO	6/27/2016	-7.22%	187
ASTE	7/12/2016	15.07%	172
MFC	9/1/2016	31.61%	121
CFFN	9/12/2016	18.53%	110
FIG	12/6/2016	-4.33%	25
Turtle	Date	Return	Days
BT	8/11/2015	-33.14%	508
DY	10/30/2015	5.31%	428
TMK	11/23/2015	23.02%	404
UPLMQ	12/1/2015	80.80%	396
CMP	2/19/2016	16.91%	316
NVR	2/22/2016	4.86%	313
ENOC	3/15/2016	-18.03%	291
AMWD	3/17/2016	10.10%	289
ESRX	6/13/2016	-9.38%	201
SFM	9/8/2016	-3.86%	114
	Since 5/31/2011	Annualized
S&P	66.43%	9.54%
Mouse	120.54%	15.20%
Rabbit	68.23%	9.76%
Turtle	117.20%	14.89%
	Previous	YTD
S&P	51.94%	9.54%
Mouse	77.79%	32.10%
Rabbit	57.21%	7.01%
Turtle	58.35%	37.16%

Final numbers for 2016.

As I mentioned before, the Rabbit and Turtle will be rejoined as a single model using the Turtle’s rotational period.  The recombined model will have the average of each.

I’ll likely also go back to the simple naming convention of “Sector Model” and “Full Model”.

Going forward, the five and a half year holding period on the full model will tone things down a bit, but working less for greater returns is the dream of any investment blog.

Let’s hope for a wonderful 2017, and celebrate the end of a very strange year…

Happy New Year!

Tim

12/26/2016 Get a Life!

Mouse	XLB	2.27%
Rabbit	Date	Return	Days
NVR	12/16/2015	0.68%	376
CASY	5/12/2016	5.32%	228
AVB	5/24/2016	-1.69%	216
AEM	6/7/2016	-22.35%	202
AMED	6/16/2016	-14.23%	193
FRO	6/27/2016	-6.96%	182
ASTE	7/12/2016	15.82%	167
MFC	9/1/2016	33.38%	116
CFFN	9/12/2016	18.46%	105
FIG	12/6/2016	-1.97%	20
Turtle	Date	Return	Days
BT	8/11/2015	-33.53%	503
DY	10/30/2015	8.67%	423
TMK	11/23/2015	23.87%	399
UPLMQ	12/1/2015	90.77%	391
CMP	2/19/2016	21.08%	311
NVR	2/22/2016	4.92%	308
ENOC	3/15/2016	-15.98%	286
AMWD	3/17/2016	11.34%	284
ESRX	6/13/2016	-9.10%	196
SFM	9/8/2016	-0.66%	109
	Since 5/31/2011	Annualized
S&P	68.29%	9.79%
Mouse	123.07%	15.48%
Rabbit	68.04%	9.76%
Turtle	122.71%	15.45%
	Previous	YTD
S&P	51.94%	10.76%
Mouse	77.79%	25.47%
Rabbit	57.21%	6.89%
Turtle	58.35%	40.64%

Closing in on the end of the year – and the end of the Rabbit.  The Turtle will absorb the Rabbit artifacts at the end of the year for the final version of the model, and the end of the “experimental” phase.

Since the launch of the models in May 2011 I’ve been trading my own funds, and will continue to do so.  But as of the close of the Rabbit I consider the experiments to be at an end.  The Turtle strategy will self-adjust, but will not require further human development.  All metrics will automatically fine tune based on which metrics perform best for long term trades.

Both the short term holding sector model (i.e. the Mouse) and the long term holding stock model (i.e. the Turtle) are right on the regression line from the 1999 to present backtest, showing that live trading and backtests are perfectly consistent.  I couldn’t ask for better.

That said, “success” for me is a way to spend less time trading and more time living.  The blog has been quieter of late as I closed in on the goal of this process.  Going forward, I plan to do two things.  I’ll continue to post trades and thoughts.  The only difference is that those trades will be much less frequent.

But that’s a good thing.  High frequency is for black boxes.  Low frequency is for investors who want to have a life.

And that’s my wish for you, dear reader, as we approach the new year.  Live that year.  We only have a limited number of them anyway.

Tim

12/11/2016 Taxes and Revenue

Just a quick note on dynamic scoring.

The bipartisan congressional budget office calculated the so-called Bush tax cuts as a way of optimizing tax revenue over long time periods.

Nothing magical there.  If the government taxes 100% of your income it should get 0 dollars in revenue because you won’t work – or at least that’s the theory proposed by Art Laffer.  The “Laffer Curve” is a hypothetical tax revenue curve between 0% taxation and 100% taxation, with both 0% and 100% collecting 0 dollars in revenue, and the optimal point being somewhere in between.

Both Democrats and Republicans accept the basic concept, but disagree on where that optimal point would be.  On the left, Paul Krugman places the top rates at a whopping 65% before “economic distortion” sets in.  The CBO placed the top rates at 35%.

But who is right?

The answer isn’t so simple.  The maximum tax revenue collected depends on how long a time frame we are measuring.

To give a real-world example, let’s compare the United States and the Soviet Union during the cold war.  The government of the USSR was able to keep pace with us for about 70 years before it finally collapsed.  Granted, the people were impoverished, but the government was able to squeeze a theoretical 100% taxation / with redistribution into a competitive arms race that left the future of freedom in doubt for generations.

How?  If the Laffer Curve truly collects 0 revenue at 100% taxation, wouldn’t people stop working?  Well, no.  People still work for the common welfare (or to keep from going to the Gulag), but they do so with less creativity and less resources.  Nevertheless, they do indeed work, and over the course of long time frames a communist regime can sustain about 40% of the resources that the United States can with our present system.

