Archive for February, 2012

Hamilton on the Fed and oil prices

28/Feb/2012 Comments off

James Hamilton has a new post reviewing the case for a Fed driven commodity bubble.  I must say that, while Hamilton knows more about macroeconomics and econometrics than I ever will, I think he is wrong here.

Behold this quote:

In general, I agree that the Fed needs to keep a watchful eye on commodity prices as one potential indicator that their efforts to stimulate the U.S. economy are having undesirable side effects. But based on the evidence so far, I think the Wall Street Journal is crying wolf a little too early.

The implciation here seems to be that if oil rose to…say $130, Hamilton would favor tightening. I don’t wish to put words in his mouth but that is how it sounds to me. Maybe I am so blinded by the NGDP targeting story that I overestimate the amount of slack in the economy, the size of the output gap, the flatness of the SRAS curve. It seems to me though, that with inflation expectations well below target, and the economy missing something like 6 or 8 million jobs, that there is broad scope for boosting AD. Assuming this is the case, I find myself wondering What is the full employment oil price? I suspect it is somewhere between $180 and $210 per barrel on the mid point between WTI and brent.

Hamilton must be taking it as given that either A) The economy is near full employment and catch up NGDP growth would mostly weaken the dollar, not boost employment and RGDP. or B) Monetary policy can lift specific prices through some financial mechanism, not through expectations of higher real demand for oil. I suspect and hope the answer is A and only A.

I recall the months just after QEII, when oil was rising with the jobs reports. I also recall that conservatives and Austrian cranks ranted and raved at this time about how QEII was lifting oil prices. No Way! I thought adding 400k jobs in one month would be neutral for oil! Again, the question is rarely asked, “What are full employment commodity prices?” Oil was $140 a barrel the last time the world economy was close to full employment. Since then, China is 20odd % richer, plus a slew of smaller emerging economies in Asia and Latin America have made good progress. Why, even the African economies are expanding through shear population growth. All while world oil output treads water.

There is definitely an at least medium term Malthusian aspects to our world economy. Incremental demand for inelastic goods with slowly responding supply will cause a relative price shift. Oil will be dearer in terms of labor and labor intensive goods. How could it be otherwise? How could the price of oil, the supply of which is more or less fixed in the short run, not rise in terms of other, more readily made goods and services?

The question is, do we want to reach our potential? or do we want artificially cheap commodities through tight money and high unemployment? I think I know the sad answer to that question and suspect we wont see a full recovery until enough shale oil is flowing from America to keep the CPI down in the context of a true recovery.

Categories: Monetary Policy, Oil

Lorenzo on the impending Greek collapse

Hyperbolic American talking heads like to use the string of words “complete” or “total collapse” when discussing markets. When I hear total collapse, I think Easter Island, or Detroit. The point being that “collapse” invokes feelings of, well, collapse. One goes from a high level of organization and development, and then something happens and suddenly things aren’t so organized. Thus I cringe at the word’s overuse. With this in mind, it seems to me that collapse is not unfitting in the context of Greece.

Lorenzo has a great post which I wish I had written: Not merely a Greek disaster. The post also links to a Daily Mail article with some gloomy pictures of the works which the Euro hath wrought. My working conclusion is now that whatever hormetic effect the Greek economy might have drawn from the “medicine” of austerity, has long since passed into the realm of poison. The real threat now is that is that the 1500 year running Greek braindrain is completed, and the country never able to truly recover.  I usually don’t think much good can come from overseas meddling, but at some point Obama should probably tell the Europeans that they are destroying Greece and make a plan to get the Drachma back.

One funny thing about this whole Euro Debt Crisis episode is that the Utopian machinations of the European social democratic consensus have killed the European welfare state. I can’t drive this point home enough, as I so value irony. France in 1995 wasn’t good enough. No, we must have an unbreakable currency union and to hell with whether or not it works.

By the time this austerity business is over, the only remaining welfare states in Europe may well be Sweden, led by a reasonable, market thinking center right government, and ubercapitalist Switzerland.  (As always, Norway is an oil fueled, super Potemkin village.) Lest I be accused of ideological prejudice, Sweden was kept out of the Euro under the Social Democrats who, looking back, were the ultimate conservatives of the 1990s/2000s.

