Every decision is a kind of prediction... and every prediction, crucially, involves thinking about two distinct things: What you know and what you don't.
- B. Christian & T. Griffiths, Algorithms to Live By
How have stock fundamentals broken down?
Were you sold a wonderful stock story of great sales, earnings, and profit margins last year that foretold of large future price gains for 2018? You're not alone.
We've all put a lot of faith in fundamental stock picking over the years as a sure way to outperform the markets. As a finance Ph.D. and MBA investor, fraud examiner, investment adviser and anti-money laundering specialist, I want to operate with fundamentals. I have a huge bias toward fundamentals. I want to find the numbers I can trust to substantially outperform the markets in the year ahead. But the best of my research, applying multiple discriminant analysis to find the strongest market variables, keeps showing me that "winning" has become less about the fundamentals. The evidence is clear, you absolutely have to look at things differently and apply new alternative methodologies.
What do I mean? Let me give you a quick analysis of several key fundamental variables for 2018 across all the major US stock exchanges and show you how poorly they have correlated with positive returns.
Sales analysis for 2018
Let's start by evaluating sales as a predictive variable for 2018. Did sales have an impact on whether a stock would have a positive year last year? Using the chi-square non-parametric test, let's take a high level look at sales across all the stocks in the major US markets.
The 2x2 chi-square table for sales in 2018 yields some surprising results. There was no significant statistical difference between positive / negative returns for 2018 stocks that had sales and those stocks that didn't have any sales.
So basically, if you were chasing a great sales story in some stock for 2018, this variable didn't give you any discernible advantage last year over stocks without sales. In fact 74.6% of stocks with positive sales numbers declined for the year. This would not have been a particularly good fundamental basis for stock selection. Let's look at another variable.
Net Profit Margins analysis for 2018
Did positive net profits have an impact on whether a stock would have a positive year last year? This has to matter, right?
Again, the 2x2 chi-square table for net profit margins did not show any statistically significant difference in yearly performance between stocks with positive or negative net profits. They all did poorly with no measurable distinction between positive or negative margins. You may have purchased a stock last year based on a strong profit margin narrative that was expected to greatly outperform the markets. Chances are that you experienced what this broad analysis confirms. There was no meaningful difference in performance between positive and negative net profits for stocks in 2018. In fact 72% of stocks with positive net profits declined for the year.
Earnings analysis for 2018
So let's move on to a review of earnings. Earnings have to matter, right? Applying the same statistical test, let's see if earnings had a measurable impact on whether a stock would have a positive year for 2018?
The 2x2 Chi-Square table for earnings gives us our first statistically significant result among the sampled key fundamental variables. This broad test between stocks with positive earnings and no/negative earnings shows that simply having some value of positive earnings contributed toward positive price performance in 2018. So even though more than 73% of stocks (2234/3044) with positive earnings had a down year for 2018, there's a meaningful difference achieved from stocks having some earnings that will give you better returns than stocks with no/negative earnings.
This broad example test of the earnings variable with 73% of "good" stocks having a negative year probably illustrates best just how incredibly difficult funds and fundamental trading has become for investors to find gains.
Let's dig a little deeper into earnings since it was our first statistically significant variable to consider for 2019.
EPS growth analysis for 2018
Does higher one-year earnings per share growth have an impact on whether a stock has a positive year? Based on what we saw above for positive earnings, we might expect that higher earnings means better one-year performance results.
For this generalized test of EPS growth, I split stocks into two groups. All the stocks with higher than the median level of EPS growth in one group and all the stocks with lower than median level of EPS growth were tested in the other group.
We would again expect earnings to be a meaningful variable since it would follow that higher earnings per share growth also would be a significant determinant of positive performance. What the chi-square analysis shows is that the level of EPS growth between the two groups did not produce a statistically significant difference in annual performance returns.
So for 2018 it really doesn't seem to matter if stocks were in the top half of EPS growth or the bottom half, they performed poorly within a statistically consistent expected level. Both groups had more than 74% of the stocks decline by year end.
Insider Ownership analysis for 2018
Insider ownership is another variable I'm routinely asked about, so let's analyze it here. Studies in the financial literature show a wide range of results on the usefulness of this variable and the timing necessary to benefit. One study found that if you are going to rely on insider trade activity, the CFO's behaviors are evidently the most profitable of the insiders. In any case, I want to run the quick analysis on this insider ownership variable to see what the chi-square analysis tell us.
