Tuesday, December 18, 2012

How to stop a hyperinflation

Hjalmar Schact and Montague Norman

If you were in charge of a central bank during hyperinflation, what would you do to stop it? Here's a brief but detailed account of how the German hyperinflation was halted in November 1923.

What is so unusual about the end to the German hyperinflation was its suddenness. Within days of a series of monetary reforms implemented in mid November 1923, price rises came to a dead stop. You have to put this into context to properly appreciate it. A loaf of bread, which cost 30,000 marks on August 30, 1923, rose to 300,000 by mid-September, fifteen million marks by mid-October, and 165 billion marks by early November. And suddenly it stabilized.

There were two not-entirely unrelated reforms implemented that month that halted the inflation:

1. the creation of a new unit of account called the rentenmark.
2. the stabilization of the existing unit of account, the paper mark.

The Rentenmark

With the Reichsbank's paper marks having lost all credibility, in August 1923 legislation was introduced in the German parliament to open a new bank of issue. After some discussion, on October 17 the Rentenbank was created. The Rentenbank was to institute a brand new unit-of-account called the rentenmark. A rentenmark unit was to be defined as one gold mark, or 1⁄2790 kg pure gold. The gold mark was the country's old unit-of-account which had been forsaken during the war. The bank would issue notes called rentenmarks which were to trade at their defined gold amount.

The catch was this. Historically when a nation's unit-of-account (like pounds or dollars) was defined in terms of certain quantity of gold, the currency issued by that nation's central bank was made convertible into gold. If inconvertible, then the currency risked floating away from its defined gold value. But the rentenmark was not convertible into gold.

How then was the definition of the rentenmark as 1⁄2790 kg pure gold to be enforced in the marketplace? It's a complex but interesting mechanism. Four percent of the value of all German private agricultural land and industrial property was to be mortgaged, these "forced" mortages handed over to the Rentenbank as capital. The name given to these mortages was Rentenbriefe. Rentenbriefe represented a first lien on property owners, and each bond paid a 6% coupon stated in gold marks ie. interest and principle payments were indexed to gold. Furthermore, all rentenmarks were convertible into rentenbriefe at a rate of 500 to 1. Rentenbriefe, it should be noted, were not convertible into gold.

To sum up this unusual structure, circulating rentenmarks were to be convertible into rentenbriefe, which in turn represented a quantity of German land and property expressed in gold ounces. Rentenmarks, it seems, were a land-backed currency.

Horace Greely Hjalmar Schact, who many of us have read about in the excellent book Lords of Finance, was appointed to run the new Rentenbank. The notes debuted on November 16 and were immediately embraced by the German populace, trading for goods and services at their defined gold value. In his book The Economics of Inflation, Constantino Bresciani Turroni called this "miracle of the Rentenmark".  A currency defined in terms of gold with no actual gold-backing and only land-backing had been successfully floated. The solidity of the rentenmark need only be remarked upon by the fact that by Jan 1924, only 1,600 rentenmarks had been presented to the bank for conversion into rentenbriefe.

The stabilization of the paper mark

The German state now had two units of account, each with a related media of exchange. The rentenmark, defined in terms of gold, was stable, but the paper mark continued to hyperinflate, even after the rentenmark's debut. On November 13 one dollar bought 3.9 trillion paper marks, 6.7 trillion marks on the November 17, and 13 trillion by the end of the month.

To help stabilize the paper mark, that November the government was officially banned from funding itself through the Reichsbank, restoring to the central bank a much-needed degree of independence from the government.

At the same time, the Reichbank changed its commercial lending policy. It had been incredibly profitable to borrow from the Reichsbank during the hyperinflation because its lending rate was kept artificially low. Businesses and speculators borrowed marks to buy stocks, goods, or dollars, and after the mark had lost much of its value, bought them back to cover their loans several weeks later. Even though the lending rate was eventually increased to 90% in September 1923, businesses and speculators were in no way dissuaded from borrowing from the Reichsbank since – at the speed at which depreciation was occurring – only a rate in the thousands would prevent them from profitably borrowing, shorting, and later buying back marks to repay their loan.