Categories: EMU, Uncategorized

Christensen on devaluation

Lars Christensen has a good post on the workings of currency devaluation. The point is that currency devaluation works through lifting M and V more than through softening the blow to exports, or improving competitiveness.  This would suggest that there is unlimited scope for devaluation in a crisis, so long as NGDP is forecast to be under trend. Which in turn implies that Poland is the great monetary hero of 2008/2009, because they temporarily annihilated their currency to keep NGDP going. [I say unlimited scope because “beggar thy neighbor” export boosting should quickly run into diminishing returns, especially in a worldwide crisis. ]

It would be interesting for someone to compare the composition of GDP between Sweden, who devalued meaningfully in early 2009, and Denmark/Finland who were chained to the Euro. If Sweden’s stronger recovery was driven by domestic spending—as memory serves— not exports, this would help us to better understand the transmission mechanism at work there.  It would also be interesting to dig into Poland’s numbers, though I don’t know which land is a good control…Lithuania? Each country in central Europe is pretty economically and monetarily unique. Perhaps just graphing net exports along with domestic spending would do it.


Maybe tonight.

A supplyside recovery?

The U.S. economy is clearly recovering at a faster rate.  Despite GDP growth coming in around 4% in recent reports, hiring is starting to look quite strong and joblessness is falling.

Lest I be misunderstood, the recovery is still disappointing. My point is just that meaningful progress is being made for the first time since that stretch between August 2010 and say March 2011, just after QEII and just before the Japan nuclear mess.

The oddest thing about today’s spike in growth is that stocks are rising, but longterm interest rates are flat:

Every school child knows that long rates track NGDP expectations. Things like flight from the Italian bond market might be leading investors into U.S. treasuries, but that should be a one time premium, after which yields again track stocks. Why the disconnect between shares and bonds?

One story might be, that “despite” the Fed’s bizarre commitment to keep interest rates low for two years (to new readers, low rates do not mean easy money by themselves), markets are still expecting weak NGDP growth, and thus staying in bonds.  This would suggest that the rally in shares and spike in hiring is due to some mix of positive supply shocks and adjustment in wages, prices and contracts to weaker nominal GDP growth. One way to test this is to look at inflation expectations, which should rise if NGDP forecasts are climbing:

Looks like the markets are not impressed by the Fed’s new “inflation target” or its commitment on interest rates. Inflation expectations are stuck around 1.3% out to 10 years, right where they have been since the summer. This makes me more confident that the Market Monetarist model is right and that the requickened 1970s “Keynesians” who say that the Fed should push down longterm interest rates are wrong. Long term interest rates are down and staying down on recent Fed fiddling. However, this is not stimulating aggregate demand because expected inflation is unmoved.

This means that:

  1. We need Greg Mankiw and Ken Rogoff at the Fed
  2. The economy seems to be adjusting to the nominal shock of 2008/2009 at a faster rate
  3. The window to speed the recovery through QE or level targeting is likely closing at a faster rate

It is hard to know how much of the recovery is due to supply side boons like the weirdly warm American winter or the rise in oil production v.s.  adjustment to the new NGDP path.

If I had to say which is the bigger factor I would go with wage and contract adjustment. American households have been reorganizing their lives since the Fall of 2008. Rebuilding their balance sheets and consumption patterns to  best work in a world where the Federal Reserve is a big fat nominal liar, only giving out 1.5% wage hikes instead of the 4% increases it dealt from 1982 to 2007.  You can see this in the higher savings rate, the increase in building of apartments relative to houses, in leasing rather than buying, in delayed marriages. All these are effective strategies for living in a world of uncertain nominal income growth. In my own life I maintain a near-Chinese level savings rate (if you count student loan payments in the “savings” box) but will at some point be tempted by those falling IKEA prices and address my underfurnished abode. Nominal adjustment.

Another possibility is that the newly reemployed are working harder to avoid layoffs in the next downturn, and to get their lives back on track through faster advancement.  This would fall under supply shock and suggest that real GDP is more trend reverting. Just speculation.

Estimating market NGDP expectations

One thing which bothers me is that nobody estimates NGDP expectations! This is the be-all-end-all of macroeconomics (or should be) and is of utmost importance to producing good forecasts, yet no one estimates them for any country. At least as far as I know.