According to the results of our Chi-Square test for 2018, you would have ended up with the same chance of poor results whether you went with high levels of insider ownership or the lower levels of insider ownership. Both groups delivered negative annual results for more than 74% of the stocks sorted on this variable.
Institutional Transaction Percentage analysis for 2018
The last of the chi-square tests I want to share with you moves away from the typical key fundamental stock criteria. The variable, institutional transaction percentage, has money flow at its core. Arguably this variable would correlate highly with technical indicators of Chaikin money flow (CMF) or the Money Flow Index (MFI) as institutional transactions change their positions in a stock.
Does institutional transaction percentage make a difference on year-end results? To run the test I separate the stocks with institutional transactions into two groups, the percentage increase group and the percentage decrease group.
According to the results of our chi-square test for 2018, there are very strong statistically significant results with a large chi-square test statistic. For example, 70.9% of the stocks with increasing institutional transaction percentages declined for the year, whereas a much larger 77.8% of stocks with decreasing institutional transaction percentages declined.
This discriminant variable could be applied in real time via continuous monitoring of technical indicators related to money flow that reflect the net flows of institutional transactions.
So the quick takeaway from this brief analysis of a few key fundamental variables is that fundamentally-based stock selections of sales, earnings, and profit margins will land you statistically no better off in the aggregate than if you had picked the stocks with no sales, declining earnings, and negative profit margins throughout 2018.
And even day-to-day there's correlations that make no sense. It's all messed up. My great hope is a we get out of this ridiculous monetary regime. And when we do things start to make sense again."
~ Stanley Druckenmiller
So what works? What are the winning criteria?
Ultimately, that's the critical question that multi-billion dollar hedge fund companies are struggling with today. Let's take a closer look at what's going on with hedge fund managers and their views on the market.
The best and brightest in the hedge fund world are struggling mightily to outperform and even differentiate themselves from traditional stock picking.
The people running your money have a lot to prove in 2019. Hedge-fund managers once again lagged the broader market and will have to show they are worth their high fees. - WSJ
Hedge Funds are failing to capture excess returns
Legendary hedge fund founder, Cliff Asness of AQR, with over $225 billion in assets under management, has expressed very strong concerns about his industry.
But all is not well, and winter may indeed have come to hedge funds. The reason to worry is the evidence, from both their realized excess (vs. their positive beta) returns and, importantly, their correlations to traditional active stock picking, that hedge funds no longer are what they once were." ~ May 31, 2018 - Cliff Asness, AQR Capital Management
Traditional stock picking is fading out
According to recent JPMorgan estimates, just 10% of trading is regular stock picking. Passive and quantitative investing accounts for about 60 percent, more than double the share of trading a decade ago. In fact, according to JPMorgan's Marko Kolanovic,
The majority of equity investors today don't buy or sell stocks based on stock specific fundamentals."
Top money managers are using alternatives with technical indicators, gut instincts and opportunities with algorithms.
I'm just going to trust my instincts and technical analysis to pick up the stuff... but I want to be clear that the major challenge for the algos for me is not some horrible market that I can actually see myself getting caught in that but I could also see myself perhaps taking advantage of it." ~ Stanley Druckenmiller
These quantitative algorithms don't necessarily care about stock fundamentals or some analyst's performance narrative about a company.
Guy De Blonay, a fund manager at Jupiter Asset Management expressed it this way,
Eighty percent of daily volume in the U.S. is done by machines, so what you get is a lack of focus on earnings, a lack of focus on outlooks and you just get short-term movements based on very specific data that is released every day and that creates noise."
Big data strategies are increasingly challenging traditional fundamental investing and will be a catalyst for changes in the years to come."
Salman Ahmed, chief investment strategist at Lombard Odier, says:
"The rise of algorithm-based trading means that there are in these algorithms some levels which trigger sell-off, i.e. sell orders.
"Yes, I can argue that we needed some kind of correction, given what has happened over the last few months. But the ferociousness of the intra-day sell-off is driven by these pre-set sell orders, which come programmed in these algorithms automatically."
Stanley Druckenmiller, former Chairman of Duquesne Capital Management, says:
These algos have taken all the rhythm out of the market and have become extremely confusing to me."
Assets have been decoupled from reality based on high predictable liquidity
Mohamed A. El-Erian, Chief economic adviser at Allianz SE, the parent company of Pimco, where he served as CEO and co-CIO, recently stated it this way:
"Asset prices that, for years, were decoupled from fundamentals amid ample and predictable liquidity, and the proliferation of passive investing, computer trading and exchange-traded funds, including some that promised liquidity in inherently less liquid asset classes, are potentially amplifying the risk of contagion both on the way up and on the way down."
