Sunday, December 16, 2012

Interest Rates and Jobs

Krugman and Kensians in general believe that if interest rates are artificially held down by the central bank printing money and buying bonds that more jobs will be created.   They argue that low interest rates drive up asset prices, which makes people feel rich, so they spend more.   It also causes inflation which can make real wages go down, so employers are earning more profits and can afford to hire more workers.  Also, low interest rates make it easier to build a factory and start a business that creates more jobs. 

There is another big effect at work though.   When a company has to choose between a robot and an employee they compare the monthly payment on a loan for the robot to the monthly salary of the employee.   For the employee they also look at all the extra government imposed costs like social security, obamacare, etc.  If the robot is cheaper they will choose the robot.

Artificially lowering interest rates makes the payments on the loan for the robot smaller.  Lower interest rates make it easier to get a loan to buy the robot.   This artificially tilts the market in favor of robots.   It means replacing people faster than the natural rate.    Replacing people faster puts a downward pressure on human salaries.    So lower interest rates are very hard on the average guy.

Lower interest rates help whoever can get the lower interest rates.   This will tend to be the owners of existing capital, not the average guy.   So lower interest rates skew things so the rich get a bigger slice of the pie.   This can be seen in the history of the last 30 years.

The Fed has now said they will keep interest rates down until unemployment gets to 6.5%.   This is just like,  “the beatings will continue until morale improves”.

Tuesday, November 6, 2012

Gradually and then Suddenly

"How did you go bankrupt?" Bill asked.
"Two ways," Mike said. "Gradually and then suddenly".
-Ernest Hemingway, The sun also rises (1926)
 When a government goes bankrupt it is after long period where finances gradually get worse and worse and then suddenly nobody wants to loan it any more money.   If it does not print money, then it will have to cut back the size of government to match taxes without borrowing, possibly after defaulting.   If it prints money, then the gradual part can go on a bit longer, and the suddenly part is hyperinflation.   This second scenario is much harder on the citizens.

Friday, October 19, 2012

Hyperinflation FAQ

This FAQ is for the frequently asked questions or objections about hyperinflation.  The statements or questions below are in bold and my responses follow.

How is hyperinflation defined?
In Cagen's 1956 paper on hyperinflation he used inflation of 50% per month as the cutoff for his study.  In a footnote he says,  "The definition is purely arbitrary but serves the purposes of this study satisfactorily.".  Since then many academic researches have taken this as the official definition of hyperinflation and ignored any cases of less than 50% per month as "not true hyperinflation".   Hyperinflation researcher Steve Hanke wrote a letter to the editor to complain when a company used 40% per year and a standards board used 100% in 3 years.      In the real world today people call 100% inflation in 3 years hyperinflation.    I don't think Cagen even meant to exclude 10% or 40% per month as "not true hyperinflation", I think it was just his way of reducing the cases to investigate down to a manageable number.  When millions of people call what they are living through "hyperinflation", I think it is wrong for academics to ignore them  because one guy used a higher cutoff in a paper long ago.

Normal  money functions as both a store of value and a medium of exchange.   A fiat currency experiencing hyperinflation is clearly not a good store of value but is still functioning as a medium of exchange.   The currency is sort of half money and half failed.  Kind of a one foot in the grave situation for money.  Much of the time such a currency dies within a few years.  That is where the name for this blog, "How Fiat Dies", comes from.  I am adding my own definition: Hyperinflation is that transition period when a paper money is clearly failing as a store of value but has not yet died as a medium of exchange.

In the International Accounting Standard of IAS 29 one of the guidelines for when hyperinflation accounting should be used is when "the cumulative inflation rate over three years approaches, or exceeds, 100%.".   They provide a number of guidelines but do "not establish an absolute rate" and "concludes that it is a matter of judgement".   As with Cagen's footnote, it is clear that the precise cutoff is arbitrary, but they are clearly far lower than what Cagen used.   If you have 26% per year for 3 years you would get 100% with compounding.  I contend that their other guidelines would be met if this one was.

There are many other definitions for hyperinflation but they are mostly something like "inflation over X per year" or "inflation over Y per month". People pick some level of price inflation as the cutoff between regular inflation and hyperinflation.  It is just the values for X or Y that differ. 

I think 26% inflation counts as "clearly failing as a store of value" and makes a good cutoff between regular inflation and hyperinflation.    This casts a wider net than most definitions, but I think it is wrong to exclude the cases between 26% per year and 50% per month.  It is the same phenomenon at 26% per year as it is at 50% per month.    When the public is faced with 50% per year you can bet they will call it hyperinflation, no matter what any academic says.   A country that is used to 2% inflation and suddenly facing 26% inflation will be in shock.  Excluding these is wrong because, after reaching 26% per year inflation, I think the odds  are they will have 100% total inflation over 3 years.  It is wrong because a few years of this wipes out life insurance, retirement funds, fixed pensions, and a life savings.   It is wrong because it gives the impression that 26% per year is nothing to worry about and 50% per month hardly ever happens.   If we use 26% per year, then hyperinflation is a common problem that needs to be better understood.   I really think it is a common problem that deserves more attention.

Note that hyperinflation is not defined in terms of the money supply alone, since the velocity of money and GNP are also key factors in the price level during hyperinflation.  

Hyperinflation is a process,  a positive feedback loop, that once entered is very hard to get out of.  This process can go on for years.  In the feedback loop the more the central bank prints money and buys bonds the less other people want to hold bonds.  At the same time, the less other people hold bonds, the more the central bank has to buy them so that the government has enough money to spend.  This feedback loop can also be called a death spiral, chain reaction, tipping point,  like an avalanche, slippery slope, or a debt bomb.  Once the conditions are right, it can just go off.

Definitions are important.  I hope this makes my position clear. 

Is there a real chance the US dollar could get hyperinflation?
Bernholz did a study of 29 cases of hyperinflation and looked at the circumstances that led up to them.  He found that the best predictor of hyperinflation is government debt over 80% of GNP and the deficit over 40% of government spending, in a country that prints its own money.  Note that 50% deficit spending would mean spending twice what was collected in taxes.  The US is over the debt number and not far from the deficit number,  so the danger of hyperinflation is real. 

The government or central bank would never decide to have hyperinflation.
In over 100 cases of hyperinflation I don't believe there has ever been a single "decision to have hyperinflation". Hyperinflation is when things get out of control. It is not something central banks or government voted on. No group in any government or central bank has had a show of hands like, "all in favor of hyperinflation raise your hands". Not the way hyperinflation happens.  Hyperinflation is a market response sometime after government debt over 80% of GNP and deficit over 40% of spending.  The central bank starts printing money and buying up government debt to help lower interest rates.  But the more new money they make to buy bonds the less people want to hold bonds.   The less people want to hold bonds the more money the central bank has to make to try to control interest rates, and so that the government can keep in operation.  You get a feedback loop that goes out of control.  Nobody votes for hyperinflation. Nobody wants it. It just happens.

Why would the Fed suddenly print trillions of dollars?
There can be a panic to get out of bonds.  There may be $3 trillion in bonds coming due in the next 12 months that might not get rolled over if people started not trusting bonds (I am trying to nail down how much short term the public holds, separate from the Fed and Social Security).   There is also the excess reserves of around $2 trillion that are earning interest, much like government bonds really.   The current deficit of around $1 trillion would also have to be financed with new money if nobody but the Fed was buying bonds.    Altogether this would be something like $6 trillion over 12 months.

Hyperinflation would never happen in America!  
There were cases of hyperinflation in America's colonial period.   There was hyperinflation during the Revolutionary War.  Remember, "not worth a Continental"?  There was hyperinflation in the South during the Civil War.     I also think that if the US had not made it illegal to own gold in 1933 that the Fed would have gone bankrupt, because they did not really have enough gold to back all of the notes they had issued, and that paper money would have become worthless then too.   Hyperinflation is more common than most people realize.  The time periods from the Revolutionary War hyperinflation, to the Civil War hyperinflation, to the 1930s currency crisis, to now, are all similar.   To me this looks like some major currency crisis cycle is about due.

