Guest Post : Stretch To Farthest Point Known - Thoughts on a Hyperinflation Event

Stretch To Farthest Point Known:  Thoughts on a Hyperinflation Event


January 15, 2010

Thinking about Extremes

Let’s assume for a moment that Goldman Sachs is wrong.  After all, at most points in time and space, predictions tend to fail—except the lucky ones.  So it’s good to think through scenarios that one would consider extremely remote.  Active risk management means low probability / high catastrophic outcome tail events must be hedged, and as importantly, gain exposure to those pesky Blacks Swans in ways that lead to advantage.  To accomplish this, it helps to obtain a quantitative sense of their impact, to get a “feel for the cloth” as an wise former boss of mine used to say.  So let’s try here.
Hyperinflation and Currency Crisis

What if the Fed more than succeeds in reflating and the end result is hyperinflation?  As remote a possibility as I think this is, they really could print a way to another, completely different type of economic destruction.  All they have to do is print proactively, not reactively. 

Hyperinflation is rare, but not inconceivable.  It happened on multiple occasions in the last century alone.  It is indicative of extreme government failure as the state fails to perform a most basic function:  monopoly provision of acceptable currency.    

Above all, hyperinflation shows that mankind holds closely some constants that will endure, even though the world around us never stops changing.  Even in hyperinflation life doesn’t lose all of its familiar contours. 

Forget Argentina 2002.
Imagine a country carrying a crippling amount of external debt.  Its core financial intermediary system remains is on life support at best, at worst it is an explicit arm of government policy.  The central government and its affiliates control a large portfolio of residential and commercial real estate, and further subsidize the housing market though administratively controlled interest rates.
Banks with large loan books and scale are iteratively recapitalized as losses are realized, and are wards of the state, with increased political control over their operations.  Smaller banking entities increasingly do not exist through liquidation or absorption into special purpose vehicle that allocate credit on an uneconomic basis to specialized industries. 

Industries with “strategic value” (read: political connections) are completely nationalized, with taxpayer-forced liability to cover the cash burn.  Further, these socialized businesses compete with private businesses with smaller economies of scale that do not have the luxury of taxpayer funded advertising budgets, or a creditor that doesn’t care about funding loss-making enterprises through tax revenues.

To cover the increasing losses of all these commitments—in particular government losses on assets related to the mortgage market—the central government resorts to the printing press in real force.  A market forms to price in capital and exchange controls.  The currency is no longer used, and people practice currency substitution, the use of another nation’s currency as a store of value and in some cases, the medium of exchange.  The combined mass exodus from the currency and hyper-monetized national debt, causes the inevitable happens:  hyperinflation. 

Is this the story of the United States in 2010 or Poland in 1989?  

There’s a lot of difference, of course.   The United States doesn’t have exchange rate and capital controls (yet).  The degree of nationalization isn’t as endemic to our economy.  There is a thin veneer glossing over certain aspects of our political culture.  But there are some strong similarities as well.  
Importantly, the Poles realized there was something unique and historic about their macroeconomics of de-control/disintegration.  To record knowledge of their fate—and to preserve the jobs of as many bureaucrats as possible—the government statistical service preserved the minutest details on a monthly basis. 

The data source is various issues of the Biuletyn Statystyczny, the Polish household budgetary survey that consisted of a rotating 8,000 households in Poland.  Just so you know, through 1990 it excluded self-employed as well as household of persons involved in military or police service, so there was an underreporting of self-employed entrepreneurs.  

Reading the Bones

The Polish hyperinflation was a short-lived event but in the space of a year inflation rocketed up 300%.  After the huge acceleration shock, price increases only decelerate, but they don’t decline.  Note the chart below:  there were only seven months of real hyperinflationary explosion.  After that, prices stubbornly increased, but there were no seismic events.   

The data suggests that modern currency collapses don’t mean dollars or zlotys or any other bearer bond with no coupon go the way of the dodo completely, even in hyperinflation.  What you get instead is currency reform—a “new and improved” bearer bond with no coupon in exchange for the old one (an extreme haircut at redemption), and high real and nominal interest rates.  If there isn’t currency reform that toilet paper can hang around as a unit of account for some time.

Exchange Rates and Capital Controls

Shorting the currency would seem a no-brainer here.   However, there’s a high probability of a snag:  capital controls and “official” exchange rates—currency inconvertibility.  Such measure would screw up straightforward and profitable currency shorting, but the data shows they don’t work to dampen inflation.  People are cunning and they develop work-arounds to a dying currency.  Even if citizens must accept said currency as legal tender, they develop efficient ways to off-load it using market mechanisms.  Note price discovery on the black market for zlotys (Polish currency) before and after the exchange rate finds a hard peg, meaning interest rates start to bite. 

Cost of Living

This is where the “minute detail” comes in.  That household survey captured consumer prices by category during hyperinflation.  They show that hyperinflation is the ultimate in living for the now.  As prices for basic necessities go through the roof, the prices of non-essentials collapse.  Not only is capital stored in currency destroyed, the cost for food outstrips other consumer categories.  For the record, non-foodstuffs means clothing and shoes and electrical/mechanical goods for the home; entertainment means  newspaper expenditures, books (including school-books), movies and related items like concerts;  fuel (heating oil and gasoline) is not included.  Subsidies and usage differences make such comparisons inadequate anyway because few had a car when Polish communism collapsed.
There are some clear and profitable conclusions to this section.  Food (and fuel) is how one will profit in the initial stages of a hyperinflation, even as rents, non-food consumer goods collapse.  This is due to the fixed nature of rental arrangements and the immiserizing effect of inflation takes luxury goods out of reach.  Items for immediate consumption are the one that really jack up.  After the initial inflation shock, returns on other goods surpass foodstuffs. 

The biggest defect of the comparison is the exclusion of petrochemical usage, as it was a heavily subsidized and very differently utilized commodity in communist and transition Poland.  I believe it is safe to assume that gasoline and distillates in general would be most sensitive to any type of inflation shock. 

Inflation Equations

However, you could long inflation risk through US CPI futures settled in another currency OTC.  So let’s assume that one is associated with Taleb’s Universa Investments and want an explicit hedge to a hyperinflation event.  Consider the following inflation option-like product:  a non-stream cap security with a strike at X% inflation.  Think of such a cap as a call option on the inflation rate implied by CPI.  I know, I know, it is not exactly an option in that CPI isn’t a tradable product. 

Using some modified RBS methodology [fn]The Royal Bank of Scotland, 2003, Guide to Inflation-Linked Products. Risk[/fn]or here for… less exotic… inflation caps, the pay-out is:

Where N is the notional, κ is the strike, omega is 1 for a cap (-1 for a floor) and ψ is the contract year fraction for the interval Ti-1 to Ti; retrofitting with the data earlier shows that quarterly fractioning is efficient.  I(T) is inflation at time T, measured by CPI.  In the case of a non-stream cap like here, Ti-1 resolves to T0.

For pricing, assume away counterparty, institutional, rounding, and seasonality risk.  The pricing formula is rather complicated, similar to a cliquet option.  

Formula aside, I’m not sure what kind of quote you would get from a derivatives broker for a hyperinflation cap.  For options on inflation one generally has to rely on quotes from brokers like ICAP