Money managers across the spectrum, who are transparent enough to openly express their frustration, are recognizing the fundamental breakdown in trading processes and operations.
I don't know where this is all going. If it continues I'm not going to return to 30 percent a year any time soon. Not that I think I might not anyway. But one can always dream when the free money ends we'll go back to a normal macro trading environment."
So what trading alternatives can we rely on for 2019?
"I don't know whether you read Kasparov's book, but he thinks that the ultimate chess player is not the machine it's the machine with the man and his intuition using the machine that way.
~ Stanley Druckenmiller
I have a lot of respect for these successful fund managers with so many years of experience and for the way they approach problems and challenges in the market place. The evidence is clear that fundamental approaches and conventional trading schemes are floundering. I offer a few alternative ideas that may be critical to our future success in finding excess returns, and welcome your thoughts toward finding even more opportunities.
Here are four machine based and intuitive approaches I have been applying to find opportunities in this current market environment:
1. Constantly checking for new variables that better explain price variance.
I gave some simple examples above using the chi-square statistical test, but I have written at length about the more complex approaches I use to find powerful discriminant variables across different cycles of momentum conditions. Testing more than 40 different variables simultaneously in combination through statistical measures to explain price behavior is a very powerful tool that anyone can apply.
(Source: Multiple Discriminant Analysis of the Price Momentum Anomaly and Reversal Event Signals, Henning 2016)
2. Look for significant macro-economic events and anomalies in the market
As Mohamed A. El-Erian stated above, these algorithms are going to amplify market effects. They are going to create unusual patterns as they build on their machine learning models of historical wins and losses while quickly moving billions of dollars (80% of market volume) in the markets based on probabilities.
A couple of my articles related these non-random macro liquidity events and anomalies possibly caused by algo trading are listed below:
(Source: V&M Breakouts / FinViz)
These ideas trading the VIX have been a rewarding strategy for many years. The signals these days are taking on even more extreme volatility and profitability as Fed monetary policy in the form of quantitative tightening continues to increase.
3. Improve your signals by monitoring momentum conditions and using top models from different streams of peer-reviewed financial research.
There are several ways that I accomplish this monitoring function to see what degree of momentum is in the market or what model is generating the best returns from different fields of finance. Ideally we may try to identify when value stock portfolios are gathering momentum to outperform growth portfolios or see when forensic algorithms are shifting performance between negative and positive scores.
Because the "big data strategies are increasingly challenging traditional fundamental investing and will be a catalyst for changes in the years to come," we need ways to become fast followers of market behavior.
One of my first evolving models of momentum detection started out with momentum gauges from my research on positive and negative momentum acceleration. It has transitioned into the momentum gauge control chart you see below indicating where we are in the continuum of momentum. The gauge switched negative (red over green) since late September (Week 39) last year and has completely validated its usefulness ahead of the market corrections since October.
The other signals I monitor are in the form of new portfolios each month of the best peer-reviewed financial algorithms from different streams of financial research. In the forensic portfolios I leverage the three bankruptcy and earnings manipulation models of Beneish, Altman, and Ohlson to identify the top adverse and positive stocks in the market. Collectively these forensic models address 22 different fundamental ratios and algorithms for stock selection.
Examples of these alternating monthly selections can be seen in the October strategy article here:
(Source: V&M Breakouts)
In 2017 when volatility was at record lows (not a single day with over a 2% move in the S&P 500 up/down) the two forensic portfolios both generated one-year returns of more than 90%. As volatility has increased the stability of returns declined as well, but the strength of the positive forensic portfolios has continued to outperform negative forensic portfolios longer term.
(Source: V&M Breakouts)
I also apply a value model from Stanford professor Joseph Piotroski with an enhancement from Benjamin Graham that has been well documented to outperform all other value selection models in the financial literature. The latest portfolio formation for January 2019 is lined here with Micron Tech (MU) emerging as a type of anomaly:
(Source: V&M Breakouts)
So I use different strategies to outperform the market and each of these quantitative approaches has their own advantage in different market climates. If fundamental stock selections in the form of forensic analytics and value scores are prevailing then they will rise to the top quickly in my live tracking portfolio dashboard for potential adjustment to my strategy.