Over shorter time frames a Communist government can collect as much revenue as a Capitalist one.  How short?

“Short” in historical terms, but a full lifetime for a typical human.  The chart above has 146 lines, each measuring how much revenue would have been collected from United States citizens at each percentage of taxation from 1871 to the present.

Obviously for the top line – measuring the full 146 years – the maximum revenue point is at an average rate of 27%.  Keep in mind that 27% is the average for total GDP for the federal, state, and local levels.  That is, all taxes of all kinds combined should be about 27% to collect the maximum amount of sustainable revenue.  Since the federal government historically collects between 15% and 20% of GDP, that leaves a little less than half for the states and local municipalities.

But what’s not so easily visible on this graph are all the lines clustered at the 100% taxation rate.  For the first 69 years maximum revenue would have been collected at the 100% rate!

And that’s why the Soviet Union was able to keep pace with us in the cold war until it was about 70 years old.  Communism works, in the extreme short term, and only for the government.  It doesn’t really work for the citizens at any point.  But the government can impoverish its citizens and maintain full revenue for about 70 years before economic distortion breaks its back.

My point is this: Republicans are right about the ideal rates, but they are wrong about the immediate gains.  Lowering taxes is better for the economy, and over the course of centuries it is better for the government too.  But that’s CENTURIES.  Immediate benefits in a single Presidential term are fantasies that only work if the demographics kick in at the right time.  Kennedy and Reagan had favorable demographics.  We don’t.

I understand that the next President plans to lower taxes to the CBO calculated rates.  He is absolutely correct to do so.

But he also has to cut spending; because if you cut taxes, in the short term you also cut revenue.

In a 4 year time frame more taxes means more revenue and less taxes means less revenue.  The Democrats are right about short time frames.

So was the Soviet Union.

But should a government only think in terms of 4 years?

Tim

12/5/2016 Trade

The Full Model sold BSET and bought FIG with a 3.5% favorable gap.

12/1/2016 Whipsaw

The sector model traded after the open yesterday into XLU (after the down gap), and then traded back to XLB into the close.  I couldn't post this because the trade didn't register on my account until after the close.

In any case, although I avoided the down gap I'll record it on the model as if I took the hit.

11/30/2016 Sector Trade

The Sector Model flipped from XLB to XLU at the close.

If there is a favorable gap in the morning I'll make the trade, but if not I'll recalculate before today's close.

11/27/2016 Where we are, and where we're going

Mouse	XLB	3.56%
Rabbit	Date	Return	Days
NVR	12/16/2015	-3.48%	347
CASY	5/12/2016	7.47%	199
AVB	5/24/2016	-9.35%	187
AEM	6/7/2016	-18.51%	173
AMED	6/16/2016	-17.31%	164
FRO	6/27/2016	-2.57%	153
ASTE	7/12/2016	15.49%	138
BSET	8/3/2016	17.28%	116
MFC	9/1/2016	29.25%	87
CFFN	9/12/2016	15.15%	76
Turtle	Date	Return	Days
BT	8/11/2015	-34.94%	474
DY	10/30/2015	-0.76%	394
TMK	11/23/2015	18.58%	370
UPLMQ	12/1/2015	82.04%	362
CMP	2/19/2016	17.06%	282
NVR	2/22/2016	0.59%	279
ENOC	3/15/2016	-13.93%	257
AMWD	3/17/2016	13.39%	255
ESRX	6/13/2016	1.30%	167
SFM	9/8/2016	9.35%	80
	Since 5/31/2011	Annualized
S&P	64.54%	9.49%
Mouse	128.00%	16.18%
Rabbit	66.60%	9.73%
Turtle	120.87%	15.51%
	Previous	YTD
S&P	51.94%	8.29%
Mouse	77.79%	28.24%
Rabbit	57.21%	5.97%
Turtle	58.35%	39.48%

So where are we?

With about a month left to the year, the Turtle continues to outperform the Rabbit, which is set to be officially absorbed into the Turtle for 2017.  I’ll continue with the same 20 stocks but rotate and select on the time schedule and fundamentals the Turtle has been using:

Small cap stocks with room to grow, low current P/E driven more by rising earnings than falling prices, and good 3-5 year projected returns in an industry that looks bottomed out on its unique business cycle (no two industries are exactly alike in this respect).

The holding periods will average about five years as well – some longer and some shorter, with a typical trade about once a quarter for the best trade in the portfolio.

The Sector Model continues to function as expected:

The current deviation from the regression line is -0.11406489.  A standard deviation between 2 and -2 is entirely normal behavior.  This is barely more than a tenth below the regression line.  It can’t get more “normal” than that.

The post-election rally hasn’t yet shown up in the sector ratios, so we are still showing a bear market configuration:

And what of Trump?  He’s an unknown.  A real bull market COULD be sparked by a perfect combination of taxes and regulation, but it would require an INCREASE in both trade and immigration.  Trump is in favor of lowering taxes and regulations, but campaigned to renegotiate trade agreements and cut down on illegal immigration.  It’s not impossible for him to increase legal immigration and increase trade by no-tariff bi-lateral agreements.

But Trump is running into the same problem that Clinton (or anyone) would be facing in these next three years: the long term demographic trends pose a serious head-wind to the market, pointing to another major bear market taking the S&P back into the 1600 range by 2019 (going into the next election).

The current value of the S&P is 2213.35.  The projected value of the S&P should be 2228.43.  We are exactly on target.

Without some kind of massive inflationary stimulus, we shouldn’t pass our current levels again until 2023.

Demographics aren’t destiny, but they aren’t to be ignored either.

If I were President I’d ramp up both trade and targeted immigration, and cut taxes and regulations.  But I’m not President, and none of the candidates promised to do all four of those actions.

We may be in for a long rough ride.

Tim