What is the math for hyperinflation?
The math for hyperinflation  starts with the equation of exchange and with one transformation you get:

P = M * V / Q

Where the variables are:
P = Price level
M = Money supply
V = Velocity of money, how many times money turns over in a year
Q = Real GNP

In hyperinflation the money supply is going up, the velocity of money is going up, and the real GNP is going down, all at the same time.  It is a tripple wammy that drives prices up really fast.

You have the cause and effect backward!
People will argue things like, "it is not the monetization that causes the loss of confidence, it is the loss of confidence that causes the monetization".  While someone else argues the other way.   I have also seen both sides of, "hyperinflation causes money printing, not money printing causes hyperinflation".

Hyperinflation is really a positive feedback loop.  It is a circle, a death spiral.  To argue about what comes first in hyperinflation is a bit like arguing which came first, the chicken or the egg. 

Why can't you stop hyperinflation by just having the central bank stop monetizing debt?
It seems obvious that if you just have the central bank stop buying government debt the hyperinflation would stop. The problem is that the government needs money to operate and is spending far more than what it gets in taxes and has debt around the size of the GNP.  The deficit is so large that cutting government or increasing taxes enough is not possible.  The only way the government can keep in operation, when other people stop buying their bonds or even rolling over their bonds, is if the central bank steps in. So the government always makes sure the central bank steps in.   This may take changing laws or replacing people at the central bank, or just ignoring laws, but the government will get the money or it will fail.  If it can not pay its employees then nobody will show up for work and the government will collapse.
How could the Fed on its own cause hyperinflation?
It is always the combination of the government spending far more than it gets in taxes and the central bank printing money and buying government debt (monetizing debt).   It is the two together that result in hyperinflation, not one alone.

Isn't hyperinflation a political event and not a monetary event?
The root cause of hyperinflation is politicians spending far more than they get in taxes.   It is correct to say that, "hyperinflation is always and everywhere a fiscal phenomenon".  In this sense hyperinflation is a political problem.  However, I think Milton Friedman was right when he said "inflation is always and everywhere a monetary phenomenon".  First off, hyperinflation is a process that can go on for years, not an event.   The exact cutoff point between regular inflation and hyperinflation is arbitrary, so it makes no sense to say that below 26% it is a monetary phenomenon and above 26% it is a political phenomenon. 

I think the idea that hyperinflation, where the monetary unit rapidly becomes worth less and less, is not a monetary phenomenon is not only wrong but dangerous.  This wrong idea keeps people from understanding what is really going on.  For example, Modern Monetary Theory (MMT) as it stands does not explain hyperinflation.  They count bonds as money and monetization as "just an asset swap".  They do not see any tipping point, positive feedback loop, sudden risk, or danger of inflation getting out of control.   They think that you could always just make a bit less money if inflation got too high.  Instead of admitting the problem with their theory (lots of experimental evidence of inflation getting out of control) they just say hyperinflation is a political event not a monetary event.      By the way, with a small correction to MMT, making what I call CMMT, I can easily explain hyperinflation

The US is very different from Weimar Germany or Zimbabwe!
Each case of hyperinflation is unique, so if you are looking for differences you will always find them.   You need to understand the common characteristics.  Hyperinflation happens because government debt gets over 80% of GNP and deficit gets over 40% of spending.   It does not matter how you get into that situation.  Hyperinflation works the same if you lose a foreign war,  a civil war, a dictator goes crazy, a government with excessive foreign debt, nationalizing too many businesses, rampant corruption, productive collapse, excessive regulation, a regime change, too many taxpayers fleeing high taxes, a massive depression, or whatever.  It just matters that the government is spending nearly twice what they get in taxes and has already borrowed more than is reasonable.   When they are in this situation they can not borrow more, except from the central bank under their control.  So they get the central bank to make money and "loan" it to them.  When the reality is the only way they can pay back that "loan" from the central bank is by first getting another "loan" from the central bank you are probably headed for hyperinflation.

Another problem is that people often compare the US before hyperinflation to some country during hyperinflation, which is not a fair comparison.   For example, after prices are shooting up and interest rates go up no banks will be making 30 year loans.   So people will say the fact that the US is making loans shows that it is different than some country with hyperinflation.  This is silly.  Of course a country that does not yet have hyperinflation is different from a county in the midst of hyperinflation.   The real trick is recognizing the circumstances that lead to hyperinflation.

When a country gets hyperinflation there are a number of stages it goes through.   Things are very different as hyperinflation progresses.  

Other countries are even worse off than the US.
Other countries have all sorts of problems too, so the US dollar may not drop against those other currencies.   Hyperinflation is not really about exchange rates.  If the Pound, Yen,  Euro, and Dollar were all getting 26% inflation the exchange rates could stay the same but we would still have hyperinflation.

There has never been hyperinflation of the world reserve currency before!
I have two contradictory answers for this one.  The first is that it is only since 1971 that the world has had a fiat currency as a world reserve currency.  In the past it was always gold or silver or a currency convertible to gold or silver.   It is not possible to get hyperinflation when using gold and silver.  Hyperinflation happens to fiat currencies.

The second answer is that the world reserve currency in Roman times had hyperinflation as they put less and less silver in the coins.  Please use whichever answer you like and ignore the other.  :-)

A helpful reader on reddit  said, "As soon as the Romans declared that the value of their coin was divorced from the market value of the silver those coins became fiat currency. So your first explanation holds for the roman situation too."

Why would the US central bank willingly destroy their currency?
If the Federal Reserve Notes become worthless then the Federal Reserve is probably done for, so why would they ever do such a thing?    Sounds logical.   But why did  that same logic not protect all  the other central banks from making hyperinflation?  The government writes the laws, appoints the people to the central bank, and controls the guns.  When the government is desperate for money from the central bank they get it.   But also remember that the central bank is created by the government.  If the government collapses  because  it does not have enough money to pay for things, the central bank will probably go down as well.  The survival of the central bank does depend on the survival of the government.   I think this is the core of why central bankers risk their currency to support their governments.

Don't we need more stimulus now and worry about the debt later?
The standard definition of GNP includes government spending.  So more government spending can increase the GNP number in the short term.  But it is not really helping and hurts in the long run.   In general anything that makes your government bigger hurts in the long run.   The government is an overhead on the productive parts of society.  The more the overhead the slower the long-term growth rate

How could the US dollar get hyperinflation as long as international commodities are priced in dollars?
It may be that as inflation picks up the international commodities will no longer be priced in dollars.  Perhaps they will use Yuan or maybe they will use gold.  But it does seem doubtful that people would keep selling things in dollars if dollars are failing fast.  As fewer and fewer things are priced in dollars there will be less and less demand for them.  So the end of the dollar dominance will contribute to the hyperinflationary pressure.

If you can't predict the exact timing then you are no help to investors!
Hyperinflation is sort of a slow motion panic where people get out of bonds and then out of the currency altogether.  But as a human panic, it depends on humans, so hyperinflation is hard to predict with precise timing.  However, if bonds are paying 1% and going to suddenly lose more than half their value sometime in the next 10 years, then getting out years early is clearly a better plan than waiting.

But our government is just borrowing, not printing and spending!
People think their own government is just borrowing money and that in cases like Weimar Germany or Zimbabwe they were just printing and spending money.   However, the truth is that Germany and Zimbabwe had central banks that were buying government bonds just like the US central bank is doing.    If the central bank loans some money into existence (which is what buying a bond is really doing) that money should go away when the loan is paid back.   However, in Germany and Zimbabwe the only way the government could pay back a bond was by first selling another one.    So effectively they were just printing and spending money.   If in your mind you put a big black box around the central bank and the government and look at what comes out of the box, what you see is new money being spent.   The pretend loans that never really get paid back are all on books inside the black box.  It makes no difference to the world outside the black box if they were just printing and spending or have pretend loans inside the black box.   The US is now in the same situation.  The only way they pay back existing bonds is by first selling new bonds.    Because the historical narrative simplifies out the central bank, people don't realize that their own government is doing the same thing that previous hyperinflation governments were doing.    People think previous cases were just stupidly printing and spending and that their government is responsibly borrowing money, when really their government is doing the same thing.  So the same mistake is repeated again and again, and hyperinflation keeps happening.

