Value SuperScreen Vs. The S&P 500

|
Includes: AFL, CBAN, CMCSA, DAL, DFS, DQ, DXC, ED, EE, ESRX, FDMO, FFTY, GM, IX, KRA, MATX, MPC, MTUM, MU, PAA, PBF, RCL, SPY, TM, TOL, WLKP
by: Michael J. Bernard
Summary

One of the 23 original Value SuperScreen components is removed as a false positive.

Backtesting the Value SuperScreen against the S&P 500 showed encouraging results.

Both the VSS 22 and the Pretty 30 Portfolio show promising market-beating potential.

Introduction

My goal thus far in 2018 has been to isolate a proprietary stock picking strategy that has a back-tested and real-time proven record of S&P 500 Index (SPY) broad market-beating results.

My first such strategy has been named my "Pretty 30 Portfolio" stock selection method, which relies on identifying stocks with PEG Ratios of less than 2, positive dividend yields (in the absence of a purely growth oriented proposition), and an Analyst Recommendation Score [ASR] of 2.5 or less.

The Pretty 30 serves a few purposes for me personally: The first is to re-establish holdings primarily in stocks that I had to sell off in the first quarter of 2015 - this missing the last two plus years of the bull market I was on time for entering in its very youth (2009). The second is to continue to use what had been a proven strategy - albeit largely during a stock picker's paradise of 2010-2013 when it would have been hard to use even the most rudimentary of rules-based methodologies and not at least garner a market-equal return.

The knock on the strategy, overall, is the over-reliance on the ASR - a conflicted metric in a multitude of ways. Since it is my goal to be, as my bio suggests, an "Independent Buy-Side Analyst" it is not sensible for me to use "other side of the house" opinion - especially as many know that this opinion has its own internal and external conflicts.

My original sourcing method for holdings during this period was a combination of sources: The Investors Business Daily IBD 50, the S&P 500, and Dow Jones Industrial Average components, stocks that at various times met the stringent criteria method of the Zweig Screen, and other random things I might chance upon via Jim Cramer, Barron's, and other media.

Later I experimented with a trio of screening methodologies that I wrote about here that, in the aggregate of time, did not prove successful overall. One of the weaknesses was that the passers of these screens were often heavily centered in a few very stressed industries - primarily offshore drillers and tanker companies - so a lot more extrapolation on what went wrong there does not seem necessary.

In full retrospect, it is my opinion now that those screens were TOO aggressively "value" driven (to the point of picking old milk well past the sell by date in many instances) as to be reduced to a closet of falling knives -although I think there were a couple winners hidden among the junk. Again, one of the weaknesses in the three screens was a reliance on the ASR to be a warning sign it has (predictably) failed to be.

So, it has been a goal of mine to create a more "proprietary" and successful screen, abandoning any and all reliance on the ASR, and to have a more rules-based approach in comparison to the Pretty 30 methodology, which still requires a bit of human subjective judgement.

One of the structural/technical reasons for the Pretty 30 to require subjective judgement was that the broker I was using for this period of stock accumulation was Motif Investing - and a "Motif" is limited to 30 stocks at any one time. So, this seemed to be a good target to shoot for, despite the fact I had 38 stocks from my previous universe of holdings that met the portfolio criteria - so I had to subjectively choose the best 30. And my solution for the ASR had not quite been solved yet...

Over the course of the last 5 months, two main things happened to help in my evolution forward: The first was that I chanced upon Folio Investing, which allows fractional share purchases (like Motif) but also has a few other features and benefits that at this time put it ahead of Motif for my needs, not the least of which being that a "Folio" can include up to 100 holdings that can be purchased in one simultaneous transaction.

So now, I can repurchase the entirety of my old universe of holdings and basically rebuild my old portfolio to its previous composition quite easily and cost efficiently, including heavily weighting those 38 that have "the best" metric qualities - still primarily PEG ratio and dividend yield (where applicable) - and have room for a new selection strategy if/when I found the right combination of factors that gave me the "winning formula" I was looking for.

The second thing I came across was the tremendous screening tool available at Tiingo. While there are more screeners around than you can shake a stick at, Tiingo allows for the creation and deployment of completely self-created metrics. I was able to create one of the metrics I have been experimenting with in just a few minutes and soon I was able to create a new list of potential investments with ease.

So now, I had almost everything I needed.

Value SuperScreen Constituents

In last week's article "Value SuperScreen Unleashed", I introduced a new screen criteria set for my universe of holdings and potential holdings, including adding in completely new source groupings from my proprietary screen on Tiingo as well as the fund holdings of iShares Edge MSCI USA Momentum Factor ETF (MTUM) and Fidelity Momentum Factor ETF (FDMO).

Where I use the term "Universe", this is a reference to the fact that a holding was, prior to the original Value SuperScreen article, in my Universe of Coverage - meaning it either was a previous holding or has recently been identified as having metrics that make it passing the original Pretty 30 selection criteria and had not been discarded (yet). One of the explicit goals of the VSS was to help in narrowing the universe of coverage in a more stringent rules-based approach.

Twenty-three individual stocks passed all 4 hurdles of the screening process: PEG Ratio of less than 2, Price to Sales Ratio of less than 3, Price to Book Ratio of less than 3, and Price to Free Cash Flow of less than 15.

Or so it had seemed.

In accumulating the below information, it became clear that one of the 23 stocks had wrongfully made the final cut:

Somehow, most likely due to a faulty data set, Citigroup (C) made the final cut when it in fact does not have a P/FCF of less than 15. I regret not having caught this last week, as this fact caused all of the below work to be done twice.

With C voted off the island, the VSS is down to 22 components.

