Scott Sumner  

If you believe in bubbles, then why are you a libertarian?

Is There Objective Truth?... Cash and freedom...

I'm not sure this will convince anyone, but I'll try anyway. As you may have gathered from my previous post, I don't believe in bubbles. In addition, I'm a libertarian. I see those two facts as being related. If asset markets are efficient, then the case for government intervention is weaker (although of course there may be other grounds, such as external costs like pollution.)

Here's what I don't understand. I often talk to libertarians who seem to see bubbles all over the place. But this implies that markets are not pricing assets at the proper level. Often this will be attributed to outside factors such as monetary policy. It would be like excusing inefficiency in soybean pricing by pointing to the fact that it was windy and rainy last week. An efficient market should price assets at the proper level, given the stance of monetary policy, as well as the expected future stance. If asset prices are obviously too high, and likely to fall after the bubble bursts, then markets are not efficient.

If markets are not efficient when monetary policy is off course, there is absolutely no reason to assume that monetary policy would be efficient when monetary policy is not off course. Inefficiency results from irrational pricing, it's either a problem or it isn't. People don't become irrational just because the fed funds rate is set at 2% rather than 3%. Either markets are irrational or they aren't.

If they are irrational, then the case for government intervention is much stronger. Of course government intervention is also prone to errors, so market inefficiency doesn't automatically make intervention desirable, but certainly the case for it is stronger.

I wonder if libertarians who advocate bubble theories understand that they are at least marginally weakening the case for libertarianism?

PS. I very much like jc's comment, after my previous post:

Markets are complex adaptive systems. As Dan Dennett notes, "evolution is smarter than you".

(Always struck me as interesting, btw, how my progressive friends love to belittle conservatives who deny evolution - from single cell organisms to monkeys to humans - when they quite literally hate most modern applications of evolutionary thought, e.g., evolved preferences, human natures, complex adaptive systems that defy centralized mandates, etc.)

Soon I'll try to do a post showing that liberals who claim Chinese exports raise the unemployment rate in America, are actually (implicitly) defending conservative arguments that structural problem and "reallocation", not deficient demand, explain the high unemployment after 2008.

Comments and Sharing


COMMENTS (40 to date)
Mark writes:

Hmm. All of this depends on how one defines a bubble.

A bubble doesn't require irrational pricing; it merely requires inaccurate pricing. The price of anything (or at least anything that isn't produced and consumed instantaneously) is dependent on future supply and demand as well as present, and in so far as those cannot be perfectly anticipated, markets will price some things inaccurately. More over, human consumption patterns are not points drawn from a standard statistical distribution; the parameters of the distribution themselves can change unpredictably, so not merely unpredictable outliers but even significant unpredictable trends can occur, so inaccurate pricing can be persistent.

So, in other words, if the market merely needs to be wrong for a bubble to occur, it is entirely possible; but it doesn't mean the market was irrational (rationality is not sufficient to be right, particularly when dealing with random variables). Even if markets perpetually make the best guess, the best guess is bound to be wrong occasionally; this does not support the case for government intervention in markets, because that case rests on the belief that governments can make better guesses than markets, but a retrospective error by the market doesn't support this case since it does not necessarily mean the market's guess at the time was not the best guess based on available information. Analogously, the market need not (and cannot) win every hand; pointing out that it lost a big hand does not justify the state taking it's seat at the table; rather, one must demonstrate the state's system would result in fewer lost hands than the market's would.

Kevin Erdmann writes:

Not to mention that they are basically demanding intervention through monetary policy. I think they don't think it is interventionist because they see it as pulling us down toward a non inflationary gold standard ideal. But the important thing about our fiat currency isn't the inflation, it's the monopoly on the medium of account. There is no way to impose their anti-bubble remedies without financial repression through a controlled currency and tightly regulated banks.

Scott Sumner writes:

Mark, You said:

"A bubble doesn't require irrational pricing; it merely requires inaccurate pricing."