We have been at these debt levels after WWII and there was no hyperinflation.
Some think there is no reason to worry since after WWII the debt was huge but the USA did not get hyperinflation.   After the war ended and the military was cut back there was no deficit.  If there is no deficit there is no reason to worry about hyperinflation.   Today there is a huge deficit and it does not seem politically possible to get rid of it.  So this is not the same situation and is far more dangerous.

Deflation is as big a danger as inflation.
 The graph below shows that when the US was on a gold or silver standard the probabilities for price inflation and price deflation were similar.  However, after 1933 when the dollar partially moved away from gold there was little price  deflation.  After 1971, when the dollar became completely unbacked, there has been no more price deflation.   It is wrong to use statistics from a time with gold money to predict dangers under a pure fiat money.  There have been 56 cases of hyperinflation with more than 50% per month inflation (12,970% per year).  There has not been one single case of double digit price deflation in a pure fiat currency (with no link to gold or silver nor returning to either). In pure fiat money the probability of price inflation is far higher than that of price deflation.

Most of the people predicting deflation are not even predicting price deflation, just a reduction in the money supply plus credit.   Hyperinflation destroys retirement funds, destroys life insurance values, and makes salaries not keep up with the cost of food and energy.   A reduction in the amount of credit that does not even impact prices is a non-event compared to hyperinflation.

Doesn't the deleveraging of huge private debt prevent hyperinflation?   Can you have deflation and hyperinflation at the same time?

If you define deflation in terms of the money supply, and you count bonds as part of the money supply, then you predict deflation when people are deleveraging.   Many people seem to see deflation coming, by this definition, and think that means you can not get hyperinflation.  This is not correct.   Hyperinflation is defined in terms of "price inflation" not "money supply including credit".  At the start of all hyperinflations bond values crash, so this definition of deflation will be met even as prices are shooting up.  So it is common to have deflation and hyperinflation at the same time, initially.  If inflation is 5% per month then a bond that pays 2% per year for 10 years is worth less than a penny on the dollar.  The bond collapse happens fast and the hyperinflation can go on for awhile.  History books will later say it was hyperinflation, not deflation. 

Bernanke says he is just temporarily monetizing debt.
You should not believe him.   He has to buy more and more to keep interest rates low enough that the government can handle the huge debt.  If he tried to sell the bonds he has the interest rates would shoot up and the government would not be able to afford the debt.

Can you simulate hyperinflation?
Yes, I have a nice web based simulation of hyperinflation.   The feedback loops of hyperinflation are a natural for simulation.   You can play with the inputs and see which ones end up making hyperinflation.  You can also clone the model and make any kinds of changes you want.  Please credit me if you do copy my work and publish it someplace.  People can easily use Google to see my very similar hyperinflation simulation, so it is probably not possible to steal it and get away with it.  :-)

If there is hyperinflation what makes you think you would be able to buy food with gold or silver? 
People will say, "you can't eat gold, what good is it?".   But commerce still goes on during hyperinflation.  In historical cases of hyperinflation people move to barter or selling in more stable currencies from another country, or gold and silver.  They do this even if it is illegal or "black market".  They do it even when people are getting their heads chopped off for selling things for gold instead of paper money.  To understand why this happens imagine two different store keepers, one that sells for gold and one that is selling for local hyperinflating paper money.  The one who has gold can go to his supplier a week later and his gold is worth enough to buy a new set of merchandise and stay in business.   The one who sold for paper money may find that his money is worth half as much next week and he can not buy a full set of merchandise, so he soon goes out of business.     If only those that ignore government rules to use paper money can stay in business, the fraction of the economy that is "black market" grows as hyperinflation gets worse.

Hyperinflation is for failed states and the US is not a failed state.
Not all hyperinflations are in failed states.  The US got hyperinflation as they won their independence.  However, there is certainly a high correlation between failed states and hyperinflation.   For a state where half of its spending power comes from taxes and half comes from printing money, it should not be surprising that as the currency fails the state fails.    Since paper money is issued by a state it is not surprising that the paper money is worthless after that state fails.  The question is, is it the failure of the state that causes the currency to fail or the failure of the currency that causes the state to fail?   The truth is that they often go down together.   Before they have failed they can both seem fine. 

They have printed trillions without trouble so far, why would more be different?
It may help to read the storey about the straw that broke the camel's back again.  There are situations where a small amount of something is ok but lots is not.  There are lots of medicines that are helpful in small doses but deadly in large doses.  If 2 people go walking on some ice they might be fine but 50 people walking in the same area might break through.  Hyperinflation has a human element.  There is sort of a panic to get out of the bonds and then a panic to get out of the currency.  It is hard to say exactly when the panic will start, but I don't think anyone ever prints more than the GNP in one year without getting hyperinflation and certainly not 3 times the GNP.

Isn't the Fed trying to push money out like trying to push on a rope?
Some note that the Fed can not give money away or spend it and that banks only want to lend to  credit-worthy businesses but those and consumers are still deleveraging.  So then how can the Fed push money into the economy? The government has no real limit on the amount they can borrow and spend.  They have a debt limit but they increase it anytime they get close, so it is not a real limit.   This government deficit spending, financed by the central bank making new money, is the real source of inflation.

Can't people just switch to another currency?
When there is hyperinflation in one country, say Zimbabwe dollar, the people there can switch to some other currency, like the US dollar, and life goes on.  This is called dollarization.  The government will probably fight this but people will still do it.   However, the world reserve currency is already dollars so switching to dollars is not an option.

If any major currency enters hyperinflation (Yen, Pound, Euro, Dollar) then plenty of people will start to worry about other similar major currencies (ones with large debts and deficits in their home countries).   People will flee bonds in these others, just to be safe.   This will cause the others to enter hyperinflation.   So soon after one major currency enters hyperinflation some other major ones will too.   So even the idea of people in the USA using Euros, Pounds, or Yen won't work out.

Another problem is that the central banks running other currencies use US dollars and other major currencies as reserves.    If they want to prop up the value of their currency they spend some dollars buying their currency and their currency goes up.  But if US dollars and their other reserves become worthless, then they will not be able to hold up the value of their own currencies.  So as the major currencies enter hyperinflation, currencies that use these major currencies as reserves will also enter hyperinflation.   This is nearly all paper money.

After the major currencies are in hyperinflation and the ones that use them as reserves are in hyperinflation even others will enter hyperinflation.    Even many currencies that are well run with governments not near the normal hyperinflation debt and deficit levels and not with reserves in dollars will be in trouble.    The distrust of paper money will be far greater than now so it will not take as much for people to flee a particular paper currency.  I would not be surprised to see international commerce clearing in gold, though within many countries I expect paper money will still be used.

There is no danger of hyperinflation as long as we are in a liquidity trap.
A liquidity trap is where even though the central bank has forced interest rates down near zero people are still paying down debts and not borrowing and spending.  In hyperinflation people are not in a hurry to pay off debts and are in a hurry to spend.  So by the definitions if you are in a liquidity trap you are not having hyperinflation.   The problem is that the liquidity trap could end next month or next week.  People might get worried about inflation, stop rolling over their bonds, and buy inflation hedges at any time.  There is no long term safety from calling the current situation a liquidity trap.  Hyperinflation often comes right after deflation.

Why would it matter if the US dollar was no longer the world reserve currency or oil was not priced in dollars?
Currencies are fungible, in seconds you can exchange any currency for any other currency, so you might think it does not matter what oil is priced in.   The US military burns lots of oil.  Right now the US can just print some money and get as much oil from the Arabs as they want.  If they could not do this but had instead to tax their people to get the money to buy oil it would be much harder.   Anyone who has an option to buy some amount of oil at some price in dollars 6 months from now will usually save up the money in dollars.   While they could save up the money in some other currency, there is a risk that the currency fluctuations over the next 6 months will be such that what they saved is not enough to buy the oil at that price.   So they can eliminate this currency risk by saving in dollars.   Having international commodities priced in dollars increases the demand for dollars.   If the Arabs and Chinese no longer took US dollars there is a good chance the US dollar would be headed for hyperinflation right away.

Bonds are money so monetization does not increase the money supply!
With most definitions of the "money supply" when the central bank makes new money and buys bonds the money supply is increased.  However,  there are some definitions of "money supply" that include government bonds.  With such a definition, monetizing government debt (exchanging new money for government bonds) does not increase the money supply.   So it may seem that by choosing this definition of the money supply that monetizing debt does not matter.

