8 Big Outliers That Should Revert to the Mean

Hedge fund manager John Paulson made billions betting against the housing market. You've probably heard the story.

Less well known is where Paulson got the idea that housing was a bubble. Wall Street Journal reporter Greg Zuckerman describes the epiphany in his book, The Greatest Trade Ever Made.

Here's what happened. 

In the middle of last decadePaulson and one of his analysts, Paolo Pellegrini, had a feeling housing was overheated. But not everyone agreed. Only in hindsight is it screamingly obvious. 

Paulson wanted Pellegrini to really prove prices were a bubble. Late one night in 2006, Pellegrini put together a chart of housing prices adjusted for inflation over time. It probably looked something like this:

Source: Robert Shiller.

"The answer was in front of him," Zuckerman wrote. Housing prices flatlined after inflation for most of the last century. Then, early last decade, they zoomed higher.

"The upshot," Zuckerman wrote, "U.S. home prices would have to drop by almost 40% to return to their historic trend line."

Which is eventually what they did.

Paulson had deep knowledge of the mortgage market, but the core of his housing thesis rested on a basic concept: reversion to the mean. It's the simple idea that, as investor Dean Williams once put it, "something usually happens to keep both good news and bad news from going on forever." Big outliers tend to be short-lived. Crowds don't tolerate excess, so something happens to pull outliers back toward a happy medium. Averages act like gravity. 

Reversion to the mean is probably the second most powerful law in finance, behind compound interest. Booms follow busts and busts follow booms. It happens over and over again, in all kinds of fields from finance to business to biology to physics. 

Here are eight trends detached from their long-term average that should revert to the mean.

1. The percentage of workers who want full-time work but can only find part-time hours.

Source: Bureau of Labor Statistics.

Millions of workers who want full-time work were pushed to part-time hours when the financial crisis hit in 2008.

The trend is getting better, with workers labeled as part-time for economic reasons falling for the past three years. It'll probably keep reverting to the mean as businesses that slashed payrolls struggle to expand. Take this recent story about Wal-Mart (NYSE: WMT  ) :

Walmart, the nation's largest retailer and grocer, has cut so many employees that it no longer has enough workers to stock its shelves properly, according to some employees and industry analysts. Internal notes from a March meeting of top Walmart managers show the company grappling with low customer confidence in its produce and poor quality. "Lose Trust," reads one note, "Don't have items they are looking for -- can't find it."

This can't last. So it won't. Returning to average would push more than a million part-time employees back into full-time work.

2. New home construction.


Source: Federal Reserve.

We're not building enough homes to keep up with household formation. We've been able to get away with this for a few years because we built too many homes during the housing bubble. But that excess is soaked up, and now we've gone the other direction, with the supply of for-sale homes at the lowest level in years. New home construction will very likely need to double from current levels sometime in the next decade to keep up with household formation.

3. Corporate profits.

Source: Federal Reserve, Bureau of Economic Analysis.

Three things happened to big businesses over the past decade: Technology allowed them to cut workers, interest rates fell, and rapidly growing markets like China and Brazil opened vast new markets to do business. Combine the three, and profit margins -- the percentage of each dollar of sales that goes to profits -- surged to an all-time high.

Profit margins have always mean-reverted. We don't know when, and we don't know what will drive margins lower -- maybe higher wages, maybe higher commodity prices -- but you can be confident they will. (Importantly, this isn't necessarily bad for stocks.)

4. People and businesses investing in their future.

Source: Federal Reserve.

When the economy slows, businesses cut back on investment. When businesses cut back on investment, fewer people are employed. When fewer people are employed, the economy slows. When the economy slows, businesses cut back on investment. It's a vicious cycle. But it eventually ends. Investment in factories and buildings as a percent of GDP fell to an all-time low in 2010. At some point structures need serious investment just to maintain business operations, at which point fixed investment rebounds. It's already done that, but we're still well below normal.

5. Auto sales per household.

Source: Census Bureau, Federal Reserve.

Auto sales are back to pre-recession levels last seen in 2007. But there are several million more households today than there were in 2007. Adjust for that, and auto sales are still good amount below historic norms.

There are two conflicting forces affecting sales here. One, Americans are driving less. That lowers demand to replace cars. But, two, the average age of cars on the road is now near a record of 11.4 years. Those clunkers will need to be replaced. Auto sales should continue to rise, which is good news for Ford and General Motors.

6. State and local government spending contribution to GDP growth.

Source: Bureau of Economic Analysis.

