How Market Crashes Happen

Charlie Munger has long championed the benefits of acquiring worldly wisdom through the development of mental models which utilize a number of different subjects — physics, biology, sociology, psychology, finance, philosophy, literature, etc. — to understand the world and make better decisions.

As far as I’m concerned, Michael Mauboussin is Munger’s heir apparent in this effort. All of his books (the most underrated, in my opinion, being Think Twice: Harnessing the Power of Counterintuition), white papers, speeches, and interviews include thought-provoking studies, anecdotes, data and ideas from a wide variety of fields.

In his latest missive for Credit Suisse, Mauboussin tackles the active versus passive investment debate. This is one of the most well thought-out pieces I’ve read on this topic. It’s an investment nerd’s (i.e., me) dream.

Mauboussin brings up a study performed by Blake LeBaron, an economist at Brandeis University who has performed in-depth research on asset prices and market crash dynamics.

One of the more compelling arguments I’ve heard about the current state of the markets is that they’re more or less micro efficient and macro inefficient. The idea is that it’s becoming extremely difficult to pick securities, but the market as a whole can still become unhinged from reality. These explanations are never quite this simple, but it seems like the past 15 years or so has followed this blueprint.

Here’s LeBaron’s take on how market crashes tend to transpire:

During the run-up to a crash, population diversity falls. Agents begin to use very similar trading strategies as their common good performance begins to self-reinforce. This makes the population very brittle, in that a small reduction in the demand for shares could have a strong destabilizing impact on the market. The economic mechanism here is clear. Traders have a hard time finding anyone to sell to in a falling market since everyone else is following very similar strategies. In the Walrasian setup used here, this forces the price to drop by a large magnitude to clear the market. The population homogeneity translates into a reduction in market liquidity.

LeBaron is basically describing the herd mentality here. Once enough people hop on a trend or idea in the markets the momentum can feed upon itself right up to the point where it can no longer support itself. No one can really predict how far these things can go, but this is a nice definition of how bubbles are created.

One of the hardest parts about the markets is the fact that you know (or should know) eventually they’ll go down. You just never know how much they’ll drop once they start to fall. For example, I took a look at all of the double digit losses in the S&P 500 from 1928 to 2016 along with a split by different magnitudes:


Breaking stock market losses down even further gives you a sense of how often these losses tend to occur over time. Here are the frequencies in which certain loss thresholds have occurred, on average, in this same time frame:

  • 5% losses three times a year.
  • 10% losses once a year.
  • 15% losses once every two years.
  • 20% losses once every three to four years.

Average historical returns never tell the whole story because so few years or cycles ever follow the averages, but these numbers can be instructive. Stock market investors should expect to lose a little money quite often, see a correction occasionally, lose a decent amount every couple years and lose a lot of money on an Olympics-like schedule.

These definitions are by no means scientific, but that’s because it’s difficult to know in advance what will distinguish one type of market drawdown from another. Certainly, there are certain variables that can make some market losses worse than others — economic conditions, valuations, investor preferences, etc. But really it’s the human element that determines when things will go from a correction to a bear market to a crash.

And the problem with trying to predict when these things will happen, why they’ll happen or how far they’ll go is because of that human element. Very few people are able to accurately predict the actions of the entire group of market participants because the markets are made up of all sorts of individuals, businesses and organizations who have competing goals, idea, and desires.

I have no idea how I will personally feel tomorrow morning. How will I ever be able to predict how millions of other investors will be feeling one week, month, year or decade from now?

So here’s my mantra on future stock market losses:

Stocks will have periods of poor returns…

…I just don’t know when.

Stocks will have corrections…

…I just don’t know when.

Stocks will go into a bear market…

…I just don’t know when.

Stocks will crash…

…I just don’t know when.

This realization that you know something is eventually going to happen, but you have no control over when or why it will happen can be extremely liberating as an investor. Understanding what you do and don’t know is a huge step in the right direction in terms of thinking like Munger or Mauboussin.

Looking for Easy Games: How Passive Investing Shapes Active Management (CS)

Further Reading:
Charlie Munger’s Investing Principles


Download PDF

Full Disclosure: Nothing on this site should ever be considered to be advice, research or an invitation to buy or sell any securities, please see my Terms & Conditions page for a full disclaimer.

  • Zaphod

    Thanks for the book tip, sounds perfect. Bought it.