However, the price level should be understood from the equation of exchange.   It is not only the quantity of money but also the velocity of money and GNP that determine the price level.  If Grandma was holding a 30 year bond for the last 25 years and then sells it to the central bank (monetization) she then has cash.    The velocity was really low (not moving for 25 years) but now she can spend it and the velocity will be that of regular cash.   So the velocity is far higher.  When this is done with a trillion dollars worth of bonds you get a far higher average velocity of money.    With a money supply definition that includes bonds, hyperinflation is mostly about increasing money velocity.

With this type of definition of the money supply then deficit spending, which is financed by creating new bonds, is the reason the money supply is increased.     So the size of the deficit tells you how fast the money supply is increasing.   Here the blame for hyperinflation is put more on government deficits and less on the central bank.   Remember always that hyperinflation needs both deficits and monetization.    In this view, as long as the deficit is out of control the money supply is out of control.

While there is some flexibility in how you define the money supply, a correct view of reality has to match the many many historical cases of hyperinflation.   In historical cases of hyperinflation, monetization is the key part of what is going on, so it must make a difference in any correct theory.

Why worry about the debt when interest rates are so low?
 The interest on the national debt is a big item when interest rates are at record lows.  When interest rates get back to normal levels the interest on the debt will be clearly unsustainable.

Can't we wait till there are signs of inflation before doing anything?  Inflation should be a continuous function not a sudden tsunami.  How can you worry about hyperinflation when things currently look more like deflation?
Deflation can come right before hyperinflation.  It has been said that "many of the worst periods of hyperinflation are preceded by deflation" but someone I respect disputes this and so far I can not find data to support this claim.

Because the hyperinflation feedback loop is so strong it is hard to escape hyperinflation once it is started.  It is much easier to prevent it than to halt it after letting it start.  In other words, if you wait till hyperinflation starts it is probably too late.  Hyperinflation really does hit like a tsunami.   It is not just a bit more regular inflation where the dials of monetary policy can be adjusted.  Things can flip from deflation to hyperinflation in a short period.

There are many other words for this kind of phenomenon: tipping point, positive feedback loop, avalanche, stampede, chain reaction, explosion, death spiral,  point of no return, adverse feedback loop, vicious circle, last straw, tsunami, passing the Rubicon, discontinuity, non-linearity, etc.    The reason for so many words like this is that the world is full of examples of phenomenon like this.    

Couldn't the same method Volcker used to stop inflation stop it this time?
No it can't.  If interest rates got to Volcker levels then the interest on the debt would use up all the taxes collected.  If this happens nobody would want to buy US bonds so the Fed would be buying everything.  So newly printed money would have to be used for all non-interest budget items.   The quantity of new money would be high enough to cause hyperinflation.   The Fed can not let interest rates go up without making the government obviously bankrupt but if it tries to keep interest rates down then inflation will keep going up.   It ends badly either way.

There is no way people could panic out of bonds as there is nothing else big enough to absorb that much money.

Most of the value of bond market just goes away. If you have a 30 year bond at 3% and suddenly there is 5% inflation per month, the value of the bond is less than 1 penny on the dollar. The 99 cents are not absorbed into anything, just gone. 

When hyperinflation starts people buy real things.  This might be extra cans of tuna, a piano, land, gold, silver, etc.   In hyperinflation you can see things like a trillion dollar bill only buying 3 eggs.  So the idea that there is not enough real things to absorb the value of the bond market is wrong two ways.  First, the bond market value collapses.  Second the currency value collapses.   I think it is foolish to believe it can't happen when it has happened so many times before in the history of paper money.

Your theory that debt over 80% of GNP and deficit over 40% of spending results in hyperinflation is contradicted by Japan's case, so science says we must throw out the theory.
Economics is not an exact science.  If in 40 cases those numbers resulted in hyperinflation and in one case it has not yet resulted in hyperinflation, then the numbers should still be viewed as a good predictor of hyperinflation. Also, if Japan gets hyperinflation in 2013 then even this outlier will be gone.

Advanced Western democracies don't get hyperinflation.  But of these Japan is way ahead of the US.
Part of the idea here is that Japan has been a mess for 20 years so probably the US has lots longer till hyperinflation.   Since the debt level is so high and the politicians seem to be taking control of the central bank now, there is a good chance that Japan does get hyperinflation first.   The myth that advanced democracies can't get hyperinflation will be shattered if Japan does get hyperinflation.  Once this myth is broken people will not be so confident that the US is safe.   So they may start getting out of US government debt.   This could start the hyperinflation feedback loop for the US.   Sovereign defaults seem to trigger other sovereign defaults, hyperinflations may also.   When investors have lost money on one government's debt  they are less willing to leave their money with others.  So Japanese hyperinflation could well cascade to other countries, including the US.

Hyperinflation and even inflation predictions have been wrong so far, why take them seriously?  How is it they have printed so much money so far and had hardly any inflation?
It is certainly true that anyone predicting hyperinflation of the US dollar before Sept 1, 2012 was wrong.   This does not mean that all predictions of US hyperinflation are wrong or nonsense.   It seems that all fiat money will come to an end at some point, it is just hard to say when.  Note that anyone predicting price deflation in the US has also been wrong.

The equation of exchange is not just some quack idea.  It is a core economics formula that is always true.  With this formula it is very reasonable to expect that prices will be going up if the money supply is increased substantially.   There are 2 main factors that have fooled people.   The first is that the way they increased the money supply was by buying bonds, which lowers the interest rate, which lowers the velocity of money.  As the equation of exchange shows, a lower velocity of money can compensate for an increase in the money supply, for awhile.   The second big reason the inflation predictions were wrong is that Bernanke had a new trick of paying interest on excess reserves.   This has kept about $2 trillion in new money from really leaving the Fed.   If you make new money but don't let it out of the building it does not cause inflation.  People have not yet adapted their models to this new trick.   I believe the right adaptation is to view the excess reserves as part of government debt.  If you do this then the model will fit with low inflation so far but increased risk of hyperinflation as the government debt is really $2 trillion larger than people realize.

It might seem that paying 0.25% on excess reserves would not be enough to get banks to leave money at the Fed.   But the Fed has so suppressed short term interest rates that this is actually a reasonable choice for a bank to keep some "risk free" and "liquid" money.

What causes the velocity of money to slow down or speed up?
The interest rates affect the velocity of money  and inflation rates affect the velocity of money.   If money earns 0% and inflation is 0% there is no hurry to get it to the bank or spend it.   If money is earning 20% you don't leave it under your mattress but put it in the bank where it moves on.  When prices are going up 20% per day you run from where you get paid to where you will spend your money.    The initial response to a central bank printing lots of money and lowering interest rates is for the velocity of money to go down.  This means that for awhile prices don't reflect the new money.  One of the positive feedback loops in hyperinflation is that as prices go up the velocity of money goes up, which further drives prices up, which further drives up the velocity, etc.

But the US central bank is different!
There have been over 100 cases where the combination of a government spending way more than they got in taxes and a central bank printing money and buying their debt to help them out (often after changes in law or leadership) caused hyperinflation. What is the core difference in the US government or US central bank that anyone could think sets them apart from these 100 cases?

Could the US gold prevent hyperinflation of the US dollar?
 Having assets could slow the hyperinflation a bit but it does not matter how much gold a country is holding if they are spending twice what they get in taxes. They will have to keep printing money, so there is not a finite amount of paper money.  With a constantly increasing quantity of paper money it is not possible to hold any fixed exchange rate of dollars to gold. It is the debt and deficit, and the fact that government will always get the central bank to print money and buy government bonds when things get desperate, that causes hyperinflation.  When France took the extensive church land holdings and used that to back a currency they still got hyperinflation.   If assets could prevent hyperinflation France should have been safe, but they were not.   The ongoing deficit meant there was not a finite amount of paper money.   I highly recommend this article on France's experience. A more recent example is hyperinflation in Iran even though they have huge amounts of oil, which should have protected them if gold could protect the US.

Even though the Fed is making lots of money the banks are just hording it so there is no inflation.
The Fed is making money and loaning it to the banks who are then using it to buy Treasuries or earn interest from the Fed on their excess reserves.  In both of these two ways for the banks to earn interest the money is out of circulation.  However, the government is spending the money it gets from selling the Treasuries.   And at some point the banks will be able to earn higher interest loaning to companies and people who will put the money into general circulation.   Then they will redeploy their excess reserves and the money from Treasuries as they come due.

