Slouching towards equilibrium

  

The Bank of Japan is struggling to ignite inflation.  “Central bank officials,” reports the Wall Street Journal, “have been putting a brave face on the slowdown in inflation resulting from the lower oil prices, saying that cheaper crude will stimulate demand, eventually adding to upward pressure on prices.”

If that quote is accurate, and if it refers to oil demand, then the Bank of Japan really is in trouble.  It’s confusing a market equilibrium with the lack of one.

In principle, the oil market tends toward a price that equates the quantity supplied to the quantity demanded.  If supply exceeds demand, then sellers will cut their price until they work off the excess.  Once they do so – that is, once supply again equals demand – the price will stabilize, lower than before.  It won’t rise in response to excess demand, because there is no excess demand.  The market is in equilibrium. 

Central Asian policymakers make this mistake, too.  The point is that the price doesn’t just change on a dime.  It responds to shocks – like international sanctions choking demand – that throw the market out of kilter.  Once it has brought the market back to equilibrium, it will stay put, until the next shock.

Maybe the Bank of Japan is referring to demand for all products, not just petroleum.  When oil prices fall, Japanese consumers save money.  They will spend much of the savings on a variety of goods, raising their prices.  That’s inflation.

Moral of the story:  The Journal should sharpen its blue pencils and edit for clarity.  –Leon Taylor tayloralmaty@gmail.com

Reference

Tatsuo Ito and Kosaku Narioka.  Japan inflation slows in blow to Abe.  Wall Street Journal.  December 25, 2014.

What a coincidence

 Last week, oil prices rose on the short-term futures market when traders cleared their positions, reported the Wall Street Journal.   A word of explanation: To profit from a falling price, a trader can borrow an asset and sell it at today’s high price, then purchase it at next month’s low price and return it to the lender.  Her profit is the price difference, minus transaction costs.  If, for example, the asset price falls from $100 to $60, then she clears $40.  If many traders clear their positions at the same time, then demand for the asset will rise, increasing its price.  So is this adjustment somehow mechanical, with no implications for the asset’s perceived value?

No.  Traders will choose to close their deals precisely when they expect the price to rise.  If I think that the price will be $60 on Wednesday and $65 on Thursday, then I will buy back the asset on Wednesday, profiting by $40 rather than $35.  Last week’s increase in price was not some mysterious happenstance; it reflected the belief of traders that oil prices are going north.  The prophecy fulfilled itself.

So, are oil prices finally rising for good?  Not necessarily.  The daily futures price is volatile, prone to temporary changes in what traders expect the market to do.  A more stable measure is the annual spot price.  (The “spot” market is where oil is bought and sold; the “futures” market is where traders speculate on future spot prices).  But this measure is backward-looking, since it is dominated by prices that occurred several months ago.  A reasonable compromise may be the quarterly spot price, which is a fifth lower now than it was at this time last year.  For September through November of this year, the European Brent spot price for a barrel was $88; for the same period in 2013, the price was $109, according to data from the United States Energy Information Administration.  In contrast, the annual spot price for Brent has fallen 12.4% -- $103 for December 2013 through November 2014, and $118 for the same period in 2012-3. 

In short, oil prices are still falling.  But the completion of short sales last week reminds us that the prices need not continue to decline for, say, two years.  Expectations may change and indeed may be changing.  –Leon Taylor tayloralmaty@gmail.com

Notes

I calculated the quarterly and annual prices as averages of monthly FOB prices.

References

Nicole Friedman.  U.S. oil prices bounce off multiyear lows.  Wall Street Journal.  December 17, 2014.

United States Energy Information Administration.  Data on prices and quantities of energy fuels.  www.eia.gov.

  

           

The Riyadh connection

Why did OPEC's decision rattle futures markets? 

In the last six months, during which oil prices on the futures market declined, they plunged most steeply immediately after the November 27 decision of the Organization of Petroleum-Exporting Countries (OPEC) not to reduce its supply, perhaps because of a prisoner’s dilemma. Brent prices fell by $6 to $71.25 per barrel, reported Reuters.  Why did speculators respond to OPEC by dumping oil securities?