Cutbacks from state and local governments have been one of the most overlooked causes of our slow recovery. State and local spending -- which includes teachers, firemen, police, etc. -- has historically added more than three-tenths of a point to annual GDP growth. Over the past four years, it's subtracted about the same amount as austerity set in. But things should start to stabilize soon as tax revenue increases. 

7. Gold prices adjusted for inflation.

Source: Federal Reserve.

Gold has maintained its value for centuries. It returns approximately nothing, after inflation. But that's all it does. That's all it should do. So when the price of gold rises well above its historic real (inflation-adjusted) price, you should expect a correction. We've already had one -- gold is down about 30% since late 2011 -- but you can make a strong case that there's more decline to come. My colleague Alex Dumortier has been making this case for years.

8. Google searches for the word "unemployment."

Source: Google.

This will eventually fall to normal levels. Seriously. Things will get better

 

For more on what might be the biggest trend waiting to revert to the mean -- federal spending and the national debt -- check out my report "Everything You Need to Know About the National Debt." It walks you through with step-by-step explanations about how the government spends your money, where it gets tax revenue from, the future of spending, and what a $16 trillion debt means for our future. Click here to read it. 

 


Read/Post Comments (37) | Recommend This Article (62)

Comments from our Foolish Readers

Help us keep this a respectfully Foolish area! This is a place for our readers to discuss, debate, and learn more about the Foolish investing topic you read about above. Help us keep it clean and safe. If you believe a comment is abusive or otherwise violates our Fool's Rules, please report it via the Report this Comment Report this Comment icon found on every comment.

  • Report this Comment On September 11, 2013, at 11:07 PM, SkepikI wrote:

    Morgan: A seminal if optimistic article. One of your better efforts, including your selection of charts/data. And, for a wonder, the specific areas you chose to review are generally immune from bad data collection because they are so broad.

    For what its worth, I believe more inquiry along these lines are better value for your readers.

    although, I do need to caution you about the implicit assumption that reversion to the mean in state and local spending as a % of GDP being a GOOD thing. In my neck of the woods, if the Oregon and Washington state legislatures succeed at increasing spending, it will most likely include a $3 Billion dollar bridge that ought to be had for less than half of that. Corruption and waste increases in government spending are not a great thing...

  • Report this Comment On September 12, 2013, at 12:45 PM, flyboy0100 wrote:

    It also depends on what mean you are referencing and what timeframe.

    Some cheeky examples:

    The global human population: http://www.thestreet.com/content/image/73485.include

    The mean atomic composition of the universe suggests we should all revert to hydrogen and maybe a few helium atoms.

  • Report this Comment On September 12, 2013, at 1:06 PM, kyleleeh wrote:

    <<The mean atomic composition of the universe suggests we should all revert to hydrogen and maybe a few helium atoms.>>

    When has that mean ever become skewed? at 70% H2O by body weight we are made up of mostly of hydrogen atoms...always have been.

  • Report this Comment On September 12, 2013, at 1:19 PM, ryanalexanderson wrote:

    I don't think one should refer to the price of gold in isolation, without talking about bonds and debt.

    Gold will not mean-revert until the idea of ZIRP and quantitative easing are once again fantasy ideas, as they were in, say, 2003.

    For ZIRP to be a fantasy, the economy is going to have to chug along happily for a few years with a fed funds rate of 4% - not the ~0% that it's been for the last five. And the bigger the debt explodes, the more unlikely that particular 'mean reversion' will ever happen with today's US dollar.

  • Report this Comment On September 12, 2013, at 1:29 PM, TMFHousel wrote:

    <<I don't think one should refer to the price of gold in isolation, without talking about bonds and debt.>>

    It wasn't in isolation; it was adjusted for inflation, which seems like a reasonable thing to do for an asset people buy as a hedge against inflation.

  • Report this Comment On September 12, 2013, at 1:47 PM, ryanalexanderson wrote:

    <<It wasn't in isolation; it was adjusted for inflation, which seems like a reasonable thing to do for an asset people buy as a hedge against inflation.>>

    Definitely reasonable. But some fans of gold such as myself have the thesis that the Fed will never be able to tighten very much again. The total credit market debt-to-GDP ratio is north of 350%, compared to its pre-eighties mean of around 150%. Our growth during this time has come from a debt growth fuelled by a steady reduction in interest rates from 1980 highs of ~20% to the near zero rates at which we are currently stuck.

    It is this trap that makes me have a reasonable amount of net worth in gold. Escape from this trap, I believe, is only available through default deflation or deliberate inflation. I think gold will prosper either way. (as part of a balanced portfolio with productive assets!)

