The Better Letter: Crash Protection

What we can do about the coming market crash.

When Wall Street returns from the Hamptons after Labor Day, the worrying starts.

Historically, September is the worst month for the markets, with late September particularly problematic. October is a bit better, but the worst two market crashes in history (1929 and 1987) took place then.

This time of year brings many “bear sightings,” both real and imagined, and lots of predictions about whether we’re headed into bear territory.

After every football victory, the California Golden Bears declare that the ground on which they stand is Bear Territory, but in their case it’s a good thing.

“You know it! What? You tell the story! What? You tell the whole damn world this is Bear Territory!” 

Watch a particularly lively rendition after a win in The Big Game over Stanford and Andrew Luck below (starting about 1:20 in). My youngest is at the top of the screen during the singing, number 46 for Cal.

We have seen a fair amount of market difficulty lately, despite a strong recovery yesterday, causing some to wonder if we’re headed toward the sort of bear territory Wall Street fears, especially this time of year. How we might think about that possibility is the subject of this week’s TBL.

If you like The Better Letter, please subscribe, share it, and forward it widely. It’s free, there are no ads, and I never sell or give away email addresses.

Share The Better Letter

Thank you for reading.


Crash Protection

Air France Flight 447, headed from Rio de Janeiro to Paris, vanished utterly very early on the first day of June 2009. In the middle of the night. In the middle of the ocean.

There were no survivors, no witnesses, no Mayday call, and no radar trail.

The Airbus A330 was among the safest planes in the sky, equipped with the automated fly-by-wire system, designed to reduce human error by using computers to control nearly all aspects of the flight. Since the fleet’s introduction, in 1994, not a single A330 in line service had crashed.

Until then.

At 1:35 a.m., the pilots routinely called in their altitude and flight plan. Three seconds later, Brazilian air traffic controllers called back to ask when they would reach Tasil Point – the spot in the middle of the Atlantic where control of the flight would shift from Brazil to Senegal. Another try after seven more seconds passed. Another six seconds, another attempt.

Flight 447 was gone.

In all, 228 souls were lost, most still strapped in their seats as they settled into a watery grave two-and-a-half miles below the surface after the plane pancaked into the Atlantic.

An experienced and carefully-trained crew flying a modern airliner abruptly lost control of their aircraft during a seemingly routine flight. The causes – it wasn’t just one thing – were remarkably prosaic. A series of small errors turned a state-of-the-art cockpit into a death trap (the gripping details are here). All of their crash protection mechanisms failed.

Safety scientists call this sort of problem the “Swiss cheese” model of failure, when the individually small holes in organizational defenses line up in ways that had not been foreseen, with catastrophic results. This mode of crisis requires rapid interpretation and response, with many ways to fail.

Aircraft automation is a major positive in the aggregate as fewer human decisions mean fewer opportunities to make mistakes. Since the 1980s, the safety of flying has improved by a factor of five. However, more managing and monitoring with less true “flying” can limit a pilot’s ability to respond in a crisis. Moreover, technology-dependence can erode basic cognitive skills.

As automation increases and improves, flying is simpler and there is less room for human error. However, most pilots’ flying experience includes lots of sitting in a cockpit seat watching the controls while computers do their thing. If (when) automation fails, crisis complexity is greatly increased and the pilot’s ability to handle it greatly decreased. There is both a flight problem, for which the pilot has much less experience, and a technological problem, for which the pilot is not likely equipped.

When complex technology fails, the humans operating it become confused and don’t always have the capabilities to deal with it. The University of Michigan’s Nadine Sarter calls it a “systemic problem with complexity.”

I once asked an airline pilot friend, now retired, if he’d ever faced anything truly challenging flying commercial jets. He said, “No, but remember that when I served in Vietnam, it was routine to engage the enemy and then land my fighter on an aircraft carrier at night, without lights, in a storm, in a war.” Few of today’s commercial pilots have anything like that sort of experience.

Poor human performance begets automation, which improves outcomes but worsens human performance, which begets more automation. Rinse. Lather. Repeat.

It is a paradox of *almost* totally safe systems – which reminds me of another sort of crash – within a far more complex system that is far from safe.

It was on a Monday, nearly 34 years ago, that the stock market suffered its worst day ever. I was still practicing law then, and I was working in my firm office. As the news filtered out (somebody had been out and heard the news on a car radio), my colleagues and I began to talk about what was going on. Ignorantly.

As the day progressed and the news got worse, we talked less and less. A senior partner brought out a tiny black and white television to watch the network news coverage, which had broken into the day’s typical programming of reruns, game shows, and soap operas.

We huddled around the screen and watched. Silently.

It was Black Monday: October 19, 1987. Coverage by The Wall Street Journal the next day began simply and powerfully: “The stock market crashed yesterday.” 

After five days of intensifying stock market declines, the Dow Jones Industrial Average lost an astonishing 22.6 percent of its value, plummeting 508.32 points after a 5-year run from 776 in August of 1982 to a high of 2,722.42 in August 1987. For its part, the S&P 500 dropped 20.4 percent. Black Monday’s losses far exceeded the 12.8 percent decline of October 28, 1929, the start of the Great Depression.

It was “the worst market I’ve ever seen,” said John J. Phelan, Chairman of the New York Stock Exchange, and “as close to financial meltdown as I’d ever want to see.” Intel CEO Andrew Grove said, “It’s a little like a theater where someone yells ‘Fire!’”

A closer look at the coverage and its aftermath is revealing, especially for what is missing. There is no “trigger” – no clear cause. The Fed later identified “a handful of likely causes.” Then Fed Chair Alan Greenspan has written a lot about Black Monday and about what he did without trying to explain why it happened.

