PsyFi Search

Friday 17 October 2014

Facing Down The Narrative Fallacy

Frenzy Time

Markets are tumbling. It's because the Fed is about to push up interest rates. Or maybe because the German economy is weak. Or perhaps it's Ebola, about to decimate global populations. Or it's geopolitical conflict - Ukraine, Syria, Iraq ... take your pick. It's a perfect storm. Head for the hills and don't spare the horses. And remember the shotgun.

Of course, it's none of these. Markets are falling because they were a bit overpriced and investors were ignoring the fact because they'd got themselves into a typical feeding frenzy, ignoring the risk and ambiguity that are always present. Now that they have recognized the issue they're fighting each other to get out the door. Which is why all of the explanations for market weakness are entirely plausible and entirely wrong - they're an example of what Nassim Taleb calls "the narrative fallacy".

Excess Volatility

Investors find it really hard to believe that stock prices can move for no real reason at all - and we have an entire industry of pundits whose job it is to produce those explanations and feed those beliefs. Yet the reality is that shares often do have a life of their own, and if they randomly start to move in a certain direction then that can cause a cascade effect. 

In fact, as we saw in Volatility, The Last Anomaly, Robert Shiller has found that there's up to thirteen times the expected volatility in markets - stock prices move around far, far more than can be explained by fundamentals. Yet we can't just accept this - we need some story that explains why each and every market fluctuation happens - it's the modern equivalent of the volcano god, spewing lava everywhere because it's had a bad day at the office.


Eli Ofek and Matthew Richardson have argued, in The Valuation and Market Rationality of Internet Stock Prices that it's heterogeneity of beliefs among investors that causes this cascade effect and, with it, the excess volatility in the system. In essence, if everyone is operating on the same set of values and expectations then everyone will tend to react in the same the way to the same signals. It doesn't help that a lot of investors rely on price movements and other investor behavior to determine their actions.

The type of events that typically trigger market falls include an expectation of interest rate rises, global health scares, concerns about economic growth and wars. As we have all of these at the moment it's not surprising that investors have suddenly woken up the fact that risk hasn't abolished itself and have started to sell stocks. Given that the markets were probably a bit overvalued to start with the cascade effect isn't surprising. Additionally, of course, when investors become blasé about risk they tend to push up the prices of poor companies along with good ones - and then panic when prices fall and they can't find another fool to pick up their stakes.


Some research by Philipp Illeditsch in Ambiguous Information, Portfolio Inertia and Excess Volatility suggests that it's unexpected news flow that investors don't know how to interpret that causes the excess volatility. Essentially investors will try to hedge against uncertainty to the best of their ability but when some unexpected event shows that they don't really understand the extent of ambiguity in the market - JFK gets assassinated, 9/11, the collapse of Lehman Brothers, etc, etc - then their reaction is to flee for the hills. As we saw in Ambiguity Aversion: Investing Under Conditions of Uncertainty, people can be surprisingly calm about risk, but absolutely averse to ambiguity. 

Taleb's argument is that we require stories in order to link our thoughts and memories together, and that they are a very effective way of doing so. He defines the narrative fallacy as our
"Need to fit a story or pattern to a series of connected or disconnected facts". 
Unfortunately this propensity can, like so many of our innate behaviors, lead us into poor investing behavior. In a situation in which not every fact has an explanation our tendency to demand one can cause us to create a set of self-fulfilling actions: stocks are falling because of rising interest rates/Ebola/ISIS/Ukraine (delete as applicable) therefore we need to sell stocks. Only the facts don't require any explanation other than the obvious one - stock prices are falling because people are selling, and people are selling because they believe that there is a reason for doing so.


Of course, it's possible that one of these current issues may turn out to be a historically defining event - World War Three may happen, we make see a repeat of the Great Depression, Ebola may wipe out half the global population. It's possible, but we have no way of predicting it, and we're probably better off operating on the basis that none of those things will happen: stocks are cheaper than they were a while ago and that means that increasingly we can find quality companies at reasonable prices.  

It's also possible that these reasonable prices may become even more reasonable in future. Indeed, they may become downright cheap: deflationary fears are a real concern, if nowhere near the sure fire certainty that many commentators are suggesting. But a sensible long-term investor wouldn't be selling at this point, they'd be buying. Carefully and selectively to be sure, but not fearfully; because good companies selling things that people actually want will always find a market - and if their stock price or even their earnings fall tomorrow then they'll come back eventually.


In fact, in a perfect world, a good investor wouldn't be holding dubious stocks at all when the market falls, and probably wouldn't feel the need to sell, or seek spurious explanations or worry overmuch about volatility. Because volatility is a fact of life in markets, and not something that requires us to create a complex and probably spurious story to explain it. 

The idea that we should brace ourselves and do the opposite of what the market is telling us to do requires that we accept the idea that the market can be wrong, at least in the short-term. It's contrarianism, and it's never going to be an easy path unless you can adopt a mindset that means you don't require a short-term story to justify short-term market movements. If you can accept that stuff happens in the short-term then the only long-term story you need is the graph showing the trajectory of the stock market over the last hundred years.

It's a story that will end one day, but we'd have to be unlucky to be around when it happens. In the meantime we're all better off facing down the narrative fallacy.


  1. "Given that the markets were probably a bit overvalued to start with"

    Given by whom? Can we get more tips from that great oracle?

  2. You're such a fool! You wiret a pretty good article and then you go and spoil it with an asinine comment about the trajectory of the stock market over the last hundred years. Don't you realise that the "trajectory" you talk about is like saying the bus travelled between New York and Chicago, therefore all the passengers went for New York and arrived in Chicago.
    It appears to have escaped your pea-brain that many of the passengers may have got off the bus long before it reached Chicago and been replaced with other passengers.
    So it is with the stock market. Do you realise that ONLY ONE STOCK (General Electric) is still with the Dow since inception?
    In fact the indices are unrecognisable today in terms of their content from the 1970s.
    So your entire thesis is meaningless. Sorry.

  3. Interesting as always with great citations. The thing is, we can't not narrate. Assembling a sentence is a kernal story. Your assertion that "Markets are falling because they were a bit overpriced and investors were ignoring the fact because they'd got themselves into a typical feeding frenzy ..." is itself an explanatory story.

    Further, your assertion "that stock prices can move for no real reason at all" conflates randomness with complexity. Forllowing Eli Ofek and Matthew Richardson argument, stocks prices change for so many disparate beliefs, reasons, intentions, actions that simple narrative cause-effects cannot account for them, but can account for cascades. James Simon and others attest that there are other, statistically significant, forms of analysis.

    I agree "that it's unexpected news flow that investors don't know how to interpret that causes the excess volatility," provided we add that a "news" event is a clear, simple and vivid image, or can be made so, such as your examples. In this neurological sense, images trump facts. Or more accurately, facts assume their place and importance - as vivid events - in a particular and simple story.

    So, the question is what stories are you listening to (including the ones you tell yourself)? You give an important wellformed investor story condition near the end - have it unfold over a longer period of time.