To get real hyperinflation you need lots of paper money and most US money is just electronic.
Anyone with electronic money can withdraw paper money.  The banks with electronic money at the Fed can withdraw paper money.  If the Fed did not have enough paper money they would have to print more.  Since electronic money can be converted to physical paper money, there is no real difference. 

The Fed could not print money fast enough to counter the deflationary forces.
In Zimbabwe they printed $100 trillion dollar notes.   Clearly the Fed could easily print $1 million, $1 billion, and $1 trillion dollar notes and buy up all  private and government bonds.  Since the Fed usually works on computers, it is even easier than printing up paper.   This would overpower the deflationary forces.

Hyperinflation is usually just a few bad years, why worry about it?
Hyperinflation destroys the middle class.  Saving are wiped out, retirement funds are wiped out, life insurance policies are worthless, etc.   It destroys the economy.   In Germany it arguably lead to Hitler coming to power and much more than a couple bad years.  In any case, it was so bad that generations later Germans still fear hyperinflation.

Why does it take both a large debt and a large deficit to get hyperinflation?  Could the US avoid hyperinflation by defaulting on the debt?
If a government just had a large debt but no deficit it could default and be fine, since it does not need to borrow.   Having lenders hate it is ok if it does not need to borrow more.   If it had little or no debt people would be happy to loan it money even with a large deficit.  When a government has both a large debt and a large deficit, such that nobody wants to loan it any more money, then it must get the central bank to make money and hand it over.

What if much of the debt is held by foreigners?  It does not upset voters so much if the government defaults on debt held by foreigners.    Instead of printing lots of new money if people stop rolling over bonds you could just decide not to pay them.   Since a large part of the debt is held by foreigners they might just say they are defaulting on the debt held by foreigners, which would reduce the effective size of the debt to levels that are typically manageable.    The problem is that nobody would buy more bonds except the central bank.  After burning some or all of the bond investors you don't get more bond investors.   So then the full deficit has to be funded by newly made money and you get hyperinflation anyway.

What if Japan defaulted?   Most of Japan's debt is held by Japanese banks and retirement funds.  The banks in Japan get deposits and then buy government bonds.   If the government defaulted on the bonds the banks would be bankrupt.   The Deposit Insurance Corporation of Japan could not handle such a widespread bank failure.   Probably voters would expect the government to backstop this organization and failed retirement funds.  So really even if the government defaulted it would still be on the hook for the money or the voters would be very angry.  

Why would people ever stop buying US Treasuries?  How could there be a "bond panic"?
The yield is less than the inflation rate.  It is a guaranteed loss in real terms.  The Chinese yuan is going up in value compared to the dollar and has a higher interest rate.  The US rates have been going down for more than 30 years and are now near zero thanks to the central bank making lots of money and buying bonds.  All this new money will probably push the inflation rate higher.   If interest rates start going up then bond values go down.  Many of the bonds are held by traders, not long term investors, who have been speculating that the Fed would drive up bond prices.  If the price of bonds starts going down then many people will sell.   This will cause the price of bonds to go down further, which will in turn cause more people to sell.  So you can get a positive feedback loop or bond panic.

The deficit is about 10% of GDP not close to 40%!
The 40% number is 40% of government spending, not GDP.   If the government spends twice what it gets in taxes, then the deficit is 50% of spending.

The US deficit is not really 40% of spending.
At the moment that is correct.   From 2010 to 2013 spending has been stable and taxes have gone up so the deficit is smaller (see graph above).  But a rise in interest rates or another recession could easily put it over the 40% number.  So while hyperinflation is not guaranteed, the US situation is far from safe.

Quantitative Easing is different than printing money!
I have a collection of over 100 euphemisms for "printing money".   If the Central Bank is making new money and buying government bonds it is monetizing debt.  It is replacing bonds with new money.   It does not matter if they do it on computers since money on computers can later be withdrawn as paper money.

I can see inflation in "debt free money" but how does it work in "debt money"?
If the central bank prints money and loans it into existence, some people call this "debt money".  The idea is that money comes into existence with a debt which should take it back out of existence when the debt is paid back to the central bank.  If the Treasury just printed money and spent it they call that "debt free money".  Clearly if the Treasury just printed and spent money you would get inflation but how does it work with "debt money"?  The key to understanding this is that governments don't really pay down their debts.  The way they pay back old bonds is by first selling new bonds.   If the government never really pays back the central bank then there is no real difference between "debt money" and "debt free money", at least as far as the government use of it.

Why not have the Treasury mint some $1 trillion coins?
This would be very inflationary.  Even if the coin "does not circulate", once it has been deposited at the Fed there will be $1 trillion more in paper or electronic money that can circulate.  So the base money supply would go up by $1 trillion.   If the politicians are able to make $1 trillion without destroying things then you should expect them to make another and another.  It is foolish to think they would stop at one.

There is a saying something like, "the only thing worse than a currency run by a central bank is a currency run by politicians".   Once the politicians get the ability to make money or outright control of the central bank, then the central bank can no longer control inflation.  This typically seems to happen shortly before hyperinflation starts.   If they can get money just by making it then controlling spending no longer seems important.  Very soon the money making and spending grows fast enough that people will want to get out of bonds.  As people get out of bonds the government will have to make more money.   The more they make the less people will want to hold the bonds and eventually even the currency.   You get a positive feedback loop or death spiral and the currency goes down.

The law says "platinum bullion coins".   A bullion coin is one where the value is determined by the quantity of precious metal.  This is different than fiat money where you can stamp any quantity you want on it.   So probably making a coin and stamping $1 trillion on it is illegal.   However, I would not count on that stopping this from happening.

The demand for money is unlimited, particularly the world reserve currency, so people will always want it.
There is no guarantee that the US dollar will continue to be the world reserve currency.  If it starts to get hyperinflation it would quickly go from the majority of foreign central bank assets to a minority just because the value dropped, even if the banks did not sell any bonds or dollars.

People want money for what they can buy with it.  As long as it can buy something they will want it.   But if it is dropping in value fast they will buy fast and not hold the money for a long time.   The usual end result of hyperinflation is that you can not buy anything with that currency, so nobody wants it.

There are no bond vigilantes so nothing to worry about.
If a currency is at risk of hyperinflation then even if the bond values drop people shorting them can lose.  The problem is that they win in dollars that are worth less in real terms.   So shorting bonds when there is a risk of hyperinflation does not work well.   This may be  why there are so few bond vigilantes.

The bankers control the Fed and it is against their interest to have hyperinflation.
The government really controls the Fed, just like in all the other cases of hyperinflation.  When the government is faced with a choice of closing its doors because it can not pay people or forcing the central bank to buy government debt they never choose to close their doors.  Note that if the government shuts down the central bank will shut down too.   So it is actually in their interest to prop up the government by buying their bonds.

You can't have hyperinflation when house prices are falling.
The idea is that during hyperinflation the price of everything must go up fast, so clearly with asset prices falling we can't have hyperinflation now.   This is not accurate.  You might argue that we don't have hyperinflation yet but this in no way protects us from having it in the next year or two.  Also, at the start of hyperinflation when interest rates have first shot up the asset types that are bought with long term loans will drop in value because you will not be able to get long term loans any more.  Eventually, when houses are bought with cash, house prices will go up with everything else.

You can't have inflation when there is high unemployment.
The 1970s proved this is not true yet many people continue to believe it. High unemployment is normal in hyperinflation.  The economy is a mess.

But there is no inflation.
The CPI shows little inflation.  However, it does this by averaging in things like housing prices that are going down with food and energy that is going up.  The average person who is not buying a house just sees the inflation.  Also, when the central bank first starts printing lots of money and buying bonds it lowers the interest rate and lowers the velocity of money.  A lower velocity of money can compensate for an increased quantity of money, so at first prices don't go up.  But eventually they will.