The most reasonable explanation would be that they believed that the decision would create an excess supply of oil.  Either supply would rise or demand would fall.  Let’s consider each possibility.

Could OPEC’s decision have increased supply?  It’s hard to see how.  Non-OPEC producers (like Russia and Kazakhstan) would not respond by increasing their own supply since, given demand, this would create an excess.  They might have stepped up production had OPEC voted to cut back, for this might have freed up some demand for themselves.  Indeed, this very possibility might have led OPEC to maintain its existing supply.

Surely, then, OPEC’s decision would lead to a fall in demand.  But this doesn’t make sense, either.  Why would buyers cut back in response to a confirmation of the usual supply?

Seeing is disbelieving

In short, it is not reasonable to think that OPEC’s policy could increase excess supply and thus cut oil prices.  So why did investors immediately respond by selling oil short?

Perhaps they considered not the actual effect of the decision but the perceived effect.  If the typical investor believes, correctly or otherwise, that the policy would increase supply, then it makes sense to get out of oil.  With apologies to Robert Shiller, one might call this an “irrational lack of exuberance.”

The weakness of this approach is that an investor who focuses on actual demand and supply will eventually profit at the expense of short sellers, if market conditions dominate prices over the course of a year or more.

For example, some analyses suggest that global oil prices are falling because of an (undocumented) excess supply of one million barrels per day in West Texas Intermediate oil.  Even if the excess supply does exist, it would be trivial in the context of a global market of more than 90 million barrels per day, according to data from the United States Energy Information Administration.  It is not reasonable to think that, given demand, a rise in supply of about 1% would lead to a fall in price of 30% or 40%.  But if investors think that such fallacies are commonly accepted, then they may sell oil short for profit, at least until a sense of reality prevails.  –Leon Taylor tayloralmaty@gmail.com   

Notes

1.  In the game called “prisoner’s dilemma,” each player chooses a strategy that most benefits him given whatever the other player would do – yet the overall outcome is worse for both players than a cooperative solution would be.  In OPEC’s decision, each country might prefer to maintain its own output regardless of whether other countries would maintain or reduce theirs.  But the general outcome might be worse for OPEC than a commitment by all its members to reduce output.

 References

Alex Lawler, David Sheppard and Rania El Gamal.  Saudis block OPEC output cut, sending oil price plunging.  Reuters.  November 27, 2014.

Robert Tuttle.  Brent, WTI slump to 4-year low as OPEC keeps quota steady.  Bloomberg.  November 28, 2014.

United States Energy Information Administration.  Copious oil statistics.  www.eia.gov

     

Cliffhanger

For two weeks, news reports have predicted off-the-cliff falls in the global price of crude oil on the spot market, based on publicized forecasts.  What the newspapers don’t note is that many forecasts are based simply on past data.  They assume that any new data trends will continue, regardless of the reasons (if any) for those trends.  

And rarely do the newspapers report anything more than the point forecast (say, $55); they ignore the range of likely prices (e.g., $30 to $80 – the “confidence interval”) which is especially large for long-run forecasts.  By itself, the point estimate is not of much use because it is rarely exactly correct.  The question then becomes: What values near the point estimate may well occur?  To answer that, we need the confidence interval.     

Finally, most news reports are based on authority, not logic.  In their view, Organization X's forecast is important because X is important, not because its forecast is well-grounded.  

Let’s go back to basics.  The long-run price of any product is an adjustment to market changes.  The price of X will fall if otherwise the market would remain in excess supply.  In particular, the long-run global price of crude oil will fall if either supply has increased or demand has decreased, recently and permanently.  The problem is to determine what might have precipitated either event.  The reason for the last price collapse, in 2009, was clearly global recession.  What reasons stand out today? 

Supply.  World oil output is rising but not dramatically.  It increased by 1.7% from 2013 to 2014 for the period January through August, the latest data available from the United States Energy Information Administration.  It is not realistic to think that a 1.7% increase in supply can lead to a 30% or 40% decrease in price. 

Demand.  The Chinese and European economies have slowed, but those trends are not new.  The Chinese growth rate has been relatively soft for two or three years, according to World Bank data.  The European economy has been sagging for at least five years. 

