End of 2017, Enter 2018

.

2017 is in the books.  For those of you who were following my work with the Optimized Connor's strategies, here's the bottom line:

Yes, the strategies make money.  This being said, it takes a considerable amount of work to do so, and this work is on a daily basis.  Take a day off and you potentially miss a trade, lowering the overall gains.

My actual returns for the period tested (through September 2017) were slightly more than 6% net gain (inclusive of fees, transaction costs, missed trades, etc.).  Had I used a bigger account (greater than $100K), I may have been able to achieve 7% for the account.  These are not annualized numbers -- these are raw gains for the period January 2017 through September 2017.

I note that there was never a major event that took the markets south, and in a mean-reversion strategy (such as all of Connor's strategies), if you are in a trade and expecting recovery when the market really turns you can get hammered (recall that stop losses significantly reduce gross gains in every tested Connors portfolio). 

As a comparison, my wife's TSP account (roughly equivalent to the ETFs EFA, SPY, VXF, AGG) is running about 18% over the past 12 months.  Management there is no more than 2x per month, and it's simply a matter of looking at the past 1-month performance of the EFA, SPY, VXF, and AGG and making a simple adjustment.  It takes me about 10 minutes tops, per month.  Of course this is an exceptional year and her account normally is in the 6% annual range, so I have no expectation that these high numbers will continue.  Nevertheless, management of my wife's account has been brainless, at least compared to the Connor's strategies.

All of this being said, I wholeheartedly support others who believe in the Connor's strategies, but suffice to say, I believe that there are better ways to trade (let alone invest) that can better utilize my time and effort.

~~~~~~~~~~~~~

Optimized Greenfield Sharp Ratio Portfolio

Since September I've been working on the development and testing of a lower-activity/interaction portfolio.  I simply do not have the time to run Connor's scans, enter trades, watch them day over day, and rinse/repeat.  I travel often, and the different time zones kill my ability to manage my accounts on a daily basis.

Yes, exposure (risk) goes up when you are fully invested.  I acknowledge this.  Volatility is a manifestation of visible risk so optimizing on risk (volatility) is one way not have a high ulcer index.

Here are the basic tenants of this Optimized Greenfield Sharp Ratio Portfolio, and all must be true at the time of purchase:

1. positive year-over-year (YoY) increase in revenues
2. positive quarter-over-1-year-ago-quarter (QoQ) increase in revenues
3. positive trailing-twelve-months (TTM) growth in revenues
4. positive revenues for the current quarter
5. positive year-over-year (YoY) increase in EPS
6. positive quarter-over-1-year-ago-quarter (QoQ) increase in EPS
7. positive trailing-twelve-months (TTM) growth in EPS
8. positive EPS for the current quarter

For those of you who have been following me for years, you should recognize this criteria as the basis of the Greenfield screen.  This criteria will (generally) keep you out of hot water, as it rejects companies who have falling revenues and are buying back their shares to increase EPS.  Rising revenues AND EPS, when brought together, cut the universe of quality companies down to about 300-400 in any one screen.

But wait!  There's more.  A few technical requirements must exist:

1. The close is above the 50d MA (price)
2. The 50d MA is above the 150d MA
3. The 150d MA is above the 200d MA
4. The 200d MA has a positive slope upward for at least the past 10 consecutive trading days
5. The average volume of the stock must be at least 150,000 shares on a 10-day average and 3-month average basis

Again, the above is taken as a refinement of my Greenfield screen.  These technicals, in combination with the revenue and EPS criteria, further reduce the scan list to 100-150 stocks at any given time.  This is quite manageable.

~~~~~~~~~~

I have software that allows you to calculate the optimum portfolio with respect to Sharpe Ratio (SR).   The software takes a list of stocks, an expected market performance (I use 12% annualized), looks at the risk-free-rate (RFR) of the market, which is about 1.26% annualized as I write this (use the 1-month treasury), and then it looks at the exponentially weighted moving average (EWMA) of closing prices over the last year or so for each stock.  Since the EWMA weights the more recent prices and price differentials (volatility) higher, recent activity at the time of calculation (and potential purchase) plays a significant role.  When you take the past performance and divide it by the standard deviation of the volatility you end up with the SR of the stock.