As part of its new lending policy, all new advances were now to be indexed to the value of the mark in forex markets. Thus, should a business borrow a million paper marks from the Reichsbank, and the mark depreciated vis-à-vis the dollar, come maturity three weeks later the business would be required to pay back the million marks plus some penalty for depreciation. This dramatically increased the cost of borrowing from the bank.

Lastly, the Reichsbank's policy of accepting privately-created "notgeld" at par was ended. During the hyperinflation, severe currency shortages had encouraged corporations, towns, and municipalities to issue private currency alternatives, or notgeld. The Reichsbank gave these currencies legitimacy by accepting them at the bank's clearinghouse at the same rate as all other paper marks. By ceasing to accept notgeld, these alternative media suddenly lost a large part of their moneyness.

On November 20, Schact formally defined the paper mark against the US dollar. One dollar was to be equal to 4.2 trillion marks. The market simply didn't believe the Reichsbank could defend this definition. A few days later the mark's value had fallen so that it required 13 trillion marks to buy one dollar. To return the paper mark to its defined value, Schact and the Reichsbank proceeded to pull the rug out from under the market. Schact can explain it best:

The speculators, however, did not believe that the Reichsbank would be able to hold this rate of exchange rate for any length of time, and bought dollar after dollar on time bargains at a much higher rate of exchange... This speculation was not only hostile to the country's economic interests, it was also stupid. In previous years such speculation had been carried on either with loans which the Reichsbank granted lavishly, or with emergency money which one printed oneself, and then exchanged for Reichsmarks.
Now, however, three things had happened. The emergency money had lost its value. It was no longer possible to exchange it for Reichsmarks. The loans formerly easily obtainable from the Reichsbank were no longer granted, and the Rentenmark could not be used abroad. For amongst the stipulations governing the issue of the Rentenmark, there was one which forbade the surrender of Rentenmarks to foreigners. For these reasons the speculators were unable to pay for the Dollars they had bought when payment became due. They were forced to sell the Dollars back, and the Reichsbank was not prepared to pay more than the official rate of 4.2 billion Marks to the Dollar. The speculators made considerable losses. A bare ten days later the rate of exchange of 4.2 billion fixed by the Reichsbank had re-established itself. (The Magic of Money)

To sum up, short sellers needed to buy back paper marks to cover their shorts, but there were no sellers except for the Reichsbank. According to Adam Fergusson in When Money Dies (pdf), speculators lost up to 60% of their capital and “took off for Paris and went to work on the franc.” The upshot is that by the end of November, just fifteen days after the initial reforms were floated, Germany had two stable units of account, the rentenmark and paper mark. 

And that, folks, is how you end a hyperinflation.


  1. Would it be correct to say that the Rentenmark was indirectly convertible into gold?

    1. Yes, I think it would. With gold linked mortgage bonds being the settlement media.

  2. Any idea why they pegged the paper mark to the dollar and not to the rentenmark?

    1. Schact disliked the rentenmark from the outset, he wanted a true gold backed currency. If the paper mark had been defined in terms of rentenmark and not the dollar, perhaps it would have hindered his ability to retire rentenmarks and remove them from circulation. That's my best guess.

    2. Schact: "On 20 November, 1923 the Reichsbank decided to maintain a rate of exchange of 4.2 billion Marks to the Dollar. This rate was chosen essentially for accounting reasons. The peace-time exchange rate had been 4.2 Marks to the Dollar. Thus it was now only necessary to remove the noughts from the 4.2 billion in order to achieve a simple conversion to the old gold Mark basis. By and large any other rate of exchange could have been chosen. The main thing was to fix the exchange rate."

  3. An excellent post, as usual!

    I'm not a huge fan of the forced mortgages or the capital controls. Would have rather they simply raise nominal rates however high they needed to be to choke off the inflation and I think it would have worked just as well.