K-traders

The problem is to explain why short-term oil prices have fallen 30% since June.  The only apparent new event of the past few months has been the tightening sanctions against Russia. It is hard to believe that these could account for a 30% fall in global oil prices.

Another possibility is that oil contracts take time to rewrite, so the recent price decline reflects market conditions of the past year or two.  But since various contracts come up for renewal at various times, this might not explain the suddenness of the price decline.   

Finally, some news reports suggest that prices have fallen because of a confluence of random factors.  In that case, the factors, being random, will disappear.  They should not affect long-run prices.

So, why have crude prices fallen so abruptly?  One reason may be speculation.  Many traders believe that the market price is determined not by supply and demand but by the average beliefs of traders.  I will call them “Keynesian” traders, because John Maynard Keynes analyzed this strategy in 1936, in his General theory of employment, interest and money.   

Keynesian traders regard an initial fall in price, which may occur randomly, as evidence that other traders believe that prices will continue to fall; so they sell the product short.  The short sales themselves reduce prices on the futures market, making the Keynesian strategy a self-fulfilling prophecy.  This may help explain the disproportionate fall in oil prices two weeks ago in response to OPEC's decision not to reduce output.  

Whether you think that the price decline due to shorting is a short-run phenomenon or a long-run one depends on the importance that you attach to supply and demand.  If you believe that long-run prices can fall without fundamental changes in the market, then you may accept the forecasts of $55 oil.  The important point is that a six-month collapse of oil prices is not by itself strong evidence that the decline will continue for years.  –Leon Taylor, tayloralmaty@gmail.com               

Kazakhstani income and oil prices (wonkish)

In Kazakhstan, a 10% change in the annual spot price of Brent oil seems to lead to roughly a 5% change in average real income in the same direction. 

Since oil and natural gas exports comprise more than a third of Kazakhstan’s economy, a simple model of global oil prices well explains changes in gross domestic product per capita, which is measured in international dollars adjusted for inflation. 

The table below reports the results of an Ordinary Least Squares (OLS) regression of real GDP per capita on the spot price of Brent oil (a benchmark for the global market) for 1999 through 2013.  Both variables are annual and in log form. 

In general, the model performs fairly well.  R-squared indicates that the model explains 92.5% of the variation in average income from year to year.  The large F statistic (159.52) suggests that the model almost certainly predicts more accurately than one that assumes that GDP per capita is constant over time.  The root mean squared error suggests that the model’s average annual mistake in predicting GDP (as measured by this statistic) is 8.5%.     

The t statistic for the coefficient on OilPrice is large (12.63).  Setting aside shocks to the world economy that are unexpected and extraordinary, we can be virtually certain that global oil prices will continue to dominate Kazakhstan’s economy in the next few years at least. 

The coefficient on OilPrice is the elasticity of average real income with respect to the oil price.  A 1% increase in that price leads to a rise in income in the same year of .46 of a percent.

The constant in the model (7.845) suggests that 2,553 international dollars of annual real income do not depend on oil prices.  In particular, if oil prices fall to $1 per barrel, then the model predicts an income per capita of 2,553 dollars, about one-seventh of actual income averaged over the period of 1999 through 2013. 

Output for the OLS model

      Source |       SS       df       MS                 Number of obs =      15

-------------+------------------------------          F(1, 13) =  159.52

       Model |  1.147         1        1.147               Prob > F      =  0.000

    Residual |   .093        13         .007               R-squared     =    .925

-------------+------------------------------           

       Total   |  1.240       14         .089                Root MSE      =  .085

------------------------------------------------------------------------------

     GDP      |      Coef.   Std. Err.          t      P>|t|     [95% Conf. Interval]

-------------+----------------------------------------------------------------

  OilPrice    |       .456         .036    12.63   0.000           .378      .534

   Constant  |     7.845         .145    54.28   0.000         7.533    8.158

The estimated model is:

LN (GDP per capita) = 7.845 + .456 * LN (OilPrice)

where LN denotes a natural log.