In princpal it should be possible to extract the market’s implied NGDP forecast from asset prices. The Cleveland fed does this for CPI inflation expectaions, why not for NGDP? When things happen in the economy which change the expected path of NGDP, markets move. We should be able to reverse the chain of causality and turn market movements into an NGDP forecast. Not just any forecast, a market forecast. Even if the forecast turns out to be wrong—all forecast are wrong in the end—knowing what the emergent being we call the market is “thinking” will actually give us insight into how unemployment and output are likely to behave in the near term because the individuals who make up the market act on their individual nominal income forecasts in a fairly predictable way.

Seeing as no one else is interested in estimating this series for me, I am setting out to do it myself. At first I thought it might be possible to estimate a New-Kenyesian Philips curve on the GDP deflator (an equation which maps an “output gap”, inflation expectations and past inflation to current inflation) and then rearrange the parameters until I had NGDP as the dependent variable. Then I could model NGDP expectations using TIPS spreads or maybe the Cleveland Fed series. This would only build only bond market prices into the forecast though and would require a forecast of RGDP and the output gap, which is problematic for a number of technical reasons and itself assumes an NGDP forecast.

I think the way to go about it is to keep it simple. I propose using a straightforward VAR to generate medium run forecasts of NGDP conditional on a selection of obvious market prices.  Before showing you my simple VAR setup and the forecast themselves allow me to get out ahead of the econometric nerds. I know you can get all sorts of fancy with a VAR. I know there are disputes about estimating non cointegrating VARs on the level rather than first difference. I know I could probably have put my variables into log form too. Keep it simple. I’m halfassing it so that other people who don’t work 70 hours a week can come up with something better. If they don’t, I will in time.

The setup:

First go to the St.Louis FRED

Next, download quarterly time series of GDP, the S&P500, West Texas Intermediate oil prices, ten year rates, five year rates, six month rates. Shoot over to the Cleveland Fed (they really need to move to a new city) and download three year CPI inflation expectations. Load them into LibreOffice. Load them into R. So far all this costs is a few years of statistics classes.

Estimate a VAR on the above variables over the period 1982Q2 to 2011Q4.  1982Q2 because you forgot to change the default “1982-01-02” to “1982-01-01” at the download page and aren’t trying to publish. Use the AIC to find that a lag length of three is pretty good.  Next, solve the model for the period 2012Q1 to 2015Q4 and you get a forecast of NGDP conditional on the level of various asset prices (as well as the then unknown first GDP release) as of 2011’s close.

From what I remember of the end of last year, this is a completely reasonable forecast. At the end of the year things were getting better but for supply side reasons, especially all the oiling pouring out of those states Obama and his followers don’t like. Indicators of nominal activity like inflation expectations and interest rates were atanking.  I will wait to further refine the model before bothering with deeper look at this simple forecast. Maybe build some sort of model averaging system and fix the obvious problem that even the almost useless initial NGDP number for the quarter is not known by markets in working time. In the end though, it should be possible to use the old BEA press releases to insert the preliminary and advance GDP figures into the NGDP series at the appropriate points needed to represent the information available in the past. This could allow someone to generate an NGDP version of the Cleveland Fed CPI expectations release. That is, produce average growth rate expectations over different yearly intervals. This could be helpful in formalizing just exactly what markets were thinking in 2008 and 2009 and help us to study the roll of market NGDP expectations on employment, investment and so on.

Now that I think about it I wish I had included a dollar index, thus the need for model averaging.

Categories: Uncategorized

Pure gold from Hamilton

12/Feb/2012 Comments off

From Econbrowser:

One of the problems with the gold standard is that when the real value of gold changes (as it does all the time) and the dollar price of an ounce of gold is fixed (as it must be by definition under a gold standard), that means dollar prices have to adjust in response to anything that happens to the gold market. With the economic and financial turbulence of the late 1920s and early 1930s, there was a big increase in the relative price of gold.

