Friday, August 31, 2012

Mish on Hyperinflation

Update:  I expanded the post below into a more general Hyperinflation FAQ.   I also update that post from time to time.  I recommend readers go to that instead of this.  


Mish wrote another post on hyperinflation, something he has called nonsense many times.  I think he is wrong about hyperinflation in many ways.    

I think Mish's biggest mistake is that he defines hyperinflation differently than everyone else. The International Accounting Standard of IAS 29 says there is hyperinflation when "the cumulative inflation rate over three years approaches, or exceeds, 100%.".  This works out to 26% per year.   I have seen many other definitions for hyperinflation but they almost all have something like "inflation over X per year" or "inflation over Y per month".     People pick some inflation rate as the cutoff between regular inflation and hyperinflation.  It is just the values for X or Y that differ.  Mish; however, looks at the typical end result of hyperinflation (currency becomes completely worthless) and then defines that as "hyperinflation". This is wrong. Hyperinflation is the journey, not the end result. It is the time of high inflation, not after inflation stops.  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. 

There are some cases where they have successfully halted hyperinflation. After having inflation over 26% per year they got back down to more sane levels and the currency was not a total loss. In these cases Mish's definition would say there was no hyperinflation (no total loss of faith in currency) while most would say there was hyperinflation but it was halted.   It can sometimes take years for hyperinflation to finish (either halting or death of the currency) and with Mish's definition we could not say till it was all over if it was really hyperinflation. 

Mish thinks there is no chance of hyperinflation for the US dollar.  Hyperinflation happens when debt is over 80% of GNP and deficit is over 40% of government spending. The US is at or near these numbers, so the danger of hyperinflation is real. What happens is that the more the central bank prints money and buys bonds the less other people want to hold bonds. But the less other people hold bonds, the more the central bank has to buy them so the government has enough money to spend. You get a positive feedback loop or death spiral. Please see my writing on hyperinflation.  

Mish thinks that the US having around $400 billion in gold would prevent a complete loss of faith in the currency.  Having assets could delay a complete loss of faith in a currency, but does not help so much at preventing inflation over 26% per year.  It does not matter how much gold the US is holding if they are spending 50% more than 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 when the 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.   By Mish's logic this should have prevented a complete loss of faith in the currency, but it did 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.

Mish thinks people would not decide to have hyperinflation.  I don't believe there has ever been a "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 government or a 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 to government debt over 80% of GNP and deficit over 40% of spending when the central bank starts printing money and buying up government debt. Everyone thinks they just need to print a bit more money to make it through the next week or month and there is nothing else they can do since the government needs money to keep in operation. Nobody votes for hyperinflation. Nobody wants it. It just happens.

Mish thinks you need some major trouble like losing a war and paying reparations to get hyperinflation.  The US won the Revolutionary War, did not pay any reparations, and still had hyperinflation. Remember, "not worth a Continental"?  There was hyperinflation in the South during the Civil War. So I think it is fair to say that America has had hyperinflation twice already.  As someone pointed out in the comments, there were more cases of hyperinflation in America's colonial period.   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.  The US was not in a war in the 30s.  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.

Mish has posted that he thinks a currency crisis is coming.   Even if this is a limited drop in value, like a factor of 2, and not leading to the total destruction of the currency, it could still result in prices going up by more than 26% per year, which many people would count as hyperinflation.  I don't think it will be so limited.   Things are just too far gone.


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?

People may think it is easy to avoid or stop hyperinflation.  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 nearly twice what it gets in taxes and has debt around the size of the GNP. 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 is bankrupt.

Mish can see no reason for the Fed to suddenly print trillions of dollars.  But there is a reason it might happen.   Mish has quoted someone as saying "More than half of the debt, however, is short term, maturing in less than a year."  With a current national debt of about $16 trillion, this would mean about $8 trillion is due in the next 12 months.  If people stop rolling over bonds then the government will need the Fed to print over $9 trillion in 12 months ($8 for old bonds coming due and $1 for the current deficit).   This bond panic and massive buying by the central bank is how hyperinflation seems to typically start.

Mish will say that the Fed on its own could not cause hyperinflation but Congress could.   This displays a lack of understanding of how hyperinflation works.  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.   It is the two together that result in hyperinflation, not one alone.

Mish will say "hyperinflation is a political event not a monetary event".  I think Milton Friendman 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.   

Mish said, "Since printing $2 trillion did not spur credit expansion, pray tell why would $50 trillion?".   Mish needs 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.

Mish will say 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 Mish can not see how the Fed can put money into the economy.   He is not seeing that 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 but Mish never talks about it.

Mish thinks it is silly to compare the US to Weimar Germany or Zimbabwe hyperinflations because the situations are so different.   Each case of hyperinflation is unique, so if you are looking for differences you will always find them.   You need to understand the similarities.  Hyperinflation happens because 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.

Mish thinks that when China and others arrange things so they can do trade without US dollars that it does not matter.   He thinks if the Arabs stopped pricing oil in dollars it would not matter.  He notes that currencies are fungible, in seconds you can exchange any currency for any other currency, so "it does not matter one iota 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.   Mish does not seem to get this.  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.

Mish notes that all the recent predictions of US dollar hyperinflation have been wrong so far.   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.

Mish notes that 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.

Mish does not think the US central bank would willingly destroy their currency.  He thinks this alone debunks hyperinflation.     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.

Mish thinks that having lots of private debt makes hyperinflation impossible.   Mish defines inflation and deflation in terms of the money supply; however, Mish has his own definition of the money supply.  Mish counts "credit marked to market" as part of the money supply.   So if the bond values are crashing Mish would say there is deflation.   However, as interest rates shoot up when hyperinflation starts, bond values crash.  So you can have 26% per year price inflation, which many count as hyperinflation, at the very same time that Mish's defintion of deflation was met.   In fact, I expect that at the start of most hyperinflations, when bond values are crashing, that Mish's definition of deflation is met.  If everyone else is looking at prices going up fast and saying "this is the start of hyperinflation" and Mish is looking at bond prices crashing and saying "this is deflation", he will look silly.

The debt deflation idea is that as people are paying back debts the total measured money supply can go down.  Since central banks loan money into existence it would seem that after the money was paid back to the central bank there would be no net inflation.  The big thing that these people overlook is that governments never seem to pay down their debts.    In Weimar Germany and Zimbabwe the central bank was just loaning money into existence.    However, those governments could not pay it back but kept borrowing more and more.    The US central government debt has gone from like $9 trillion to $16 trillion in the last 4 years. During the hyperinflation people talk about the central bank is buying government bonds to stimulate the economy, stabilize interest rates, or show faith in the government.  After the hyperinflation is all over, people forget about the central bank buying bonds and just talk about how the government was just printing and spending money.  The historical narrative simplifies out the central bank so it becomes much clearer and looks so foolish.  However, this simplification makes is so that the next victim of hyperinflation does not see they are doing the same thing.   They think of the historical hyperinflations as just stupidly printing and spending money while their current government is just borrowing it.    So the same mistake is repeated again and again.