Low interest rates prove the market is not worried about hyperinflation.
The current low interest rates are because the Fed is artificially driving up the price of bonds, which lowers the interest rate.  This is not a market rate.  If the Fed were to stop buying bonds the rates would shoot up, particularly now after they have printed so much money in the last few years.  However, it does seem the market is not yet worried about hyperinflation.  When the Fed is buying bonds faster than the Treasury is issuing bonds, then the private market will be exiting bonds.

I thought inflation was just from printing money.
Hyperinflation comes from an increasing money supply, an increasing velocity of money, and a decreasing real GNP.   But the increasing velocity of money and the damaged GNP are really the results of the increasing money supply.   So printing money is the key part of it.  But to me the interesting thing is why they keep printing money when it causes such devastation to a country.  The reason is the government is spending far more than it gets in taxes and the only way to fund the deficit is with new money.

The increase in prices for oil or food is from supply and demand issues and not due to monetary policy.
There is a chance that the price of one commodity going up is due to supply and demand factors for that commodity.   But if commodity prices across the board are going up then either it is an increase in the quantity of money, an increase in the velocity of money, or a reduction in the GNP, or some combination of these.    The equation of exchange is always true.

Don't you need a black market and loss of taxing ability to get hyperinflation?
When your plumber or the retail sales lady tell you they will give you a discount if you pay cash, you are probably dealing with people who are not paying all their taxes.  During hyperinflation there is usually a growing black market and lots more barter.  If you trade some sweet potatoes that you grew for some fish that your friend caught, neither of you will be paying VAT, sales tax, income tax, nor FICA on the transaction.   Loss of taxing ability increases deficits and so increases monetization, which probably drives up tax rates, which drives more people into the black market, which increases deficits, etc.    So loss of taxing ability is part of the hyperinflation feedback loop.  However, it need not be there to start hyperinflation.

People/countries won't get out of bonds because it would drive the price down.
The idea is that a place like China will not sell their US Bonds because selling would make the price go down.   This is just wrong.   If China understands that if Japan sells first then China gets less money they will be in a hurry to sell first.

Right now the Fed is propping up the price of bonds.  That makes now the  best time to get out.   People who wait a year or two will probably not do nearly as well.   In this situation there could be a rush for the exits at any time.  First ones out will be better off than the last ones out.  The logical move is to be first out.

Look at Japan.  The people that sold JGBs earlier did better than those selling now.  I expect those selling now will do better than those selling later.  The best move was to get out already.

People won't get out of US bonds because they are not worried about default.  
The idea here is that the reason people ran away from Greek bonds is that they were afraid they would not get paid but since the US can print money people don't have to worry about not getting paid, so they won't run.   This is not correct.  People also get out of bonds when they are afraid that inflation will destroy the value of their bonds.  This has happened in many other countries that printed their own money.

Don't you need to increase the money supply by 26% in a year to get 26% inflation?
Nope.  The most important thing to understand about price inflation during hyperinflation is that the velocity of money and GNP become important factors.  You might have 26% inflation where 10% was due to increase in the money supply, 10% was due to increased velocity of money, and 6% was due to lower GNP.

If a country had hyperinflation then clearly they had stupid politicians and/or stupid people running the central bank.  The US leaders are not stupid.
This is not correct.  The US leaders are no smarter than the ones in countries that had hyperinflation.  Leaders all over the world think they can control markets and make things better.  The idea that they can force the interest rate lower and make the world better by deficit spending and printing money has fooled many very smart people, including Bernanke and Krugman.  The problem is that in the short run it seems to work but in the long run you are worse off.   Leaders are too focused on the short run.

France in 1920s did not get hyperinflation why should we worry?
The vast majority of the countries which printed their own currency and had debt over 80% of GNP and deficit over 40% of spending got hyperinflation.  The fact that a few did not might be cause for hope, but it is not sufficient reason to ignore the danger.  

Paper currencies backed by gold have also fallen.
When a paper currency that is backed by gold stops being backed by gold, as is common at the start of a war, then the currency can fall.  Paper backed by gold is not as good as a gold coin in your hand.

The government could take everyone's gold like they did in 1933.
I think there would be a lot less cooperation and a lot more shooting if they tried that again.

If China sells their treasuries then the renminbi goes higher and higher and their companies that export go bust.
It depends on what they do with the dollars as they sell the treasuries.  If China sells treasuries and buys renminbi, then the renminbi would go higher and higher.  However, China could sell treasuries and buy gold, and print renminbi fast enough to keep the value as low as they wanted compared to the US dollar.  If they did not know what else to spend renminbi on they could use some of those to buy gold as well.  In this case they end up with gold and get to spend a bunch of renminbi, but the value of the renminbi does not go up compared to the dollar.

The central bank controls the interest rates why would they ever let them go up?
If the central bank is willing to buy all bonds for sale then they can put a lid on interest rates.   If they do this they may have to make huge amount of new money to buy the bonds.  This can make people afraid that the value of the money will be dropping, so people may decide to sell their bonds.   Since there is a danger of an "adverse feedback loop" where the more people sell bonds, the more the central bank makes money, it is sometimes wise to let interest rates go up.   However, if the government will be clearly bankrupt when the interest on their debt is more than they get in taxes, the central bank is sometimes left without any good option.   If they let rates go up people will stop buying bonds because the government is clearly bankrupt.  If they keep forcing rates down then people will stop buying bonds because there is so much new money.  If the government has huge debt and deficit and the central bank becomes the only buyer of government bonds you get hyperinflation.  

What about debt deflation?
When the central bank loans money into existence there is more money.  Since demand deposits also count as money by most definitions, when   fractional reserve banks lend against demand deposits the money lent out and the demand deposit balance both count as money, so there is more money.  One definition of inflation is an increasing money supply.   If debts are being paid back faster than they are being created you could have the money supply decreasing, which is one definition of deflation.   When the government debt is going up by more than $1 trillion per year, this will eventually eclipse any finite amount of private debt that is being paid down.  At some point private debt will start to increase again and the government will still be borrowing a trillion a year, so there will be lots of inflationary pressure.     It is hard to say when inflation will really kick in though.  Note that in fiat money "price deflation" is very unusual.

By printing money they lower the value of the dollar and help the economy by increasing exports.
The US imports much more than it exports.  If the value of the dollar goes down it hurts the economy more than it helps.  In 2010, U.S. exports amounted to $1.3 trillion and imports amounted to $1.9 trillion. Trade deficit was around $600 billion.    A weaker dollar means that $1.3 trillion part of your economy can do a bit better but that a bigger $1.9 part of your economy suffers.   This is not a win.

The original definition of inflation was an increase in the money supply and it seems wrong to use the newer definition of inflation as an increase in prices.
In calmer times, where the velocity of money and GNP are more stable, it is a reasonable point of view.   If the velocity and GNP are not changing, then the equation of exchange shows that the quantity of money and prices are linked.   But during hyperinflation you need to understand how increasing velocity and  dropping GNP cause prices to go up too.

Doesn't much of the pain from the US money printing go to the rest of the world?  Does the US have an inflation tax on people outside the US?
Yes, many countries peg to the dollar using a Currency Board where they end up buying dollars and supporting the dollar.  It lets the US export some of the inflation.

The world reserve currency can win by printing money, for awhile.   If the US prints another $1 trillion dollars they get $1 trillion dollars.   Now existing dollars and bonds everywhere will eventually lose about the same total value, $1 trillion dollars.  But about half of the dollars and bonds are outside the US.  So the people inside the US lose $500 billion and gain $1 trillion, for a net win of $500 billion.  The people outside the US lose $500 billion.  In effect the US is able to collect an inflation tax from the rest of the world.   At some point the rest of the world will no longer put up with this and so will stop holding so many dollars.

The government/central-bank has a monopoly on the issuing of currency, so they have full control over the value of the currency.
The value of the currency is totally the responsibility of the government/central-bank and theoretically they could maintain the value.  However, the government/central-bank often does not show enough self control on the issues spending, debt, and deficit and so loses their ability to limit how much currency they issue.

A government budget is not like a family budget as they can always issue more currency.
True.  A government that issues a currency is really very analogous to a public company that issues stock in their own company.    A company/government with good income/taxes and low expenses/spending can issue more stock/currency and it will hold value.  But a company/government that is losing-money/running-a-deficit and still issuing more stock/currency will see the value of their stock/currency go down.  If the company/government is so desperate that it can not stop issuing stock/currency then the value of the stock/currency will plummet until there is a major reorganization of the company/government or it fails completely.