I strongly disagree, and so does Nobel Laureate Eugene Fama. If bubble doesn't mean irrational pricing, then the term is completely vacuous. Pricing is always wrong, ex post, or at least 99.999% of the time it's wrong.

Lorenzo from Oz writes:

Asset markets are unstable =/= bubbles exist. But it is remarkable how many people cannot parse the difference.

Rises and collapses in asset prices (retrospective "bubbles") rest on unpredictable turning points. For if the turning point was reliably predictable, people would not buy prior to the turning point at prices about to collapse.

Therefore, asset price rises and collapses rest on unpredictable turning points.

Which is to say that asset prices cannot incorporate information which does not yet exist.

So, the issue is simply whether asset prices reflect existing information. If they do, then EMH. If they don't, please nominate the block to them doing so.

Complaining that asset prices do not reflect information that does not yet exist is remarkably pointless.

The upshot of which is unstable asset prices =/= bubbles. Not in any useful sense of "bubble" other than "retrospectively unstable asset prices". Which is not a very useful sense and typically does not mean nearly as much as folk think it does.

As I understand it, a libertarian embraces the non-aggression principle NAP. That is the primary definition.

One logical consequence of the NAP is free trade. Many economists have shown many benefits of free trade. For example, in a model with perfect competition everything comes out perfectly.

This promised perfection tempts many libertarians, who are eager or paid to engage on the margin of public policy debate, to equate the supposed benefits of free trade with their libertarianism. One of these supposed benefits is efficient markets.

But this equation should be spelled out as an assumption, should we want to keep our heads clear for deeper dives into theory.

Silas Barta writes:
If they are irrational, then the case for government intervention is much stronger. Of course government intervention is also prone to errors, so market inefficiency doesn't automatically make intervention desirable, but certainly the case for it is stronger.

How so? You can believe that market actors can make poor decisions while believing government actors make worse decisions.

Imagine someone saying, "Beanie baby manias strengthen the case for price controls."

Robert Evans writes:

I'm not an economist.

I'm curious about two things touched on in this post:

1) What is this valuing of "efficiency" about, and what is the end output of monetary efficiency supposed to be?

2) Most of what I read on economics (especially of the market) deals in currency and things called markets (which are generally currency or barter driven). Are economists in general paying attention to non-currency economics? Because the current market system seems kind of crappy at advancing total material well-being, at least when I look at how badly the status quo utilizes the skills, and fails to facilitate the ideas, of so many people.

Julien Couvreur writes:

Although I worry about the ambiguous meaning of "efficient" in your question, I think you raise a valid point.

"Why Don't Entrepreneurs Outsmart the Business Cycle?" offers a few factors that resolve this apparent paradox. Read it at

In short, if you override the signals by force, the information about what the natural signals would be disappears and can only be imperfectly guessed.
Also, the overriding of the natural signals creates an incentive to go along to some extent, even if you think it is a bubble. Those that short too soon loose.

Andrew_FL writes:

On the contrary, if strong versions of EMH are correct, it rescues Market Socialism from its deserved place in the trashbin of history.

John Hawkins writes:

Andrew_FL is onto something with his comment.

Scott, this is Chicago libertarianism over-modeled garbage. Free markets are like Churchhills quote about Democracy. The worst except everything else tried.

Levi Russell writes:

Andrew_FL is right and Kevin Erdmann is way off base.

Scott, the problem is that your analysis, like most of what Coase called "blackboard economics" lacks nuance. Just because we don't believe markets live up to some impossible platonic ideal doesn't mean we can't view it as the best among alternative real world systems (see Demsetz, 1969).

"Inefficiency results from irrational pricing, it's either a problem or it isn't. People don't become irrational just because the fed funds rate is set at 2% rather than 3%. Either markets are irrational or they aren't."

That's way too simplistic. Information, after all, is costly to obtain. The implication is that no one will ever be perfectly informed on anything, even people who spend all day looking at one stock won't know everything about the distribution of future prices.