  • Report this Comment On September 12, 2013, at 1:47 PM, MAACPRIME wrote:

    Another stellar piece of writing.

    I'd like to see an article on how the big banks and emerging markets may revert to the mean - both tend to boom and bust with a fair amount of regularity.

  • Report this Comment On September 12, 2013, at 1:48 PM, 45ACPbullseye wrote:

    Nice one Morgan! I do wonder a bit about #1, part-time workers. My sense is that we are seeing a "new normal," driven primarily by the cost of benefits for full time employees. Best, Bill

  • Report this Comment On September 12, 2013, at 1:58 PM, ryanalexanderson wrote:

    Morgan,

    ...just a follow-up to what I wrote. The total credit market debt to GDP ratio is shown in this chart here:

    http://i.imgur.com/1s0nach.png

    Would you count this among your mean reversion charts? It's a terrifying thought.

  • Report this Comment On September 12, 2013, at 2:00 PM, TMFHousel wrote:

    <<The total credit market debt-to-GDP ratio is north of 350%, compared to its pre-eighties mean of around 150%.>>

    Total debt-to-GDP will come down, but there's no evidence that it's mean reverting to a pre-1980s levels. In any case it already has come down, for five years straight.

    http://research.stlouisfed.org/fred2/graph/fredgraph.png?&am...

  • Report this Comment On September 12, 2013, at 2:01 PM, TMFHousel wrote:

    I actually take back what I said. It's reasonable to assume it will mean revert over time.

  • Report this Comment On September 12, 2013, at 2:02 PM, TMFHousel wrote:

    <<It's a terrifying thought.>>

    I don't think it's that terrifying. The last time we had a big deleveraging was 1945-1975. That was one of the best economic periods in American history.

  • Report this Comment On September 12, 2013, at 3:18 PM, ryanalexanderson wrote:

    Fair enough. My only response is that since 1980, we are used to this number going up. It's been a big tailwind, that has helped to replace the demographic/women-entering-workforce/manufacturing tailwind that the US had in the 1945-1975 period you mentioned.

    Anyway, thanks for your article and subsequent responses.

    Ryan

  • Report this Comment On September 12, 2013, at 6:44 PM, kakkineni wrote:

    The mean you are using is a historical mean that includes future periods. Hence, why it's a flat line, that overstates the mean following above mean periods, and understates the mean following below mean periods. For example, the mean of the 50s could be around 15% in your 4th section. What this means is that it could be implied that we are at the mean right now for private fixed investment.

    But because you are using an all-time mean, the 2000s and 1970s, which were way above mean, are dragging the mean up. What would be useful is a cumulative average UP TO but not including each year, to compare that year to.

  • Report this Comment On September 12, 2013, at 6:52 PM, FriedBrain wrote:

    <<The mean atomic composition of the universe suggests we should all revert to hydrogen and maybe a few helium atoms.>>

    Cheekiness aside, flyboy0100, there's a problem with those examples.

    In order to apply the principle of mean reversion, one must be able to *establish* the mean to which something will revert. In the case of population growth, it's more helpful to look at the growth rate, rather than the overall total, which fails to control for things like discovering new resources and habitable land.

    And your second example forgets to take several scientific principles into account, starting with entropy. Which gets back to the previous point about understand what the mean actually *is*. Statistics should follow reality's lead, and not dictate terms to it.

    Humor can be a powerful teaching tool, but not if you're establishing a straw-man to use it.

  • Report this Comment On September 12, 2013, at 7:02 PM, AnsgarJohn wrote:

    Ryan,

    What's wrong with only having a diversified mix of productive assets (that create cash flow). Why is it important to have unproductive assets that swing wildly in price as well?

  • Report this Comment On September 12, 2013, at 7:04 PM, FriedBrain wrote:

    So, Morgan, one question: what exactly does -1 Google searches for "unemployment" mean?

    Clearly, I don't *quite* get the meaning of the y-axis in the 8th graph.

  • Report this Comment On September 12, 2013, at 7:18 PM, optimist911 wrote:

    Statistics 101 tells us that the average is a poor choice as a measure of central tendency in all of the above cases, mainly due to the outliers that occur frequently. The median works much better. I would recommend a course or self-study in statistics, because the conclusions of this article only paint a very misleading overall picture.

  • Report this Comment On September 12, 2013, at 7:49 PM, TroyCash wrote:

    I am a strong believer in reversion to the mean and use it as a critical part of my own investments. However I do not share your optimism on the Western World employment situation. For a decade I watched my high tech company reduce its US workforce only to hire three times as many employees for the same price in India and China. Plumbers and retail clerks need to be local but pretty much any job that can be done with a computer can be done in another country at a lower price. I don't see US companies reversing that trend any time soon.