Caveats.  A plain-vanilla OLS model assumes that the strength of the relationship between the dependent variable and an independent variable is the same whether the latter rises or falls.  Thus our model predicts that if oil prices rise 10%, then income will rise about 5%; and if oil prices fall 10%, income will fall 5%.  In reality, annual oil prices have fallen sharply only once since 1998; and income has fallen only once in that period.  Both declines occurred in 2009, during the Great Recession.  At that time, oil prices fell 36.3%; income, only 1.4%, at least partly because the government stepped up spending to cover the loss of private consumption.  So experience suggests that the model may overstate the loss of income due to a large decline in oil prices.

The overstatement occurs because OLS is a linear model and because it assigns the same weight to each observation.  Of the 15 observations, only one is of a decline in average income.  –Leon Taylor, tayloralmaty@gmail.com 

Technical notes

OLS assumes that the independent variables (which are on the right-hand side) capture all important systematic influences on the dependent variable (on the left-hand side).  Unimportant or random influences show up in the error term, which is the difference between the actual value of the dependent variable and the predicted value.  

The assumption of independent errors might often be wrong.  The error term might correlate with the independent variables since these might influence its variance (heteroskedasticity). Or the error term may correlate with itself; e(t), for example, might correlate with e(t-1) (serial correlation).  Let’s check our model for these possibilities.

Heteroskedasticity.  I ran the Breusch-Pagan/Cook-Weisberg test.  Its null hypothesis is homoskedasticity (that is, the variance of the error is the same for each observation, which is what OLS assumes).  The probability of homoskedasticity, given estimated parameters, is 11.9%, so I did not reject that possibility.

Serial correlation.  I ran the Breusch-Godfrey (LM) test.  Its null hypothesis is no serial correlation.  The probability that the null is correct, given parameter estimates, is .364, so I did not reject it.

Nonstationarity.  A model that changes over time may predict poorly because it is based on obsolete data.  Dickey-Fuller tests found that the GDP variable was stationary but the oil-price variable was not.  (The test statistics were -4.24 and -1.35 respectively.  Both variables were in log form.)

When rewritten, nonstationary variables may have a stable – that is, stationary – relationship to one another.  In that case, the revision may predict well.  To check for this possibility of cointegration, I ran the Dickey-Fuller test on the error term of a model regressing GDP on oil prices.  The test statistic was -2.81.  Given an indefinitely large sample, the critical value under the Engle-Granger procedure at the 10% level of significance is -3.04.  (The level of significance is the largest probability of error that one is willing to tolerate by rejecting the null hypothesis.  In this case, the null is nonstationarity.) 

Although I have a small sample, I concluded that the model may border on cointegration.  I preferred this model to one that first-differences the oil-price variable in order to get stationarity, since differencing would eliminate an observation.  But one should bear in mind that in the model I use, nonstationarity may affect forecasts.   

Omitted variables.  If we fail to control for a systematic influence with an independent variable, then it will show up in the error term.  If it also correlates with an independent variable, then it may bias the coefficient estimate for that variable.  This problem is more serious than are serial correlation and heteroskedasticity since these do not produce bias. 

I applied the Ramsey RESET test, which examines the possibility of omitted variables that are higher-order forms of the independent variables already in the model.  The null hypothesis is that there are no omitted variables.  The probability of the null is .72, so I did not agonize over the problem. 

Warning shot

The communist faction in Kazakhstan’s legislature proposes to require the central bank to give advance notice of a devaluation of the tenge so that people can prepare.

The communists are justly famous for bad ideas, but this has got to be one of their 10 worst.  Suppose that the National Bank announces, on January 1, that it will weaken the tenge to 200 per United States dollar on March 1.  Speculators will immediately cash in their tenge while they are worth something (roughly 180 to the dollar at present), selling tenge for dollars.  This will drain the Bank’s dollar reserves, which are its only means of protecting the value of the tenge in the last resort.  The Bank will have to devalue to 200 on January 2.  So much for advance notice.

The Majilis would be more sensible to require the Bank to issue, in plain language, monthly reports of the state of the economy.  –Leon Taylor tayloralmaty@gmail.com

Reference

Zarina Karymova.  Kommunisti predlagaut preduprezdat’ nacelenie o deval’vatsii zaranee.  [The communists propose warning the population about devaluation in advance.]  Panorama.  November 14, 2014. 

How much money do you want?