Charlie writes:

Here's a Rorty quote I found on wikipedia:

"Truth cannot be out there—cannot exist independently of the human mind—because sentences cannot so exist, or be out there. The world is out there, but descriptions of the world are not. Only descriptions of the world can be true or false. The world on its own unaided by the describing activities of humans cannot.”(5)

I'd be very interested in a big debate about the objectiveness of truth. My guess is that Scott has thought a lot harder about his position than DH, but that it is also a harder (less intuitive) position to defend.

David N writes:

I believe the market is efficient, except in certain cases when it isn't. "Either markets are irrational or they aren't" is a false dichotomy. Why can't markets be mostly rational with occasional pockets of irrationality?

The core of the EMH, at least according to Malkiel, is that the actions of individuals trying to take advantage of inefficiencies in the market drive its efficient outcome. If everyone just accepted the EMH and never questioned the market price, there would be no forces, no movement of capital, to produce an efficient outcome.

So, are all the actors in the market, acting against EMH to produce EMH, therefore irrational? Is the net product of millions of irrational actors, by definition rational?

Two people want to buy the common stock of a corporation. One wants a nice dividend, the other wants to take control of the corporation and get seats on the board. Must they agree on a common, "efficient" price?

I'm a libertarian too, and I see nothing wrong with admitting that markets are sometimes irrational. As a libertarian I believe that individual liberty is the best way to minimize irrational market outcomes. Government is clearly more likely to base decisions on non-economic factors.

tl;dr, the weak form of EMH usually holds but not always, and believing in observable instances of market inefficiency does not undermine libertarianism.

Nick Wernicke writes:

For those in the Von Mises orbit (many non-monetarists), the word 'bubble' has come to imply "any situation where prices would be lower absent government interference". Yes, I know it's a bit of a cop-out to redefine the term, but that's really the source of the 'bubble' in bubble-talk over the last 10 years.

People at FEE or Mises Institute or ZeroHedge would understand you if you talked about hypothetical "barber bubble" caused by some insane occupational licensing requirements. Of course a haircut in San Diego IS $30... and its not because hipsters here are cutting their hair more.

The differences between libertarian pundits has to do with how they personally perceive the impacts of central planning intervention.

Some (like Peter Schiff) attribute a very strong affect to the economy by CB interest rates and open market actions. They directly attribute large scale market moves to central bank policy. For example: "Greenspan set interest rates too low, and this triggered a rush into housing, refinancing, vehicle loans, and caused the housing bubble."

Some (like John Tamny) attribute only a weak link between the economy and CB actions. They look for more subtle reasons. For example: "The 2001 market drop and subsequent low rates caused old and institutional money to re-allocate to housing because it is a safe bet with a good dividend. A perceived central bank safety net caused many investment banks to assume more risk than they otherwise would have."

K Partyka writes:
I often talk to libertarians who seem to see bubbles all over the place.
I do not agree on the main statement. Bubbles do not have to be contradictory to the efficient markets. They can just express the overall over optimism about the price of risk in the economy or an agreement on the "status" of an asset( like treating an asset as information insensitive). Such over optimism can dissipate quickly and cause bursting of a Bubble. The fast reaction of the market is rather an indication of the efficient market( fast change in utilities, demand etc is causing fast adjustment in prices).

I do agree however on the intervention part of the statement. If you believe that monetary policy causes irrational pricing, then you are not a believer in efficient market.

But if there is an external effect of the monetary policy on the price of risk(optimism) ( and there probably is, but small) then monetary policy could cause bubbles, but the market is still efficient,as it is pricing the changes in the attitudes on the market.

Dan W. writes:

Many good posts here with which I agree.

EMH means that at any point in time the pricing of assets in the market are internally consistent. If this were not so then one could immediately profit from the inconsistency.

EMH says nothing about the future of prices and yet assets are priced based on expectations of the future. Bubbles are an outcome of faulty expectations. They are a manifestation of prices being based on human emotion and the collective dreams of what might be.