For example, to get an ounce of gold in 1929, a farmer would need to deliver a little over a hundred pounds of cotton. By 1932, it would take more than three times as much cotton to get that same ounce of gold. Whereas 18 bushels of wheat would be enough to buy an ounce of gold in 1929, you would have needed more than twice as much wheat to get gold in 1932. And since the price of gold in terms of dollars was fixed between 1929 and 1932, that means you’d need to produce about three times as many pounds of cotton or two times as many bushels of wheat in order to earn one dollar in 1932 as you would have needed to earn one dollar in 1929.

It is hard to know where to being when debating with gold standard people. Do you go after the ABCT? Do you instead try to show them that an inflation indexed bond spread target would be better? Or just point to the unworkability of a deflationary monetary system? Certainly the policy framework was much  more progrowth in the 1920s, and that era just happened to coincide with a gold standard. However, if we live in a world where even U.S. monetary policy doesn’t work well for a place like Hong Kong, or German monetary policy doesn’t work well for Denmark, why should we forwait that a completely random monetary policy should work for America?

This is what a gold standard would be. The price of money set by happenstance. Money is kind of important, I insist its value not be set by the number of weddings in India (Hindu weddings are big on gold jewelry), and  political stability of Mongolia or Congo (from where marginal gold output comes). Gold was a great medium of account in the premodern world but it would not work today.  That individuals and companies do not write contracts indexed to the gold price is evidence of this. The monetary system should be built to fit the world, not the world to the monetary system. Level target NGDP and ready the gold kooks for the knacker’s yard.

[A final note, the price of money is not the interest rate. The price of money is the GDP deflator, or if you like, your wages, or the price of unsalted grass fed butter. Money has many nominal prices but they are all for goods and services. Interest rates are the price of credit, as Sumner reminds us.]

Categories: Uncategorized

AS Incarnate

8/Feb/2012 Comments off

From Bloomberg

Crude production in the U.S. is already increasing. Within three years, domestic output could reach 7 million barrels a day, the highest in 20 years, said Andy Lipow, president of Lipow Oil Associates in Houston, a consulting firm. The U.S. produced 5.9 million barrels of crude oil a day in December, while consuming 18.5 million barrels of petroleum products, according to the Energy Department.

North Dakota — the center of the so-called tight-oil transformation — is now the fourth largest oil-producing state, behind Texas, Alaska and California.

The growth in oil and gas output means the U.S. will overtake Russia as the world’s largest energy producer in the next eight years, said Jamie Webster, senior manager for the markets and country strategy group at PFC Energy, a Washington- based consultant.

I’ve been long the dollar for about three months, after learning there was a field equal to about 20% of Saudi Arabia deep under the land along the Ohio/Pennsylvannia border. I have wanted to blog on this but haven’t made the time. Maybe meager posts are the way to go. Anyway, the U.S. is under rather different conditions than the U.K. To spell it out, we have 1. clearly better immigrants  2. A captive upper middle class (only Vancouver can compete with our yuppie utopias, Brits flee to Oz, North America) 3. Oh…and rising oil output.

No 3. might be enough to explain the rising jobs numbers despite slow NGDP growth. Households should eventually adjust to the new 4% per year NGDP growth path (after they pay off their 7% interest loans with 1% yearly pay hikes, when they were implicitly promised 3.5% hikes by the Fed from 1985 to 2007), but you wouldn’t expect to see such a rapid surge in hiring. We will have to see if the recovery continues to gain speed but the key might be shale oil. Ohio, Texas, Montana, Wyoming. There are sites spread throughout the country that need workers, steel, trucks and all sorts of equipment.The problem may be nominal, but the—second best—solution is real.

I’ll dig into the numbers later but the U.S. is clearly at the beginnings of a positive supply shock.  The most hopeful thing about rising oil production is that it will keep a lid on that pesky energy component of the CPI, which talkingheads, Austrians and dishonest antiObama conservatives brought up during the period just after QEII when the recovery gained speed.  The problem with the U.S. recovery has been that, at roughly full world employment, oil is about $140-$200 per barrel, judging from the price we saw mid summer 2008. That means a temporary bout of CPI inflation, and much gnashing of teeth at the petrol station if the economy should recovery quickly. It also means that if the Fed does its job, it will needfully drive up oil prices. Giving this guy ammo. With rising oil supplies, we might just get the sustained CPI weakness needed for enough political cover to print some permanent base money. Plus all that oil is real wealth.

Categories: Oil