According to the Big Board’s Chairman, at least five factors contributed to the record decline: the market’s having gone five years without a major correction; general inflation fears; rising interest rates; conflict with Iran; and the volatility caused by “derivative instruments” such as stock-index options and futures. Although it became a major part of the later narrative, Phelan specifically declined to blame the crash on program trading alone.

The Brady Commission, appointed by President Reagan to provide an official explanation, pointed to alleged causes like the high merchandise trade deficit of that era, and on a tax proposal that might have made some corporate takeovers less likely. Yet there is precious little evidence to support those claims.

Survey evidence accumulated by the Nobel laureate Robert J. Shiller confirmed generally that traders were selling simply because other traders were selling. It was classic herd-behavior and, essentially, a self-reinforcing selling panic. As Shiller explained, Black Monday “was a climax of disturbing narratives. It became a day of fast reactions amid a mood of extreme crisis in which it seemed that no one knew what was going on and that you had to trust your own gut feelings.”

Here’s the bottom line: the Black Monday collapse — the biggest in history by a country mile — had no definitive (or even clear) trigger. The market dropped by almost a quarter for no obvious reason. So much for the idea that the stock market is efficient.

That shouldn’t be news. As Shiller has demonstrated, “The U.S. stock market ups and downs over the past century have made virtually no sense ex post. It is curious how little known this simple fact is.” Here is Shiller’s graphical proof.

As Paul Singer, of hedge fund fame, noted, “The important turning points in markets are never identified with precision in advance by ‘experts’ and policymakers. This lack of foresight is not surprising, because markets and the course of the economy are not model-able scientific phenomena but rather are examples of mass human behavior, which are never predictable with anything like precision.”

As Ruth Gordon’s Maude says to Bud Cort’s Harold in Harold and Maude: “Consistency is not really a human trait.”

Markets were already crazy-complex when the internet and our dependence upon it ratcheted up the complexity exponentially, thus offering the potential for much, much more chaos and unintended consequences.

“@Jack wants a text-based version of CB radios and the media and political systems get upended. Mark Zuckerberg wants to let Harvard kids hook up more easily and the entire news business blows up.”

In essence, there is all too much complexity, chaos, and chance built into a system of too many variables for us to be able to explain it very well, much less predict it.

Barron’s claimed recently that every major downturn requires a “trigger.” That’s clearly and obviously false, unless the last grain of sand landing on the pile triggered the avalanche that followed.

Once in a great while, though, somebody gets a major downturn prediction right, perhaps coincidentally. Lots of big careers have been made being right about some big thing once in a row.

But getting it right once doesn’t make someone any more likely to keep being right (see Paulson, John and Josh Brown’s excellent analysis here). Because of our general loss aversion, we are remarkably vulnerable to strong claims of impending doom. The usual suspects continue to expect and predict the worst. Of course.

Someday one of them might eventually (if only coincidentally) be right but, even as and while we’re waiting, they get both regular hits on “business” television and attention from the press. And there is precious little accountability when they are wrong. We are no better in the aggregate (and are probably substantially worse) at predicting market crashes than we are at predicting the markets in general — which is to say, we are truly dreadful at it.

While it’s counterintuitive, these observations are wholly consistent with catastrophes of various sorts in the natural world and in society. Wildfires, fragile power grids, mismanaged telecommunication systems, global terrorist movements, migrating viruses, volatile markets, and the weather are all complex adaptive systems that evolve to a state of self-organized criticality. Upon reaching the critical state, these systems then become subject to cascades, rapid downturns in complexity from which they may recover but which will be experienced again repeatedly, if somewhat differently.

It sometimes ends up looking a lot like this.

There is persistent and unavoidable uncertainty inherent in market activity, which means that we should not expect ever to be able to identify the trigger of any correction or crash in advance or to be able to predict such an event with any degree of specificity, no matter how low interest rates go, how overvalued the stock market is, or how many alleged experts predict gloom and doom.

Crashes are inevitable, of course, despite Janet Yellen’s hopes and dreams. But crash protection isn’t a thing. Unless you get very lucky.

The most we can expect are inventive and powerful narratives created after the fact to try to explain what we all missed at the time.

Today could be a day the market crashes. However, there is no reason to expect to see it coming. Crash protection is not a thing.


Totally Worth It

Broadway is back! Be sure to catch Glinda’s opening line and the reaction to it (at about the 5:30 mark).

This week brought Earth, Wind & Fire Day!

After a brilliant but turbulent career, Cat Stevens is reemerging on the public stage. How should we feel about him — and his music — now?

This is the best thing I saw or read this week. The most heart-wrenching. The most insightful. The most important. The most powerful. The most interesting. The most comforting. The craziest. The most bizarre. The coolest, unless it was this or this. The most predictable. The most California. The most ridiculous. The most horrific. The most riveting. The kindest. The stupidest. The ickiest. The best (societal) news. Quite the lesson. 180 million views can’t be wrong.

Feel free to contact me via rpseawright [at] gmail [dot] com or on Twitter (@rpseawright) and let me know what you like, what you don’t like, what you’d like to see changed, and what you’d add. Don’t forget to subscribe and share.

Share The Better Letter

Of course, the easiest way to share TBL is simply to forward it to a few dozen of your closest friends.

Here is some beauty for its own sake.


Benediction

A subscriber created a TBL playlist on Spotify here. Thank you, Perry! Be sure to play it and turn it up.

Last, but best of all, I have a beautiful new grandson, born to my youngest (shown in the video up top) and his wife this week. He inspires this week’s benediction.

Thanks for reading.

Issue 81 (September 24, 2021)