    Indexing the central bank loans to a foreign currency (or a CPI index, or the price of anything that's not hyperinflating) is a fantastic idea in this situation: you get to keep your optical interest rates at a reasonable level whereas in fact you could be in the thousands of % nominally..

    1. Thanks. Yes, it seems to me they could have stopped the hyperinflation without creating a rentenmark, just fix the existing paper mark using the techniques they would eventually adopt.

      I haven't gone into the politics the situation but my guess is that a dual pronged approach was necessary for political reasons. Rudolf Havenstein was the head of the Reichsbank and was not willing to tighten credit. Nor could he be removed as he was appointed for life. Havenstein denied the idea that inflation was due to loose credit, so they had to build around him via the Rentenbank. Coincidentally, Havenstein died on November 20, the same day that the new definition of the paper mark was adopted.

    2. Sounds like that guy was really in the way of keeping the currency stable.

      Curious what you think about the impact of the WW1 victors demanding repayment on the hyperinflation?

      Do you view it as a spark that could have been contained had interest rates been kept super high to discourage domestic consumption, thereby directing a large fraction of the nation's real output to the creditors?

      That's my view, but I feel much less informed than on you the specifics of the situation.

    3. I don't know the history behind the reparations very well but theoretically I can see them having a very significant influence on the hyperinflation. Reparations forced the government to rely on the Reichsbank for funding and this would have been inflationary.

  4. You have just taken over the position of BMBOT (Best Monetary Blogger Out There). Let's just hope that the blog doesn't get ruined by popularity.

    Is it 4.2 trillion or 4.2 billion?

    Thomas Sargent, in "The Ends of Four Big Inflations" noted that in every case, inflation ended after fiscal reforms but before money growth was reduced. In other words, hyperinflations are ended by a restoration of backing, and not by curtailing money growth. Greg Mankiw's textbook (5/e p. 369) prints the charts that clearly show inflation ending before the reduction of money growth. He then proceeds to observe only that "money and the price level move closely together...consistent with the quantity theory of money". One of my favorite examples of the misuse of data.

    1. Mike, actually money growth continues to rise because now money demand, which "dissapeared" during the hyperinflation comes back with force.

    2. Thanks Mike.

      It's 4.2 trillion.

      For anyone who is curious, here is the Sargent paper Mike is talking about.

      On the empirical side, there are a lot of variables here to tease out. Although the growth in paper marks and rentenmarks continued to explode, notgeld were simultaneously being canceled and the use of foreign exchange reduced. I'm not saying backing isn't important... it is. So too is instituting a new definition for the unit of account and enforcing that definition. Schacht must have had a fascinating time, standing at the middle of all this.

  5. You said:
    "What is so unusual about the end to the German hyperinflation was its suddenness. Within days of a series of monetary reforms implemented in mid November 1923, price rises came to a dead stop".
    That´s NOT unusual. It was how ALL hyperinflations ended (see Sargent: The End of 4 Big Inflations). It´s also true for the "modern" type:

    1. Fair enough. What I find interesting/odd is how a phenomena that has built up for so long can suddenly just stop. With a click up the fingers. Compare this with other economic phenomena which are slow and sticky... they cannot be turned on a dime.

    2. My interpretation: When you get into "fantasyland" (hyperinflation") a "click of the fingers" "wakes you up" (back into the real world)

  6. In economics, the one thing that does turn on a dime is price: The price of commodities, stocks, bonds, real estate, even labor. Price is about the last thing one would expect to be sticky.

  7. However, his name was Schacht, not Schact.

  8. - Garbage. There was 1 and only 1 reason why Hyper-Inflation stopped. The german central bank (literally) stopped printing banknotes. And then the Hyper-Inflation was over within 2 weeks. Nothing more, nothing less.

    - And the name of the president of the german central banks was Schacht, not Schact.