Can the crowd be wrong? Of course! Can one profit by betting against the crowd? Yes but clearly it is not easy to do. The crowd can be wrong for a very long time. For example, betting against the Nasdaq crowd between Nov 1999 and April 2000 would have felt like a long, long time. The Nasdaq 5000 crowd was wrong but it took many months before this was made evident.

Jose Romeu Robazzi writes:

Markets are processes, things don't happen instantaneously. This is where the "invisible hand" idea is actually misleading. In order for prices to move, there must be an informed trader willing to take on risk on that view. It does not matter if someone "knows" a price better, but cannot act on it. Over time, as information spreads through the system, prices move. Kudos to Hayek.

BJG writes:

Or perhaps libertarianism has nothing to do with a dogmatic faith in the perfect efficiency of markets.

RPLong writes:

What? Doesn't this argue the exact opposite of what Prof. Sumner wants to argue?

The bubble theories state that economic calculation can't happen accurately (i.e. markets cannot be efficient) when the state interferes with price signals. In other words, markets are efficient unless something (say, the central bank) disrupts their inherent efficiency by undermining pure market pricing information.

So, we can't say, "You're saying markets aren't efficient!" as a response to that argument, because that is the exact opposite of what bubble theories claim.

On the other hand, if Prof. Sumner wants to claim that monetary policy does not influence the efficiency of markets, then why does he suppose monetary policy is effective at all?

Joe Dokes writes:

[Comment removed pending confirmation of email address. Email the to request restoring this comment. A valid email address is required to post comments on EconLog and EconTalk.--Econlib Ed.]

Brad W writes:

I've never believed in Strong EMH, as a libertarian. I think the long run, EMH is accurate. Prices tend to generally reflect the "correct" price via the collective actions of many people reasonably informed about market information.

But the short term is always different, and the market can remain inefficient longer than you can remain solvent... Every price is someone betting on future asset values. And we're human. We sometimes collectively bet on expectations that aren't accurate. The people who are right that the rise in asset prices is not justified are a smaller group than the people who are wrong. So prices rise unsustainably. Eventually those numbers flip, and they fall.

Bubbles are an excursion from complete efficiency, and the popping of those bubbles is the excursion being corrected.

Jeff writes:


You said: "If markets are not efficient when monetary policy is off course, there is absolutely no reason to assume that monetary policy would be efficient when monetary policy is not off course."

I think you meant to say " reason to assume that markets would be efficient..."

Daublin writes:

To be a libertarian, all you have to believe is that the most likely government responses to a problem are also likely to make things worse.

It depends on how you define "bubble", but it seems like people really do have buyer's (or seller's) regret. If they regret buying something at a high price last year, based on what they have learned this year, and if it happens pervasively, it seems like that would be a bubble.

Going with that definition, it is hard to see how the federal government is going to be helpful in fighting bubbles. By definition, the most knowledgable people in a given sector are all deluded and are going to change their mind before long about what the appropriate price levels are. The people making decisions for the government can't be *smarter* than the smartest people in the given sector, can they?

As such, the best ways the government can help are more about setting the rules of the game, than attempting to identify and deflate bubbles. That's a long-winded way of saying, the government should define reasonable markets and then step away while they do their job.

Hazel Meade writes:

Just because markets are irrational, doesn't mean that governments are less so.

I'm totally prepared to acknowledge that people and markets are often highly irrational. The problem is that I don't believe that governments are going to do any better. A bubble, for instance, is nearly by definition, impossible to identify before it pops. That is because bubbles are psychological phenomenon driven by the belief that prices WON'T fall. That belief, especially in a democracy, in invariably shared by the government agents that presumably would be put in change of bubble-prevention. Imagine the Fed attempting to "pop" something like the tech stock bubble in the 1999s. If people believed that the bubble would continue, they would not stand for it. And if they didn't believe that, it wouldn't be a bubble. End of story.

