I See Patterns

Includes: AMT, DDM, MOS, MS, T
by: George Acs

"I see patterns," is much less of a hook than "I see dead people." But recognizing patterns is an invaluable skill for an investor, if only it was really possible.

PatternsI've often wondered about patterns and our ability to recognize them or even care about them.

Caring about patterns is important if you buy into the Santayana aphorism "The one who does not remember history is bound to live through it again."

We'll never know if Santayana's use of the word "bound" referred to the likelihood of occurrence or to a predestined obligation.

"Likelihood" offers the possibility of deviating from the pattern, while "pre-destined" offers no such possibility of reprieve. Why bother to care if it's pre-destined? Perhaps a respected theologian could weigh in on that question.

Imagine ancient man and his struggle to understand his environment and the forces that shaped it.

Understanding was all that could be hoped for, because man could never hope to control that environment or the natural laws that pertained to it.

Seasonal change, for example, is generally gradual, although there may certainly be outlier kind of days. So I don't believe that our ancient ancestors, especially before they learned to cultivate crops, would have thought much about that pattern.

Surely the pattern of day becoming night and then night becoming day again was one that occurred with enough frequency and regularity that they sooner or later put it all together and stopped their daily panic attacks and human sacrifice rituals in their attempt to appease the Gods.

Eclipses? That's a different issue. Irregular and far spaced, probably spaced well beyond the typical lifespan, despite the fact that it seems as if there's a new one everyday.

Haley's Comet? An 86 year interval? Why in the world would anyone care about that?

The one that always caught my interest was how early humans come to understand that there was a pattern of behavior involved in the birth of children?

Lots of false negatives in that one, and as far as we know, only one reported false positive. Again, an issue for theologians.

It's relatively easy to see patterns when the events are closely related in time, but nine months is a long time, especially when there are additional confounder events taking place during that time frame.

If the term "confounding events" is itself a confounder to you, I was referring to "sex."

It took me many years of having my account managed by a trusted and wonderful stock broker to start feeling a sense as if "I'd seen this happen before." That was well before the internet put data and charts at anyone's fingertips.

The "buy and hold" approach he practiced resulted in observing the passage of many apparent cycles and patterns of price movement. In addition to missing many opportunities to sell shares at one of those regularly recurring price peaks, it seemed as frequently the ultimate sale of shares occurred during the downswing, despite the fact that there was invariably an upswing to come.

It's probably the same feeling that the "Griswold Family" felt when trapped for hours in a traffic circle in London.

It then took me a very long time to methodically look at the data and convince myself that there might be a way to profit from what appeared to be patterns, albeit sometimes with the predictability of the adolescent mind.

What bothers me about a formal approach to stock patterns and their visual representation in charts is that we, the individual investor, are often shown the instances in which it appears that an outcome is predicted by a pattern. Yet what we don't know, or are not shown, is whether that fulfillment is actually a reflective of the pattern.

In the world of science and research, measurement tools must always undergo validity testing. False positives, false negatives, specificity and sensitivity are terms thrown about. Don't get me started on Kappa.

In other words, how often to the patterns turn out to not result in the expected outcome, despite the charting indicators all being in place?

The mathematical proof must exist someplace, but if I'm going to be limited to a single conspiracy theory, then this is the one.

The problem, though, is that unlike patterns based on natural phenomenon or even upon human behavior, stocks have become an altogether different class. Of course, overlay the imperfection of human analysis and there may be a problem, even when so called experts are the ones doing the observations and interpretations that may themselves just fulfill a pre-existing bias.

Take the recent case of American Tower (NYSE:AMT), the company that rents out its tower space to cell phone providers. For purposes of this piece, I'll conveniently ignore my superficial discussion of 5 year charts and Bollinger Bands that appeared in my book, Option to Profit.

When word came of AT&T's (NYSE:T) plan to purchase Deutsche Telekom's (DT) T-Mobile service provider, shares of AMT dropped by 7% as analysts were quick to point out that the newly combined cell phone powerhouse entity wouldn't be needing as many cell towers, as there would be existing duplicate installations that could be eliminated.

I'm not going to bore you with the details of how quickly the price recovered and how little time it took for a Morgan Stanley (NYSE:MS) analyst to put an "overweight" on AMT shares. For my purposes, I bought the depressed priced shares, sold the calls and was quickly assigned.

That's my pattern.

Nine months to the day later, AT&T announced that the transaction was being abandoned in the face of US Department of Justice anti-trust action. Of course, the analysts were quick to opine that shares of AMT would greatly benefit from the need to keep all of those tower spaces leased. That translates into a "strong buy."

You may know where this is going. You would think that when the initial hypothesis was proven wrong and shares of AMT ultimately outperformed the market, they might curb in a bit of the "smug factor" and maybe even do a public mea culpa upon it's subsequent "unexpected" underperformance after the deal fell through.

You would have guessed wrong and with little excuse, since an analyst's pattern is to not acknowledge missed calls.