If inflation got as high as 10% then in only a few years it would eat away at the real value of the national debt and reduce it to manageable levels.  If you have a fixed rate 30 year mortgage at 4% and inflation goes to 10% you will find it gets easier and easier to make your payments, so why can't the government do the same thing?
I am sure that no country that had long term fixed rate debt in its own currency ever had hyperinflation.  If the debt was in 30 year fixed rate bonds then a few years of 10% inflation would destroy those bonds and the trick would work.  However, that is not the US situation.     The US debt is mostly short term, not long term.  The Fed has worked hard at buying up the long term debt and selling short term debt with their "operation twist".   So now it is like the US has a variable interest rate on their debt.  If inflation went to 10% and interest rates went to 12% the interest the US paid on $18 trillion (Treasury and Fed)  would be moving towards $2.1 trillion but the total taxes last year were only $2.4 trillion.  The deficit would increase drastically with the higher interest payments and it would be clear to all that the US was going to have to print money like crazy.  People would flee US bonds.  The Fed would step in as the government's lender of last resort and print money and "lend" it to the government.  Given the amount of short term debt, increasing inflation to 10% does not result in a stable situation that the government could keep for a few years, things would fall apart fast then.   It looks like we will see this sort of problem unfold with Japan.

I think that hyperinflation is caused by X, not high debt and deficit or the central bank monetizing debt during a bond panic.
If you think debt is not a factor, then why did the historical hyperinflations all come to governments with large debts and deficits?

That you think the Fed Prints Money shows how clueless you are.
In common jargon the phrase "printing money" means "making money" or "debasing the currency".  The fact that the actual printing is done at the Treasury instead of the Fed does not make a bit of difference to the reality that when you make a lot of new money then eventually the money won't be worth as much.   It also does not matter that it usually starts out as electronic money as the owner of the money could make a withdrawal in paper currency.   There are more than one hundred ways to say debasing the currency.

Doesn't the Phillips Curve prove that inflation reduces unemployment?
When Phillips wrote his paper in 1958 he was using data from under a gold standard.  Under a gold standard wages go up when labor is in short supply.  But it is the short supply of labor that is causing the wages to go up, not inflation that is causing full employment.    The people using this paper are getting cause and effect backward.  People have taken this study under a gold standard and totally misinterpreted the results to mean that printing money under a fiat standard is a good thing.   It was not a study of the effects of inflation under fiat money.   The only truth to this idea is what Keynes pointed out,  if people are not used to inflation, then inflation may be used trick people into accepting a lower real wage.   If people are working for lower real wages, then employers can hire more people.   So if inflation makes people poor enough then employment can go up.  

Conditions now are like during the deflationary period at the start of the Great Depression.
Nope, things are not the same at all.  Back then the Fed could only have   $1 paper dollar in circulation for every $0.40 they had in gold, even though they told everyone they could turn in $1 paper dollar and get $1 worth of gold.  This was a Ponzi gold scheme that was bound to fail.  As other countries took out gold the Fed had to reduce the money supply.  Eventually they had to outlaw gold or the Fed would have been bankrupt.  Once US citizens could not turn in their paper money for gold they were able to print more money and end the deflation.  Nothing like that is acting to reduce the money supply now.  With pure fiat money, no ties to gold, they can print trillions per year if they want to.

Can't inflation linked bonds protect you from hyperinflation?
Nope.    The Treasury Inflation-Protected securities (TIPs), do not adjust the interest rate, only the principal that the interest is calculated on.   So if the going interest rate is 1,000% you might still be earning 2%.  What is paid out will seem very small.  The currency may fail completely before the bond becomes due.  So the larger principal paid out at the end of the bond will be in worthless money.  Indexing the principal will not save you.  You do not want your money tied up in any kind of bond during hyperinflation. 

Nothing can replace the US dollar and bond market as nothing else is big enough.
After the bond market crashes it won't be so big.  Having huge amounts of debt does not make the US more attractive to investors it makes it less attractive.  The Chinese Yuan is going up in value relative to the dollar and pays higher interest rates and the government has far less debt than the US relative to GNP, and their GNP is growing much faster.  So over time it makes sense for investors to move away from US debt and toward something like Chinese debt.

How is hyperinflation stopped? 
There can be many things that cause government spending to get out of control (war, depression, foreign debt, etc) and so start hyperinflation.  So people often put the blame for hyperinflation on these many different things.  However, the only way hyperinflation ends is when central bank stops monetizing a large government deficit.  I think this is the best proof of what the core cause of hyperinflation really is.   This can happen if the government reduces or eliminates the deficit or if the government fails completely.   Along with cutting the deficit, governments often make a new currency and have new laws limiting the central banks ability to monetize debt.  But the key to success is controlling the deficit.  I think the only way the deficit is ever cut enough during hyperinflation to stop it is by cutting the size of government, as taxes are usually already so high that people are pushed into the black market or out of the country to avoid them. Governments that are not able to make the cuts seem to fail completely.

Debt currency issued by central banks is inherently deflationary.  
Since central banks charge interest on all the loans they make and make safe enough loans that they almost always get the principal back, it would seem they tend to reduce the total currency outstanding over time.  So in theory the central bank debt currency method should be inherently deflationary.  The problem with this idea is that governments just keep increasing their debt to the central bank.  Also, at least in the US case, any profits the central bank makes on this interest are paid to the government, not withdrawn from circulation.  So, in practice, this method results in more and more government debt and an ever increasing money supply.

There is always a buyer for each seller, so how can you have a bond panic where people get out of bonds?
Much of the debt is in short term bonds.  People can just not roll over their bonds.  So there does not have to be a buyer, people can get out of bonds by just waiting for the bond to mature and getting paid in cash.

When the central bank is trying to control interest rates it buys bonds.  So people can get out of bonds by selling to the central bank.

The government is running a huge deficit so it can not just pay off the bonds with tax money but in fact needs to sell more and more bonds all the time.  If others are getting out of bonds then, assuming the government is still getting money, it is the central bank that is buying.   As the central bank's holdings of bonds goes up there is more and more money in circulation.

People have been predicting the end of the world for a long long time.  It never happens.
Hyperinflation is not the end of the world, just the death throws of some fiat currency.   Inflation of 5% per month or even 50% per month is, amazingly, something that most people adjust to.   Crime will probably go up a bit, but hyperinflation is not a "mad max" thing.  Hyperinflation has happened many times around the world and even several times in America.

How does Krugman explain hyperinflation?
Inflation is really a government tax on those holding money.   In a country where a large part of a government's budget comes from an inflation tax it, like any other high tax,  can affect people's behavior.   The way people avoid a high inflation tax is by holding less real money.   But as people hold less real money the government has to make the inflation tax even more extreme to get enough real money to spend.   But the more extreme the inflation tax the more the public reduces their real money holdings to avoid the tax.  The process can easily spiral out of control.    As the government prints faster and faster the public reduces its real money holdings more and more.   Eventually the public is unwilling to hold any money at all.  Then the government has to abandon its use of the inflation tax. 

I am summarizing from "Krugmans Macroeconomics for AP*" pages 324 to 327.

And in a blog post he wrote:  "Hyperinflation is actually a quite well understood phenomenon, and its causes aren’t especially controversial among economists. It’s basically about revenue: when governments can’t either raise taxes or borrow to pay for their spending, they sometimes turn to the printing press, trying to extract large amounts of seignorage — revenue from money creation. This leads to inflation, which leads people to hold down their cash holdings, which means that the printing presses have to run faster to buy the same amount of resources, and so on."

I would not say there is anything untrue in either of these explanations.  However, hyperinflation always involves large debt and deficit and I am impressed that Krugman could give two reasonable explanations of hyperinflation without mentioning these.   A large fraction of Krugman's posts argue for more deficit spending and more monetization of bonds, or argue against austerity.   It is no doubt easier to advocate debt and deficit if you don't think they are part of hyperinflation.

If we added to his explanation "a high inflation tax makes people want to hold less money and also less bonds" and also "on top the inflation tax for the deficit the government also needs an inflation tax for all the bonds that people don't roll over" we could get a more accurate explanation, though at the cost of making some of Krugman's favorite things look bad.   He is missing out on the "tipping point" problem of bond monetization getting out of control, which is the real key to understanding hyperinflation.