The software looks at all the different combinations of stocks that you are considering and develops a "space" that defines what is called the "efficient frontier".  Here's an example of an efficient frontier for a set of stocks that I am current considering:

The plot shows volatility on the x-axis and historical (~100 day) return on the y-axis.  You can see that SQM has both a high return as well as a high volatility, so it's located way in the upper right corner.  HON has the lowest volatility of the lot, and is at the midpoint of the scale in return, so is located on the y-axis right around the 12.11% mark.

The blue line defines the efficient frontier -- there is a combination of stocks and weights to those stocks that will place you on that blue line.

The red line is defined by the current RFR, and where it intersects the blue line is the historical return and volatility for the optimized portfolio.

Determining the exact portfolio with the software is done with a push of a button -- the math is quite complicated and while I understand it for two stocks, understanding it for 10, 20, 100+ is beyond me.  That's why I bought the software  :)

If you are not familiar with Sharp Ratio then I urge you to click here.

If you are presently invested (I am), the software tells me what the existing numbers are for my existing portfolio and the new portfolio numbers.  The output looks like this:

I didn't include the column headers -- "R" means return and "V" means volatility.

My present ownership of stocks is running about a historical annualized return of 16.87% and a standard deviation of volatility of 14.59%.  Divide the two and you get the SR = 1.07.

The software is suggesting a slightly modified change to the portfolio that takes me to the numbers shown in the upper part of the table.  The shift from an old SR of 1.07 to a new value of 1.338 is significant and I've already placed my orders for Tuesday.

Note that I've constrained the software to not recommend any SELLING, only adding more stocks to improve the SR.  Selling is a different topic and I'll cover that in another blog entry.  It's easy, but it may not be wife-TSP-account easy (see my opening above).

~~~~~~~~~~~~

Unfortunately, I didn't start developing this Optimized Greenfield Sharpe Ratio Portfolio approach until September 2017, so data is early.  This being said, I'm going to run a real portfolio with it, with actual money management, so we'll see how it does.

Another caveat:  I'm only going to invest into dividend stocks, and those that are higher than the mean yield reported by the scan.  For example, the current mean yield of all the dividend stocks meeting my Greenfield criteria is 1.67%.  If you are curious as to what stocks and ETFs are returned, here's the list (I will NOT be posting this often but you can see the quality of the stocks returned by the scan from this list):

BBL
BHP
CLX
CM
DEO
DNKN
DRI
EMR
EV
EVR
FHN
HASI
HD
HI
HLI
HLS
HON
ITW
JPM
KAR
LM
MANT
MC
MGA
MNR
MS
MSFT
PAYX
PBF
PETS
PFG
PKG
ROK
SCHN
SQM
STT
TD
TEO
TROW
TRST
TXN
UNM
VALE
WBS
WM
WRK
WYN
XLK

So, if I am going to build a NEW portfolio, to start on January 2nd, the software gives me the following as the optimum portfolio and weights:

XLK	36.66%
JPM	8.29%
HON	6.21%
WRK	4.35%
PAYX	4.28%
CLX	3.88%
HD	3.78%
WYN	2.89%
HLS	2.88%
MGA	2.61%
KAR	2.33%
EMR	2.33%
BHP	2.31%
WBS	2.30%
HASI	2.25%
MSFT	2.18%
CM	2.03%
MNR	1.49%
DRI	1.25%
DNKN	1.16%
LM	1.13%
PFG	0.90%
HLI	0.85%
MS	0.78%
ROK	0.55%
STT	0.36%

If I buy each one of these positions with the weights shown, I should be in the ballpark of the following historical performance (past performance is no guarantee of future returns):

"Current Portfolio" is ##### because I don't own anything in this ideal portfolio.  Note that the NEW portfolio structure does not perform as well as my present holdings -- this is simply because I have ownership of stocks that are not in the current solution set as of the run of December 29th.  

The key takeaway here is that 14% gains at a volatility of 7.69% has been achieved with this portfolio, for a massively improved SR compared to what I'm holding right now, so this is a good portfolio.

Of course, I could feed just bonds or bond funds to the software and it would give me great SRs but poor returns -- far under 12% target -- so the input to the system matters.

~~~~~~~~~~~~

Stay tuned.  We should learn something in 2018 about this portfolio.  It will also provide dividend income too, which is important for many.

~~~~~~

As with all my ramblings, you are responsible for your own investment decisions and I am not.  Please do your own diligence, and please take ownership for your actions.

Regards,

pgd