El Paso Electric Company (EE) {Tiingo}

Plains All American Pipeline, L.P. (PAA) {Tiingo}

Royal Caribbean Cruises Ltd. (RCL) {Tiingo}

Westlake Chemical Partners LP (WLKP) {Tiingo}

Colony Bankcorp, Inc. (CBAN) {Zweig}

Daqo New Energy Corp. (DQ) {IBD 50}

Kraton Corporation (KRA) {IBD 50}

Micron Technology, Inc. (MU) {IBD 50, MTUM/FDMO}

Aflac Incorporated (AFL) {Universe, MTUM/FDMO}

Comcast Corporation (CMCSA) {Universe}

Delta Air Lines, Inc. (DAL) {Universe}

Express Scripts Holding Company (ESRX) {Universe}

General Motors Company (GM) {Universe}

ORIX Corporation (IX) {Universe}

Matson, Inc. (MATX) {Universe}

Toyota Motor Corporation (TM) {Universe}

Toll Brothers, Inc. (TOL) {Universe}

Discover Financial Services (DFS) {MTUM/FDMO}

DXC Technology Company (DXC) {MTUM/FDMO}

Consolidated Edison, Inc. (ED) {MTUM/FDMO}

Marathon Petroleum Corporation (MPC) {MTUM/FDMO}

PBF Energy Inc. (PBF) {MTUM/FDMO}

Value SuperSystem Vs. The S&P 500

For this backtest, I screened the 505 components of the S&P 500 and came up with 32 that passed the screen (6.3%) with an overlap of 10 that were already a part of the VSS (2%).

I used Portfolio Visualizer to compile the following results:

In the above backtest, the 3 portfolios were broken down as follows:

Portfolio 1 [blue]: The S&P 500 components that passed the VSS but were not a part of the VSS 22, a total of 22.

Portfolio 2 [red]: The S&P 500 components that passed the VSS and were a part of the VSS 22, a total of 10.

Portfolio 3 [orange]: The VSS 22 that were not a part of the S&P 500, a total of 12.

In the above backtest, the 3 portfolios were constructed as follows:

Portfolio 1 [blue] : All S&P 500 components that passed the VSS screen, 32 total holdings.

Portfolio 2 [red] : The complete VSS 22, 22 total holdings.

Portfolio 3 [orange] : My Pretty 30 Portfolio, a total of 30 holdings none of which pass the VSS.

Analysis

It did come as a bit of a surprise how handily the VSS 22 beat not only the S&P 500 (SPY) but also the S&P 500 components that also pass the VSS - in both scenarios, the first isolated to remove the overlapping components and then in the combined scenario.

In the case of the VSS selection strategy, 12 of the 22 stocks were sourced from the IBD 50 (FFTY), Zweig Screen passing stocks, and holdings of MTUM and FDMO - In the end, giving the components a healthy bias toward growth, comprised within a value strategy.

Both the VSS 22 and Pretty 30 have +2 Sharpe Ratios, with the VSS having less market correlation. According to Investopedia:

Usually, any Sharpe ratio greater than 1 is considered acceptable to good by investors. A ratio higher than 2 is rated as very good, and a ratio of 3 or higher is considered excellent.

I only used a year's worth of backtesting because I see the VSS as very metric based, and in the event that any of the 4 metrics became non-passing, the component would be immediately removed. I anticipate rebalancing the portfolio at least quarterly moving forward and, by the end of Q3 2018, will have real-time data to assist in performance evaluation.

Conclusion

The VSS 22 passed the first of my backtesting performance hurdles - beating the S&P 500 (SPY) and a similar cohort of VSS passing SPY components that are not a part of a recognized growth/momentum portfolio strategy.

The next few backtests will be testing the VSS 22 versus different sub-sets of stocks not within the S&P 500, broken down into different universes. I have not come up with a final decision on how to break them down (market cap, average volume, price-per-share) yet and would be open to suggestions. The goal is to keep it manageable (anything resulting in more than 50 passing stocks gets quite cumbersome to work with).

Moving forward, I have developed a great deal of confidence in the VSS and am even more encouraged now that the overall selection process - focusing primarily on growth/momentum universes of stocks - has been able to eat the cake of the most widely followed market benchmark.

One challenge I have with fully committing to the VSS 22 is the reluctance to throw my lot in with a few companies or industries that I otherwise would avoid in almost all circumstances - Automotive Manufacturers such as GM or TM, airlines like DAL - or that I know very little to nothing about yet. Very certainly, the final step prior to deployment of the VSS 22 will be a final deep dive on every component to determine, despite what confidence and hope the back-testing "proves", whether each holding does not present a danger reminiscent to my three previous failed screening methodologies.

It is not lost on me that GM, TM, and DAL had become members of my universe of coverage, not because I fully expected that they would one day challenge to become a member of the Pretty 30, but that each would eventually be proven to be, despite initial metric passing, inadequate. To expect now that they could be "the best of the best" is still somewhat laughable to me. Only time will tell.

At no point in the last 5 months have I been closer than I am now to selling my SPY/DIA holdings and moving those directly into my Pretty 30 and VSS 22 portfolios. Presently, I am accumulating holdings in a handful of ETF's with the assumption that at some point, if a market-beating strategy was identified and proven out, those could be liquidated and redeployed with ease.

The plan is to not have to make that decision until 2019, but if the data continues to be as positive as it has been thus far, the choice might make itself more obviously apparent than ever anticipated.

Disclosure: I am/we are long SPY, FFTY, DIA. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

Editor's Note: This article covers one or more microcap stocks. Please be aware of the risks associated with these stocks.