Isn't this the Quantity Theory of Money and wasn't that discredited?
The Quantity Theory of Money uses the equation of exchange, as do I, but then says that over long periods of time the velocity of money does not change much, so that prices are proportional to the quantity of money.  

Over a short time period the quantity theory of money does not hold as the velocity of money can change.  Over a short time period you can say the quantity theory of money does not apply or that it is discredited.

The theory also fails badly during hyperinflation.  In hyperinflation the velocity of money is going up fast so that prices go up much faster than the money supply. So this is different than the quantity theory of money.

Your usual explanation of hyperinflation relies on The Equation of Exchange.  Could you explain it in other terms?
Yes, I can explain hyperinflation using the Real Bills Doctrine, also hyperinflation with Corrected Modern Monetary Theory, as well as many other ways.   It should be possible to explain hyperinflation using any theory of money that fits the historical experimental record, as hyperinflation has happened many times.  If you have some other theory you would like me to use, just say so in the comments.

A central bank buying government bonds is "just an asset swap" and does not matter.  
If the government was not running a deficit and the central bank were just buying short term bonds it would not matter much.

However, if the government is spending twice what it gets in taxes, making and selling new bonds to get the needed cash, and the central bank is making new money to buy bonds, the net result is as if the government were just making huge amounts of new money and spending it.   This does matter.  History is clear on this point.

The problem is that the government never pays down the total debt.  They pay off old bonds with money from new bonds, so it really is just like they are printing money and spending it.

You don't understand our monitary system.  If you did you would realize that Government debt is not a huge deal, we will never run out of money, we aren't in danger of hyperinflation, and there is no chance we will run out of buyers of our debt.  The government can always meet its obligations in US dollars as it is the issuer of US dollars.
Our monetary system of fiat money created by a central bank is the same as the systems used in more than 100 cases of hyperinflation.    In all cases they did run out of buyers of government bonds.  The central bank monetized government bonds like crazy and they had hyperinflation.   In most cases the money died.  Government fiat currency was no longer accepted as money.   Think about this.   A government that prints fiat currency really can run out of money, as that paper can fail to be money at some point.  History contradicts your theory of how things work.   You don't understand history.

Many of the obligations, like social security, medicare, obamacare, unemployment, etc are real world obligations and not just debts denominated in US dollars.   As the government prints more money the prices for these real world obligations goes up, which means they have to print even more, which makes the prices go up further, etc.    If the government just had debt, and no deficit for real world obligations, there would be no risk of hyperinflation.

Isn't this a self fulfilling prophesy?  If everyone believes the dollar is going to crash then they will get out of the dollar and it will crash, so isn't this just a self fulfilling prophesy? Doesn't publishing this scare stuff help bring about the end?
 I think that the more people that understand the danger of hyperinflation the better chance we have of avoiding it.  So explaining the dangers could help save us from hyperinflation.  Preventing hyperinflation is far easier than stopping it.  I have worked for years to explain hyperinflation as clearly as I can.  Remember not to shoot the messenger.   In any case, at the end of 2012 this blog has 2 followers and one of them is the author.   Update: There are now a total of 3 in June 2013.  So if hyperinflation does happen it is not because I scared millions of people.

If the world gets hyperinflation I think Dr. Paul Krugman deserves more  credit than anyone else.  He has cheered money printing and deficit spending for many years.  He has been extremely critical of attempts by governments to cut spending.   He has ridiculed anyone who is worried about deficits, calling them austerians.   He writes for the New York Times.   He has millions of readers.  He has a Nobel prize in economics, so politicians feel justified in following his advice.   Even as Japan went crazy with money printing and deficit spending, bringing hyperinflation much closer, Krugman cheered Japan on.  Krugman has been providing the intellectual cover for the political choices leading up to hyperinflation, so when it comes, Krugman should get the most credit.

Productivity improvements will keep hyperinflation at bay.
Productivity improvements are on the order of 1% per year.  Hyperinflation rates are far higher than this, so this can not make even a noticeable dent, let alone prevent hyperinflation.  Also note that all the historical cases of hyperinflation normal productivity progress failed to prevent hyperinflation.  So this claim fails by both the magnitude of the numbers and historical evidence.

Couldn't you get hyperinflation without a huge debt because of a war?
Yes.  If a country was attacked and clearly going to lose, the currency could fail even if it did not have a huge debt before the war.  The embargo of Iranian oil caused Iran to get hyperinflation even though it did not have a huge debt ahead of time.   If a government is overpowered their currency can be killed.

This is a simple to understand case, and worth mentioning, but I find general case of hyperinflation much more interesting.

Real hyperinflation is one where the currency does not survive.  It is a complete loss of faith in the currency to where nobody accepts it any more.
The problem with this type of definition is you can only use it after it is all over.   When people are losing their life savings and ask you, "is this hyperinflation" all you can say is "I don't know yet".     What you are talking about is "death of a currency".   Hyperinflation is the process a currency  goes through when it is dying.  It is the decaying sickness of the currency, not the death itself.

The equation of exchange is at the core of your position but this formula has been discredited.
The velocity of money in the equation of exchange is usually solved for, as it is easier to get measurements of the other factors.  So some have argued that velocity is just a made up number to make things work.  But the concept is reasonable.  And using algebra to solve for things is not cheating.   The formula does explain what is going on and help understand how things work.   While I have seen the claim that it was discredited a number of times I have never seen anything that came close to really discrediting it.  I don't believe there is a respected economist anywhere who would say the equation of exchange is not true.

Collapsing asset bubbles are deflationary.
Collapsing asset bubbles usually contribute to deflation.  However, when a bond bubble collapses and the central bank starts buying up huge numbers of bonds with new money, you don't have deflation.  This flood of new money makes hyperinflation.

Hyperinflation is a kind of extreme thing, why are you so focused on it?
Hyperinflation is the reason that the Keynesian, MMT, MR, and other economic schools that advocate "government stimulus" are wrong.  These schools ignore the end that their ideas lead to.   If every time there is a problem you grow the size of your government then eventually the spending is too much for the productive parts of the economy to support and debt and monetization get out of control.   If people understand hyperinflation they can see the flaw in these schools of economic thought.   So I think understanding hyperinflation is very important for understanding economics in general, yet far too few people do understand hyperinflation.  So it seems like an area I could contribute to.

This FAQ is really long.  What is the shortest reasonable explanation of hyperinflation you can give?
Printing money can give you inflation but to get hyperinflation you need out of control money printing. For this you first need a large government debt (over 80% of GNP) and a large government deficit (over 40% of government spending) in a country that prints its own money. You can get a large debt and deficit in many different ways.  Next, the government needs to get the central bank to monetize debt.  This may require replacing people at the central bank, changing laws, or just breaking laws.   Once conditions are right, a death spiral or positive feedback loop emerges where the more people get out of government bonds the more money the central bank has to print to monetize bonds, but the more they print the less  people want to hold government bonds.  This may be triggered by some event, but in most cases this is just a trigger or a last straw.  Politicians will use price controls and higher taxes to try to regain control of inflation but in doing so will hurt the real GNP of the country.    The faster prices go up the harder people work to spend their money fast.  With the equation of exchange we can understand that a lower real GNP, a rapid increase in the money supply, and the rapid increase in the velocity of money makes prices go up really fast. 

The above paragraph summary is still too long.  I want a 2 sentence summary. It is ok to skip over some stuff.
When a government that prints its own money gets too much debt, and the central bank starts monetizing debt, you can get a positive feedback loop where the more the central bank buys bonds the less anyone else wants to hold them or buy them and the less others buy bonds the more the central bank has to so that the government can have money to operate.   Hyperinflation is where this out of control monetization and an increasing  velocity of money combine to make prices go up really fast. 

Others Questions
Feel free to ask other questions in the comments. I will move some of the best questions from the comments into this FAQ.  If any of my answers does not seem satisfying enough, let me know.  If anyone knows of any site on the web that does a respectable job of defending deflation or attacking hyperinflation please mention them in the comments.   I would really like to debate hyperinflation with another economics blogger.    I will link to any post that replies to this post. 

8/11/13  Armstrong Economics