  • Report this Comment On September 12, 2013, at 8:18 PM, jtmc1 wrote:

    I would have liked to see where we are today in the Nationwide Homeprices Adjusted for Inflation chart.

  • Report this Comment On September 12, 2013, at 8:56 PM, ryanalexanderson wrote:

    Hi AnsgarJohn,

    It's a fair question.

    My answer would be that precious metals would outperform productive assets for the following two reasons:

    1) Tools such as quantitative easing are ineffective at combating the deflationary effects of the deleveraging shown in the linked charts. (Morgan's linked chart is more up to date.)

    2) The Fed is going to do it anyway, because it is the lesser of two evils.

    This would be the stagflation scenario. Global economy slows down, so productive assets and consumed commodities are inherently worth less, but the fed prints up paper to compensate for this drop.

    As non-consumed commodities, silver and gold get the upside of the printing, without the downside of the reduced consumption.

    The holding period wouldn't be forever. I would hold it only until we get to the conditions I mentioned earlier - basically, once we get back to a 'strong dollar' policy. It was this policy that Paul Volcker instituted in 1980 that stopped the 1970's bull market in gold in its tracks.

    But the damage that such a strong dollar would do to productive assets -today- would be epic. I don't think it would happen, and therefore I think gold is a prudent holding. And apparently China's leaders do too.

  • Report this Comment On September 12, 2013, at 9:04 PM, kyleleeh wrote:

    << I do wonder a bit about #1, part-time workers. My sense is that we are seeing a "new normal," driven primarily by the cost of benefits for full time employees.>>

    that may be true, but if so it may end up causing the reversion. 100 years ago 60 hours was a normal work week. When overtime was introduced by FDR it was to provide a financial incentive for companies to hire more people instead of squeezing the ones they had. So instead of having two people work 60 hours a week they had 3 people work 40 hours a week. The result of this was that unemployment went down, which made labor become scarce, which caused wages to go up, which made a 40 hour work week all you needed to live off of. The same may end up becoming true if companies try to cut everyone to below 30 hours.

    This is an example of how "something usually happens to keep both good news and bad news from going on forever."

  • Report this Comment On September 12, 2013, at 11:27 PM, spcd7 wrote:

    if we're not careful,deflation will take care of coporate profits.

  • Report this Comment On September 13, 2013, at 12:31 AM, HistoricalPEGuy wrote:

    Morgan - you highlight a wonderful tool for investors. My handle (HistoricalPEGuy) is founded on this exact premise.

    However, it's just one of many signals that must be interpreted given the environment in which you read it. For example, I would be hard pressed to short GLD right after the announcement of QE2, even if it was way above its historical mean.

    There are many who claim that "you can't time the market", but in my experience, it just isn't true. Your analysis exposes a clear measure that can help get in or out at the right moment. Great post, I think you've enlightened many an investor.

    -- HPEGuy

  • Report this Comment On September 13, 2013, at 1:01 AM, dsciola wrote:

    Not sure Im as enthusiastic about this...

    At what point is the mean or even median noise and in actuality we've reverted to a new normal?

    Perhaps tech changes will permanently keep profit margins high...perhaps individual tastes will move more ppl back into the city vs the 'burbs.

    Perhaps i read too much Of Nicholas Taleb...who to me very accurately showed how soft sciences are not hard sciences...the law of gravity in physics is a factual law that won't change...the 'law' of mean reversion is based on human behavior.

    My 2c.

    Dom

  • Report this Comment On September 13, 2013, at 11:17 AM, knighttof3 wrote:

    Excellent article.

    I too don't understand the -1 in the Google search for unemployment. Maybe Google searched the people instead of people searching Google? :-)

    One striking thing about mean reversion is that you cannot predict the cause (or the catalyst - because sometimes it's hard to establish cause-and-effect) and yet the law seems to hold. The problem, as always, is to get the right period to compute the mean. In the first 7 charts, every recent trend started between 2004-2007. It's hard to know if this is a structural change or cyclical. Hey, Rome went on for what, 1000 years until all of a sudden, there were barbarians at the gate? It's rare but "this time is different" can happen once in a while, and nobody can tell you that beforehand.

  • Report this Comment On September 13, 2013, at 11:18 AM, knighttof3 wrote:

    Excellent article.

