Noise - Chapter 5
Sources Of Noise: Price Action
Briefly: Price Action
Noise. Changes in the prices of securities are a source of noise. If you succumb to noise, your performance will be worse than it otherwise would have been.
Extreme. Extreme price action can trigger neurobiological reactions that deprive you of your ability to make sensible decisions.
Unexceptional. Unexceptional price action can lure you into screen-watching and overtrading, which takes small and frequent bites out of available returns.
Changes in the prices of securities can take various forms: choppy volatility, extended trends, or sudden reversals. And they can represent any number of things: disaster, success, warning, opportunity, distraction, or irrelevance. Small intraday price movements might be completely irrelevant to a long term investor, but small intra-second movements might convey valuable information to a high-frequency trader. A change in price is made meaningful to you by its relationship to your investment process and your portfolio positioning.
So changes in price are not necessarily noise, in and of themselves. What makes price action a source of noise is whether it causes you to do something that is inconsistent with your investment process. If changes in price force you to diverge from your process, then those changes are clearly not something to be waved away as an irrelevance, those changes are a source of noise that are likely to lead to the destruction of alpha.
Price action can be measured in two dimensions, amplitude and frequency, and each of these dimensions is the source of a different type of noise, and each has different consequences.
Extreme Price Action
Sudden big price moves are a source of acute noise because they tend to catch you by surprise. This type of price action can trigger in you a neurobiological and psychological chain reaction that results in errors of judgement and in the destruction of alpha.
The neurobiological effect of acute noise, such as that which arises from extreme price action, is that you lose the capacity for conscious and deliberate thought. When you experience this kind of amygdala hijack your metabolic resources are diverted away from the executive centre of your brain, and you are physically unable to think properly. Acute noise radically deprives you of your capacity to think well; it robs you of your ability to generate alpha.
When your psychological equipment is ill-prepared for a sudden surge in cognitive load that is caused by an extreme price move, you will attempt to cope with the situation by relying on simplifying rules of thumb, some of which are operating outside of your immediate awareness. These heuristics are sometimes helpful but they are always imperfect, so your reliance on them makes you prone to errors of judgement. Extreme price action can seriously cloud your judgement precisely when you need to be thinking most clearly. You can temporarily take leave of your senses, abandon your investment process in the heat of the moment, and do things that you later regret.
This might seem like an amateur error, but even the greatest fund managers are not immune to the acute noise that arises from extreme price action.
Example: Soros Fund Management
Stanley Druckenmiller was the manager of Soros Fund Management’s $20 billion Quantum Fund for over a decade. Despite his skill and experience, Druckenmiller became a victim of extreme price action when the tech bubble burst in 2000.
Ironically, he had been long tech shares during 1999 and had then sold out all of his tech exposure in January 2000, a couple of months before the peak. But the prices of tech shares continued to run up very hard, and this price action enticed Druckenmiller back into the market.
He later said, “I bought the top of the tech market in March of 2000 in an emotional fit I had because I couldn’t stand the fact that it was going up so much, and it violated every rule I learned in 25 years.”
He describes the action he took and the consequences: “I put billions of dollars in within hours of the top. And, boy, did I get killed the next couple months.” The Nasdaq peaked at a level of about 5,000 in March 2000 and bottomed out at about 1,200 (-75%) in September 2002.
By his own admission, Druckenmiller could not withstand the noise that was generated by the price action of a market going up so fast. He had spent two and a half decades designing and refining a highly successful investment process, but this episode of acute noise made him abandon his process at a critical moment. That lapse of judgement led to severe losses for his investors and it also led to an abrupt end to his partnership with George Soros.
This seems to illustrate something simple enough: design a decent process and follow that process. It may be simple, but it’s definitely not easy. Even if you are able to design a decent process, what makes this all so very difficult is the ever-lurking threat of noise. Even if you have the mental toughness to stick to your process when it is going through the inevitable fallow patch and your performance is lagging the market and/or your peers, when there is a sudden spike in noise from extreme price action it can be too much to withstand, as Druckenmiller said. But even if you can stick to your process, it is no guarantee that things will work out well.
Despite being long tech at the end of a massive bull market, Druckenmiller had earlier voiced his concern over the possibility of it being a bubble and its inevitable bursting. In making this prediction he had something in common with another giant of the hedge fund world, Julian Robertson. Where they differed was the manner in which they approached the phenomenon of this extreme price action.
Example: Tiger Management
Robertson ran the highly successful hedge fund firm, Tiger Management, whose assets peaked at $21 billion in 1998. He was forced to shut down his firm two years later after a period of underperformance and client withdrawals.
In the late 1990s Robertson refused to play the rapidly rising tech market from the long side, preferring to invest in so-called Old Economy companies that better suited his orientation as a value investor. His dogged determination to stick to his process and not be sucked into extreme price moves showed great ability to withstand this source of noise. But noise is a terrible and cunning shape-shifter that will retreat from a battle that looks unwinnable only to appear in another guise elsewhere, perhaps more ferocious and destructive.
Robertson’s apparent immunity to extreme price action was admirable, but this same attribute had the unintended consequence of triggering other sources of noise. His non-participation in the sharp run-up in tech share prices was also part of the reason his funds suffered from a period of severe underperformance, which also turned his clients into a source of noise.
Investors who were dissatisfied with Tiger’s returns started a trickle of redemptions which quickly turned into a tide that the firm could not withstand. Just as the tech bubble burst in spectacular fashion in March 2000, Robertson returned the remaining $6 billion to his investors, pulled down the shutters on his firm, and retired to New Zealand.