Secondly, governments operate by means of force, and always entail a small number of people who have their own interests forcing their will on a large number of people. That is going to have a greater distortive effect on markets than the irrationality of millions of independent people. When the small number of people are wrong, the effects are more widespread, more disastrous, and more immoral because they were forced upon the victims involuntarily.

Hazel Meade writes:

Government is clearly more likely to base decisions on non-economic factors.

This sums it up in a nutshell.

Robert Schadler writes:

This is the sort of discussion that gives "libertarianism" a bad name.
Prices, markets, efficiency, rational, government.
Prices are almost always "wrong" but almost always less "wrong" than other ways of pricing something.
Consider: a painting is being auctioned. No one is interested in it. Technically, in that immediate "market" it is worthless. One person comes in who likes it. Its market value rises somewhat. A second person comes in determined to buy it. The price rises dramatically. That person then ... has a heart attack ... or thinks it is a fake. The prospective price drops.
At any moment, the price was either "right" or pretty accurate, given the proximity and number of potential buyers and the partial information they had, and the subjective value of the money they possessed relative to owning that object.
Conceivably the government could have intervened: forbidding the sale, setting a cap or a floor for the price, disqualifying certain potential buyers, etc. OR the owner could do any of these things. And the price would fluctuate accordingly.
Was there a "bubble" or did the price simply change based on changing, often imperfectly perceived, circumstances?

Phil writes:

I fail to see the link between irrational prices and libertarianism, unless being a libertarian means you must believe in efficient markets. If so, that's news to me and I might need to rethink labeling myself a libertarian.

Eli writes:

If the stance on monetary policy is unclear or leaves a lot of variability within, then markets might do the best that can be done, but still have bubbles.


AMW writes:

Funny I should read this post now, as I am about to conduct a laboratory experiment on price bubbles. I expect to find them in this session, as I did in the last.

Have you read the literature on bubbles in experimental laboratories? If so, what is your take on it?

Thomas B writes:

Responding to this post is like shooting fish in a barrel. Not that I've ever shot fish in a barrel - frankly I'm puzzled by how that ever became a thing.

I'm a libertarian: I advocate that government should adopt a strong, but rebuttable presumption in favor of liberty in its policymaking.

I can believe in bubbles: experimental economists have demonstrated their existence under laboratory conditions. There is NO doubt that bubbles exist.

Still, I can be a libertarian, because I can believe that no mechanism exists whereby governments can do better than markets in setting prices in the real world (unlike artificial experiments where the correct price can be known): even accepting that markets may misprice, there is no reason to believe that government policy could do better, and a good deal of reason to believe it would be systematically worse.

Have I answered the question?

Robert Schadler writes:

Seems a "bubble" is defined by a sharp increase in the price of something followed, after a period, by a sharp decline in price. And those some things need to be of sufficient number and interest to enough people so that the price drop is broadly noticed.
If that is what a "bubble" is, then it is not hard to believe that they exist.
No doubt some few are caused by govt action: for example, the price of slaves in N. America probably dropped precipitously during the course of the U.S. War of Southern Secession. But most due to change in fashion, the "madness of crowds," new information, etc.

Alexis Wadsley writes:

I'm interested in the statement 'not believe in bubbles'. This may be semantic, but the idea of Bubbles has a well studied history, Tulipmania, South Sea Bubble, US Stockmarket heading up to the 1929 crash. They are an observed phenomena that is collectively identified as 'bubbes' to allow analysis.

Whether these phenomena (which i will call bubbles) require markets to be rational or irrational, individuals to be rational or irrational is a fair question of debate. Given the historical pattern of bubble endings coinciding with a negative macroeconomic shock, the question of whether there is an appropriate policy framework before or after is also raised. The policy question leads to a requirement for a definition of bubbles that helps identify and justify intervention.

Libertarians shouldn't deny bubbles exist, just have a very high threshold for intervention and challenge definitions that implicitly justify intervention by their construction.