Otherwise I will rely on my breakout stock algorithm that has outperformed the S&P 500 by 28% in 2017 and by 40% last year. My momentum portfolios leverage the multiple discriminant analysis work of Altman and Taffler to identify strong momentum characteristics as explained here in my new picks for 2019:
Breakout Stock Portfolio Total return +76.02%
So far the model has returned nearly 10x the S&P 500 returns over the two year trading period from inception.
The strategy I'm ultimately pursuing is designed to create a machine based and intuitive learning model that acts as a fast follower of meaningful market changes. Like most traders I'm looking for the greatest benefit with the highest risk mitigation strategy. I'm not seeking to generate my own AI model, but simply to stand on the shoulders of those before me who have developed amazing algorithms and well-tested models I can use as indicators and spotlights on excellent profit opportunities.
(Source: V&M Breakouts)
With my quantitative investing models that screen the entire market throughout the day, I basically rely on an inductive investment approach to follow where the strongest momentum, value, and forensic signals lead me. I stay out of a lot of trouble by not pretending to be some brilliant prognosticator of a generally unpredictable world.
Historically, the technology and healthcare sectors provide many of the best growth stories and I don't see that changing, especially as the Baby Boomer generation enters retirement. Always keep your eye and a healthy portion of your portfolio on medical and technological companies working on so many amazing new breakthroughs. Many of the best stocks in these sectors appear in each of the different quantitative portfolios I manage.
Lastly, I use all this information and monitoring to move to individual stock selection.
4. Pick stocks where the timing, positive market conditions, and top models converge.
One way I do that is by leveraging a live trading board that pulls the best stocks from all the different portfolios I publish and monitor each week, month, and year. A snapshot of top stocks is shown below with repeated selections showing how a stock emerges from the validation of several different models.
(Source: V&M Breakouts)
Some key observations from the Top Stock list taken today show several important ways to benefit.
First, take for example Hi-Crush Partners (HCLP) that was selected yesterday for the January Piotroski-Graham enhanced portfolio and surged over 30% intraday. It has maintained very high value scores for months and now shows early breakout conditions that further validate its potential.
Second, look for what portfolio types dominate the Top Stock list on a given day when markets are highly positive or highly negative. Currently there's a very high representation of stocks from the Forensic portfolios that rely on 22 different fundamental algorithms for selection.
These include Rosehill Resources (ROSE), NovaBay Pharmaceuticals (NBY), Mitcham Industries (MIND), Workhorse Group (WKHS), Novan (NOVN), and Nymox Pharmaceutical Corporation (NYMX) to list those over 18% today.
Third, the particularly large gains across these top selections are relatively higher than I have seen for quite a while in current market conditions. This may indicate some broader based rally that pulling up stocks like Genocea Biosciences (GNCA) from the Russell 30000 Anomaly Portfolio along with old weekly breakout stock selections that have not been seen for months like Presbia PLC (LENS) and Restoration Robotics (HAIR).
These stock selections that I publish regularly are designed with the understanding that we may not be able to trade at levels of high frequency as the best machine based algorithms do. But new strategies and timing indicators may give us all the signals we need to be fast followers who can benefit greatly while avoiding harmful noise and hype in the marketplace.
As always, I trust this will be a profitable contribution to your investment objectives!
All the very best in your trades and have a Happy New Year!
JD Henning, PhD, MBA, CFE, CAMS
Altman, E. I. (1968). The Prediction of Corporate Bankruptcy: A Discriminant Analysis. The Journal of Finance, 23(1), 193–194. doi:10.1111/j.1540-6261.1968.tb03007.x
Amor-Tapia, B. & Tascón, M.T. (2016). Separating winners from losers: Composite indicators based on fundamentals in the European context *. Finance a Uver,66(1), 70-94.
Beneish, M. D. (1999). The Detection of Earnings Manipulation. Financial Analysts Journal, 55(5), 24–36. doi:10.2469/faj.v55.n5.2296
Beneish, M. D., Lee, C. M. C., and Nichols, D. C. (2013). Earnings Manipulation and Expected Returns. Financial Analysts Journal, 69.2, 57-82.
Graham, B. (1949). The Intelligent Investor: The Definitive Book on Value Investing
Ohlson, J. A. (1980). Financial Ratios and the Probabilistic Prediction of Bankruptcy. Journal of Accounting Research, 18(1), 109. doi:10.2307/2490395
Piotroski, J. D. (2000). Value investing: The use of historical financial statement information to separate winners from losers. Journal of Accounting Research, 38, 1-41.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. 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.