    • Ben

      More Than You Know worth a read too

  • Paul LaVanway

    I believe that should be “Brandeis University.”

    • Ben

      noted. time to fire my editor (me)

  • Alan

    Thanks for sharing this post, the “Looking for easy games…” piece is very interesting. Do you agree that generally investors moving from active to passive does increase market efficiency, and therefore represents weak players leaving the poker table? I would have thought these players are still at the poker table, but are now playing in a robotic / mindless fashion. For example the vast numbers of passive investors in the S&P 500 will presumably lead to inefficient over valuation when money is pouring in, and undervaluation when it pours out. Whereas for companies outside this index (say slightly smaller US companies) if they have a greater proportion of active investment one might expect them to be priced more efficiently.

  • “During the run-up to a crash, population diversity falls.” Measured how?

    Nick de Peyster

    • You can probably measure it using correlations, or some form of cluster analysis… a high performing cluster will become temporarily dominant over everything else.

    • RichardU

      Measured perfectly with hindsight. That’s always the rub. You never know until it is too late. Or you become rich because you did know before everyone else.

      If HFT causes the next crash, it will be obvious you could measure the lack of diversity by looking at the percentage of trade volume dominated by HFT. If debt causes the next crash, you should have measured how much money was invested in debt, and negative rates should have been an obvious signal of the absurdity. If protectionism is the cause, you should have known global trade diversity would fall. Brexit and Trump would be the obvious signals. Or maybe it is central banks buying so much debt, and even stocks. Obviously trillions owned by a single actor isn’t diversity. Or all the money in ETFs.

      So, ask me in five years and I will give you the answer.

  • MG

    I wish more investors really accepted this. I tell people that over the decades they should make more money in a diversified ETF than any other investment, but that they should also expect one or more times to witness their capital plunge by half.

  • There’s a lot more to say here about the dynamics of crashes:

    * where liquidity is, and where it is lacking.

    * how bubble assets usually have to be fed to hold onto them near the end

    * hidden leverage, short dated nature of leverage near the end of bubbles, and other credit market dynamics

    * How prudence typically gets punished before a bubble pops

    and more…

    • Ben

      True, there are certainly signals to get you a little closer. I just look at things like the tech bubble or japan bubble and marvel at how far out of whack things truly got

  • won’t be any crashes for a very long time .this time really is different

    • RichardU

      Said the majority before every major correction.

      1999 really is different because technology doubles in productivity every twelve to eighteen months, and that translates into much faster growth in the economy, and stock market isn’t growing nearly as fast as technology.

      2006 really is different because housing is a safe investment, and people will always need a place to live.

      So please do tell why this time really is different. Courageous investors would like to know what to short.

      • simranjeet singh gandhi

        I think he was being sarcastic 🙂

  • RichardU

    The real problem with predicting how the mob will act is not really being able to predict them, but being able to predict the events which will cause a swift and immediate reaction. Imagine an alternate reality where Amazon filed for bankruptcy in 1998. Or Sun, or even Do you think the runup would still continue, and the crash would still have waited? Or would we have had a lower peak and a smaller decline? What if Bear and Lehman ran into their troubles a year earlier?

    Lately, imagine if Greece didn’t unconditionally surrender to the EU for a bailout. Or if the EU reacted to Brexit by basically immediately ceasing Britain’s membership, and trying to clean up the total mess that would have caused. Or if a major central bank decided enough was enough, and pulled the plug on QE, negative rates, etc. These things were far closer to happening than we probably want to accept. Perhaps ten years from now some of the insiders will write books about just how close we came.

  • John Stanton

    So, what is the conclusion? Just buy in, and hang on, realizing that you will suffer losses of the magnitude written? Or, realizing that this is the case of stock markets, have a strategy to take advantage of this realization?

    • Ben

      The conclusion is to take these facts into account when building your portfolio and do what you have to do to survive them.

      • Oliver Lee

        “and do what you have to do to survive them” Does not mean sit by and lose your short. Have a strategy to protect yourself

  • Nate Eslinger

    Great article Ben. Related to this topic, I’m curious about your thoughts on what impact millennials – who seem to be shying away from the market in droves – might have on future returns. I believe 70% of stocks are held by institutional investors, but if upcoming generations of workers stay out of the market, could that be catastrophic for individual investors such as myself?