David N writes:

The sharp increase/sharp decline definition of a bubble is insufficient. Those events can occur for all kinds of reasons. My argument, which I trot out in the comments every time Scott brings up bubbles, is that the hallmark of an asset bubble is when existing methods of valuing that asset no longer explain the price, and some new method or "paradigm shift" must be introduced to justify the new, higher prices. Mis-using EBITDA, for example.

Chris Taylor writes:

Pricing is never wrong or right or inaccurate or irrational. In economics, prices are what they are, no more, no less. A bubble has nothing to do with "wrong" pricing. A bubble becomes "evident" only in retrospect when the assumptions underlying pricing (usually multiples, sometimes the cash flow projections, or a mix of both) change sharply. The "proof" of a bubble only happens in the subsequent crash. If the "inflated" multiples don't reverse sharply, by definition there was no bubble.

Chris Taylor writes:

For a given cash flow, its value is determined by dividing it into the riskfree rate plus the risk premium less the expected growth rate. All three variables are highly influenced by monetary policy. And any one, or more likely all 3 at the same time, can swing sharply if monetary policy - or even the perception thereof - changes dramatically. Thus one can say a certain monetary policy is "irrational", or that the market's perception thereof is "irrational", while still believing the market itself is perfectly rational or efficient. It is very important to make this distinction. Prices are never wrong or irrational. The assumptions underlying them are almost always wrong (since no one can predict the future), regardless of whether the reasons for these assumptions appear to be rational or logical or whatever. No competent economist should ever be caught saying prices themselves are wrong!

Hazel Meade writes:

A bubble becomes "evident" only in retrospect when the assumptions underlying pricing (usually multiples, sometimes the cash flow projections, or a mix of both) change sharply. The "proof" of a bubble only happens in the subsequent crash. If the "inflated" multiples don't reverse sharply, by definition there was no bubble.

Good description. Bubbles are only identified when we find out that we were wrong about the assumptions underlying pricing. That causes the bubble to pop. But by necessity, we never know we are wrong at the time.

In order to have a policy intervention to prevent bubbles, government would have to know which assumptions are wrong at a time when the vast majority of investors believe them to be right. And it would have to be able to make such predictions reliably enough, so as not to counteract price changes based on correct assumptions.

David N writes:

Bubbles are only identified when we find out that we were wrong about the assumptions underlying pricing.

This is a cop out. If bubbles exist, and have observable causes, then we, or at least some of us, should be able to identify them as they are happening. I'm trying to think of another useful economic concept that can only be applied in hindsight and struggling to do so. If bubbles can only be identified after the fact then I agree with Scott; they're a useless concept.

Some of the discussion here sounds to me like a quest for objectively-true bubbles. Such discussion may be an instance of the more general search for objective truth.

But if we drop the assumption that we can attain perfect, objective knowledge, and rather see that people make up words in their effort to communicate about things that separate people believe they see, then we see that "bubble" is a useful and apt metaphor for something that many people will swear they have seen.

Hazel Meade writes:

If bubbles exist, and have observable causes, then we, or at least some of us, should be able to identify them as they are happening.

Some of us may be able to, but it could never be common or accepted knowledge, because then investors would act on it and pop the bubble. Bubbles only exist when there is a knowledge gap between reality and generally accepted belief. That doesn't preclude a small minority from identifying the bubble.

I suppose in theory someone could come up with a way of identifying bubbles easily, but then they wouldn't require government intervention. They could just publicize the bubble-detection formula and let investors follow it.

Glen writes:

Scott, why not just say what you mean and leave the ideologies aside: Any market that can be manipulated can't be an efficient (or “correctly priced”) market — even if the manipulator is the market operator itself (e.g., the government).

And why can't bubbles occur in efficient markets? Aren't expectations a major component of asset prices? If so, how much influence could they exert before markets became “irrational?”

Finally, why should less than perfectly efficient markets be a reason for greater government interference? Why would government bureaucrats do a better job of setting prices than individual market participants?

Maybe efficient market theories are just an excuse to not work hard to try to “outsmart” markets? After all, why try when you can't win?

Comments for this entry have been closed
Return to top