The question sounds odd.  Surely the answer is:  “All the money in the world.”  But economists are literal-minded:  To them, “money” refers to the number of tenge available for spending, and “money demand” is the number of tenge that we want to hold, not spend.  Suppose that you had 5,000 tenge in your purse this morning and spent 1,000 on lunch.  Then your money demand fell from 5,000 tenge to 4,000.  Of course, the demand depends mainly on how much you’d like to spend later.

Money demand can determine whether the central bank can navigate the economy to harbor.  Normally, during a recession, creating tenge will spur spending and thus production.  But if people are willing to hold any amount of money, then they won’t spend the new tenge.  Such a “liquidity trap” may have blocked recovery from a long depression in Japan and in other countries where households save an unusually large share of their income.  In any case, we’d like to know how much money people want.

 

That question is hard to answer.  We know what the answer should look like.  Your demand for money should increase with your income, since people who earn more now will want to spend more later; and with prices in general, since you must pay more to buy your usual groceries when prices rise.  Also, money demand should fall as interest rates increase, since you would rather hold bonds, which pay interest, than money, which doesn’t.  (Money in this sense is usually defined as cash and checkable accounts, or M1 money.)  Indeed this model works pretty well in the long run – but not in the short. 

 

The next-door solution

 

Kazakhstan is no exception to this conundrum.  A 2010 study finds that the demand for money (broadly defined) in this country depends in predictable ways on output (which generates income), the interest rate, and on foreign exchange rates (which affect global demand for Kazakhstani exports), in the long run.  But this sensible model breaks down in shorter periods.  The authors speculate that rising interest rates, for example, may have induced people to move their wealth from dollars to tenge.  Thus the relationship between interest rates and tenge demand may be positive, contradicting the usual money-demand model.

 

The difficulty of short-run predictions hinders monetary policy.  Four or five decades ago, macroeconomists led by Milton Friedman argued that the central bank could reduce uncertainty in the economy, and thus spur purchases and production, by targeting the money supply.  The bank should create no more money than is needed to buy new output at the usual prices.  In October 1979, the central bank of the United States, the Federal Reserve, adopted this policy.  But money demand suddenly skyrocketed, outstripping money supply.  Interest rates rose, spending fell, and a wintry recession set in.  By the late 1980s, the Fed had stopped targeting the money supply.  Since then, it has usually targeted interest rates.

 

Sweep stakes

 

New research suggests that we may be able to explain short-run money demand after all.  In 2012 Lawrence Ball of Johns Hopkins University studied American demand for M1 from 1959 to 1993.  He found that it depended less on general short-run interest rates such as the Treasury bill rate than on the rates paid on such close substitutes for M1 as savings accounts and mutual funds in the money market.  In the late 1970s, money holders were sensitive to rises in the interest rate on mutual funds because the Fed’s Regulation Q capped the interest rates paid for conventional bank accounts.    

Should central banks again target money?  Ball notes an intriguing new complication in measuring M1.  Most central banks, including the Fed, require commercial banks to set aside a given share of their demand deposits; otherwise, the amount of money created by new loans, in the form of checking accounts, may get out of hand.  Beginning in the early 1990s, the banks got around the Fed’s reserve requirement by shifting money from demand deposits to money-market accounts.  (This cut in demand deposits reduced the amount of money that the banks were prohibited from lending out.  Suppose that the reserve requirement ratio is 10%.  If a bank reduces its demand deposits from $2 million to $1 million by shifting a million into money-market accounts, then its reserve requirement will drop from $200,000 to $100,000.  Voila!  The bank now has another $100,000 to lend out.)  Since money-market accounts were not part of M1, the banks’ “sweep” campaigns artificially reduced that measure of money supply.  After 1993, M1 data became unreliable. 

In Kazakhstan, could sweeps help account for the stagnation of M1 over the past two years?  -- Leon Taylor, tayloralmaty@gmail.com

 

References

 

Lawrence Ball.  2012.  Short-run money demand.  Journal of Monetary Economics 59:  Pages 622-633.  Informative.   

Mesut Yilmaz, Yessengeli Oskenbayev, and Abdulla Kanat.  (2010). Demand for money in Kazakhstan: 2000-2007.  Romanian Journal of Economic Forecasting 13: Pages 118-129.