    I too don't understand the -1 in the Google search for unemployment. Maybe Google searched the people instead of people searching Google? :-)

    One striking thing about mean reversion is that you cannot predict the cause (or the catalyst - because sometimes it's hard to establish cause-and-effect) and yet the law seems to hold. The problem, as always, is to get the right period to compute the mean. In the first 7 charts, every recent trend started between 2004-2007. It's hard to know if this is a structural change or cyclical. Hey, Rome went on for what, 1000 years until all of a sudden, there were barbarians at the gate? It's rare but "this time is different" can happen once in a while, and nobody can tell you that beforehand.

  • Report this Comment On September 13, 2013, at 11:24 AM, Charles82 wrote:

    Hi,

    Sorry, I disagree completely with the concept of "adjusted for inflation" in your real estate graph. When I first started buying property for rental purpose, gas was .25 per gallon or five for a dollar. Real estate went up between five and 10 percent almost every year until we finally sold our last holdings in 2004 for 1.5m that we originally paid 160K for in 1982. So what is that adjusted for inflation except financial hocus pocus?

  • Report this Comment On September 13, 2013, at 11:38 AM, TMFGortok wrote:

    TMFHousel: Looking at DC home prices, does that mean we can expect prices to fall slightly? It looks like they're 'over' their historical adjusted-for-inflation range: http://www.aboutinflation.com/inflation-adjusted-charts/us-r...

  • Report this Comment On September 13, 2013, at 11:59 AM, flyboy0100 wrote:

    <<In order to apply the principle of mean reversion, one must be able to *establish* the mean to which something will revert. In the case of population growth, it's more helpful to look at the growth rate, rather than the overall total, which fails to control for things like discovering new resources and habitable land.>>

    The point I was (poorly) trying to make is that in investing we are biased toward human lifespan time scales or shorter. Even a tiny population growth rate can't be sustained forever on Earth.

  • Report this Comment On September 13, 2013, at 2:27 PM, sibtrag wrote:

    We don't revert to the mean when there has been a fundamental change to whatever is being measured.

    As TroyCash mentions, communication technology has brought a fundamental change to the labor market, making it easier to distribute work across a large area to reap the lowest labor costs for each role.

    I think that #5 may be affected by a general rise in automobile quality, and thus a longer productive life of each automobile.

  • Report this Comment On September 13, 2013, at 5:35 PM, adamcohen wrote:

    the average is not the mean!

  • Report this Comment On September 13, 2013, at 8:10 PM, Lucaskasan wrote:

    The difference between the housing chart and all the other charts is that the housing mean is pretty obvious. The other means---not so much.

  • Report this Comment On September 15, 2013, at 2:59 PM, mastermiki wrote:

    I really dislike opinions that ignore the fact that the world just 50 years ago was a very different place economically. Ditto for silly 100 year charts going back to the early 1900's... Assumptions based on such "historical data" overlaying the global economy and financial markets today are laughable at best.

  • Report this Comment On September 15, 2013, at 11:22 PM, SkepikI wrote:

    ^ indeed. The reilaibility and veracity of data before 1900 ought to set off alarm bells for anyone who has ever "collected" real data. Just think about the complications of getting accurate numerical information, let alone complete coverage in 1900.

    And, for all you fans of "reversion to the mean" you should contemplate the statistical foundation on which this concept is built. (AS should you Morgan) Mean and average are mathematical constructs built upon the bell shaped curve produced by random variations about a real data point. There are several "notorious" statistical "exceptions" to the well accepted mean and reversion thereto.

    Bottom Line: bust the statistical bell curve presumption and the reversion to the mean is nonsense. No bell shaped curve? No mean, no average, and no reversion to either. Famous random example: population curve of Dinosaurs.

  • Report this Comment On September 15, 2013, at 11:24 PM, SkepikI wrote:

    ^ which I have become by not accepting reversion to the mean as fundamental theory...ha ; -)

  • Report this Comment On September 25, 2013, at 10:36 AM, Ragingmoose wrote:

    Morgan, I keep a folder of your articles. This one is one of the best. We should send it to all the short term oxymorons of Wall Streetm ... and shot down CNBC for ever.

    From Montreal,

    a city as far of New York as Omaha, Nebraska.

Add your comment.

Sponsored Links

Leaked: Apple's Next Smart Device
(Warning, it may shock you)
The secret is out... experts are predicting 458 million of these types of devices will be sold per year. 1 hyper-growth company stands to rake in maximum profit - and it's NOT Apple. Show me Apple's new smart gizmo!

DocumentId: 2633577, ~/Articles/ArticleHandler.aspx, 9/30/2014 12:40:57 PM

Report This Comment

Use this area to report a comment that you believe is in violation of the community guidelines. Our team will review the entry and take any appropriate action.

Sending report...


Advertisement