Although Robertson was able to withstand the acute noise created by extreme price action and stick to his investment process, in doing so he effectively shifted the vulnerability to noise from himself to his less resilient clients. While he did not abandon his process, his clients abandoned him. He had been fighting noise that arose from extreme price action, but it came back and bit him in a different form.
Neither of these two legendary fund managers, Stan Druckenmiller and Julian Robertson, was able to escape the effects of extreme price action. Although each reacted differently, both were injured by their encounter with noise.
Noise affects everything that you see and think, including your perception and analysis of risk. When asset prices have risen, say at the late stage of a bull market, perceptions of risk change. Despite elevated prices, you can perceive risks as being low when they are actually high. This is a latent form of noise because its threats are veiled and obscured. It’s noise, but just not easily perceptible noise, and it’s noise for the reason that it will lead you into temptation, it will lure you away from your investment process. This latent noise makes you more prone to all sorts of cognitive biases, chief amongst them being overconfidence and confirmation bias, which is simply another way of saying that you can be duped by noise into misperceiving risk, often with seriously bad consequences.
So, precisely when risks are rising, perceptions of risk can be falling. And the converse is true: after abrupt falls in prices there is an elevated perception of risk when risk might have fallen. Here the acute noise that arises from extreme price action is manifest rather than latent, it’s noisy and, boy, don’t you know it. But even knowing that it’s noisy is not enough; just knowing that there’s a hungry lion in the long grass is not enough of a defence against being eaten by the lion.
At the end of the Global Financial Crisis the massive decline in prices was accompanied by the perception of elevated risk levels, when in fact risk was ebbing away precisely because prices had over-reacted to reality. There is no denying how scary things were for a while and there was no shortage of credible scenarios indicating that everything would end in a smoking heap, but the scariness had outpaced the reality on the ground. This meant that the risk/reward relationship had become attractive but it was very difficult to see at the time because price action had created noise that interfered with clear thinking. Price action had triggered heuristics and biases that were very difficult to overcome for those who were less able to manage noise. The cognitive biases that are most closely associated with the acute and manifest noise that arises from falling prices are loss aversion (as price declines extend and accelerate), and irrational extrapolation (as panic sets in and noise traders capitulate).
Those who were best able to manage this acute and manifest noise, which meant not being suckered into equating falling/fallen prices with rising/elevated risk levels, were the ones who were able to see the opportunities more clearly and take advantage of generalised noise and turn it into significant alpha. Very few fund managers participated in the extraordinary opportunities that existed; most had experienced severe amygdala hijack and had begun to fear that everything was going to zero. And there was plenty of opinion that underpinned this erroneous belief, which made it even harder to take advantage of the opportunities that this dislocation had presented because the noise created by the severe decline was so acute and destructive. It’s very hard to stand apart from the crowd when prices are accelerating down and opinion is following those price moves and affirming and confirming what has already happened.
Unexceptional Price Action
It doesn’t have to be all sturm und drang for price action to be a source of noise. The small and unexceptional moves that reflect the market’s normal behaviour on most ordinary days are less obviously threatening to your ability to generate alpha but, even though this price action doesn’t make headlines, it still can be a problem. It is the ongoing normalcy of these small moves that makes them seem benign when in fact they are a source of chronic noise that detract from your ability to generate alpha.
Any move in prices that distracts you from your process is noise. So if you find yourself watching small intraday moves on your screens, even if they are not dramatic, that price action is noise. The very act of diverting your attention to doing something which is unlikely to result in alpha-generation indicates the presence of noise.
There are two related ways in which unexceptional price action destroys alpha, the first of which is screen-watching. Fund managers know that this activity is thoroughly counter-productive and yet most fund managers do it, and those that acknowledge doing it tend to vastly underestimate the amount of time that they spend doing it. This activity captures your attention and diverts scarce resources away from properly alpha-generative work, like doing some deep analysis in a low-coverage area. It’s way less taxing to watch a screen than to think critically or creatively, but that is part of its narcotic appeal and it naturally carries a significant opportunity cost. It is also viciously circular: the more that you are sucked into watching unexceptional price action, the more price action you will be exposed to and the more compelling your screen-watching becomes. Little wonder that so many fund managers are addicted to this activity. You’re an addict when you know that something is bad for you but you can’t stop doing it.
The second way in which unexceptional price action destroys alpha is that when you allow yourself to be lured into screen-watching, the frequency of your price observations naturally rises; and the more frequent your observations, the more frequently you will tend to trade. If your trades are executed without a sound basis in your investment process or without an informational edge, then your overtrading is going to eat up alpha.
Screen watching is an apparently benign activity but one that needlessly squanders scarce cognitive resources, and it is a gateway to overtrading and higher portfolio costs. While this makes for a far less dramatic story than Stan Druckenmiller clipping the very top of the tech market, the indirect and direct costs associated with the chronic noise that arises from unexceptional price action are costs that are unnecessarily incurred by far too many fund managers and which compound over time and take a bite out of available alpha. The good news is that this is one of the more ready sources of additional return that you are likely to encounter; it’s practically begging to be captured.
References
Kahneman, D. (2011) Thinking, fast and slow. New York, New York: Farrar, Straus and Giroux.
Interview with Stanley Druckenmiller on Bloomberg Television in November 2013.
Letter from Julian Robertson to the limited partners of Tiger Management, LLC, 30 March 2000.
Article published by Bloomberg on 17 April 2000 by Gary Weiss: ‘What really killed Robertson’s Tiger’.