In the case of AMT, there was a pattern, even though it was based on just a single point. Yet that couldn't be recognized by the best and brightest. They ignored the past, as well as their failed thesis in favor of their continued belief in the very same thesis that failed yet again.

At least that part is an easily recognizable pattern.

Maybe it was the "9 months" that threw them. After all, ancient man struggled due to that temporal delay before realizing the pattern related to child birth.

In their defense, however, even though I saw the pattern, I failed to follow up on that pattern with a repeat of my own and let a good, predictable opportunity go to waste.

There was certainly a time when fundamentals, things like earnings and price to earnings ratios were all the inputs that you needed to bake into a stock's price movement, as it was further massaged by the intangible investor behavior and mood, that was further influenced by herd mentality.

But these days, the"patterns" really are based upon nothing but casual causality, as they become a self fulfilling prophecy thanks to the elimination of the human factor.

The incorporation of notions regarding patterned stock behaviors into trading algorithms increasingly make these patterns become reality, especially as individual behavior and "group think" is removed from the equation.

If an algorithm is programmed to execute an action when a particular "predictive" event occurs, the result of that event's occurrence will have been pre-destined.

I don't need a theologian to tell me that it was a human hand that altered the path of the markets. This human hand is different from that of insider trading manipulation, yet it's legal.

Sure, the human factor still exists, but only as a vestigial triviality. Whereas human behavior in the markets would feed upon itself in the past, now it is fed by trading programs fueled by algorithms.

The human mind tends to fill in the gaps.

Ask any criminal lawyer and their experiences with eye witnesses. For good or bad, we often see patterns where they really don't exist, so it would seem to be helpful to take the human factor out of that equation and replace it with a dispassionate algorithm, that only carries the subjectivity of its creator.

Maybe that's why we as humans, and perhaps as individual investors, are "bound" to fail.

I look for patterns all of the time, knowing that even if discovered, they may not withstand the test of time and may be nothing more than casual associations, such as I recently mentioned between Goldman Sachs (NYSE:GS) and ProShares UltraDow ETF (NYSEARCA:DDM)

The axiom that an investor shouldn't be net long going into a weekend must have started with some kind of empirical observations recognizing a pattern, perhaps followed with actual statistical analyses. However, over the past year, there's been little reason to heed that advice as large gains on Friday were more likely to be followed by gains of any magnitude than by losses.

By the same token recognizing that over the same recent time period a large loss on a Friday was not likely to be followed by a similar loss on the following Monday can prevent panic selling at the lows on Fridays in order to avoid the event that's not likely to occur.

Instead, ignoring the pattern would lead to a pattern of selling low and buying high.

That's a bad pattern to follow.

A recent observation that the Barclays Volatility Index ETN(NYSEARCA:VXX) when sharply higher, often gave back those gains in the last hour of trading may also not stand the test of time, much less any kind of rigorous scrutiny. Yet, it offered the opportunity to sell calls in the afternoon and then buy them back in the afternoon in multiple occasions over a short period of time over the past month.

I believe that it's far easier to see a pattern, even a fleeting one, than it is to uncover a hidden gem of a stock and then know at precisely what points to buy and when to sell.

The human element, though, is still vexing.

Herd mentality can be protective, but it can also blind you to rational thinking.

Yet somehow, lemmings have not gone extinct.

A perfect example of the herd mentality was the universal agreement that Mosaic (NYSE:MOS) was going to explode upward in price following its most recent earnings release in March 2012.

At the time, that argument was backed up by the observation that there was unduly large activity in buying the next two higher strike level out of the money options contracts.

Almost always owning shares of Mosaic and almost always selling calls on those shares, I revel in seeing my shares assigned and collecting both capital gains and option premiums. Mosaic has been an excellent company to consistently apply that pattern of trading.

However, admittedly, I gloated when shares of Mosaic tumble by 5% after the earning's release. I gloated, despite the fact that my shares ended up going unassigned.

Yet unlike the lemmings that chased after Mosaic, I kept the shares and got to repeat the time tested pattern that simply follows Mosaic's pattern.

Mosaic simply goes up and down and up again. Coming up against the edge of the cliff isn't frightening, because I have free will. I don't need to jump. There's no algorithm to deliver me to that pre-destined abyss, nor spur me into action. The only pattern that matters to me is that after I add all of the options premiums, dividends and capital gains or losses together, it exceeds the index.

I can understand that pattern, and at least to some degree, I can control it, as well.

I have no clue as to the precise mathematical correlations and the velocity of Mosaic's moves. All I know is that it does so on it's own time and I'm patient.

These days, patience helps, holding shares of a number of stocks that haven't fared terribly well the past couple of weeks.

Patience and faith, if you believe in that sort of thing.

Why can I never find a theologian when I need one? I saw one just nine months ago.

Disclosure: I am long MOS, MS.

About this article:

Author payment: $35 + $0.01/page view. Authors of PRO articles receive a minimum guaranteed payment of $150-500. Become a contributor »
Problem with this article? Please tell us. Disagree with this article? .