Monday, February 22, 2016

Connecting the Dots: the Demand Side of Oil's Decline

Why have oil prices declined so sharply since mid-2014? For many observers the answer is obvious: there has been a surge in global oil production and it has pushed down oil prices. It is hard to argue oil supply has not been important, but the slowdown in emerging economies and in the United States also suggests that global demand is playing an important role too. If so, this understanding raises several questions: First, how much of the decline in oil prices since mid-2014 can be attributed to weakened global demand? Second, why did global demand begin falling in mid-2014? 

On the first question, Stephen King of HSBC believes that weakened global demand is a key reason for the recent decline in oil prices. Here is what he said to CNBC:
"You have a situation where emerging markets in general are extremely weak, that in turn is causing commodity prices to decline rapidly, including oil prices, so rather than saying lower oil prices are a stimulus for the commodity consuming parts of the world, I think you should see lower oil prices as a symptom of weakness in global demand," 
Jens Pedersen, an economist with Danske Bank, shares this view and provides following figure as evidence:
  

This figure is certainly suggestive of a link between global demand and oil prices, but exactly how much does this link explain?

Ben Bernanke had a recent post that attempted to answer this question. In it, he presented evidence from an estimated model for the demand of oil. The model builds upon the work of James Hamilton who estimated a regression where the demand for oil is determined by changes in copper prices, the 10-year treasury yield, and the dollar. Here is Bernanke's explanation for why this model approximates the demand for oil:
The premise is that commodity prices, long-term interest rates, and the dollar are likely to respond to investors’ perceptions of global and US demand, and not so much to changes in oil supply. For example, when a change in the price of oil is accompanied by a similar change in the price of copper, this method concludes that both are responding primarily to a common global demand factor. While this decomposition is not perfect, it seems reasonable to a first approximation.
Using this approach, both Bernanke and Hamilton find that 40%-45% of the decline in oil prices since mid-2014 can be attributed to weakening global demand. I reestimated their model, but added the BAA minus the 10-year treasury spread as another indicator of demand conditions. The idea behind its use is that weakening demand increases the credit risk of firms and therefore causes this spread to rise. Using this specification of the model, I was able to attribute 51% of decline in oil prices to weakening global demand. This can be seen in the figure below:


My estimates, which are similar to Bernanke's and Hamilton's, suggest that about half of the decline in oil prices since mid-2014 is due to weakening global demand. So yes, global demand does seem to be an important part of story behind the descent of oil prices.

As an aside, these results help shed some light on a puzzle with "breakevens"-- the spread between nominal treasury yields and real treasury yields from TIPs. Breakevens are supposed to capture the bond market's view of inflation (though it is sullied a bit by a liquidity premium for TIPS). Many observers have often noticed that these breakevens track oil prices and therefore conclude that either oil prices are driving inflation or that the breakevens really do not tell us much about expected inflation.  


I have always had a problem with this understanding. It ignores the possibility that something else might be driving both breakevens and oil prices. Moreover, breakevens also tracks the BAA-treasury spread mentioned above.


It always seemed to me that breakevens, oil prices, and the credit spreads may be responding in part to something else. The results above suggest that the something else is global demand. 

Anyways, this still leaves us with the second question: why did global demand and therefore oil prices began to tank starting around mid-2014? Readers of this blog should already know my answer. Beginning in mid-2014 the Fed started talking up interest rate hikes and continued to do so through 2015. This signalling that future monetary policy would be tightened got priced into the market and affected decisions well before the December 2015 rate hike. 

The talking up of interest rate hikes, therefore, amounted to an effective tightening of monetary policy for the United States and all the countries that peg their currency to the dollar.It explains the sharp rise of the dollar as shown below:


This tightening of monetary policy got ahead of the U.S. recovery and precipitated a slowdown in U.S. economic activity as I document here. It also put the noose around China's neck--who pegs to the dollar--as I show here. It is no surprise then that the oil prices began their dramatic fall after mid-2014. 


There is a rich irony to all of this. One of the reasons Fed officials talked up interest rate hikes is because they believed inflation was about to take off. They believed the only thing preventing it from happening sooner was the low oil prices which they saw as transitory. What they missed is that their own actions were a key reason for the decline in oil prices in the first place. Fed officials failed to connect the dots between talking up interest rate hikes, weakening global demand, and low oil prices. More and more, it seems Fed made a huge mistake in 2014-2015

P.S. Andres Ariza Meneses makes the same argument here.

21 comments:

  1. These last few posts have been interesting. What I don't get with the thesis of appreciable recent tightening, however, is why it woulnd't be showing up in realized NGDP since 2014, which is sticking very closely to the straight-line path fitted over 2010-2013 (either log space or the raw number).

    ReplyDelete
    Replies
    1. On reflection, I'm guessing your reply would be that the problem is Emerging Market pegs, not US monetary policy locally.

      Delete
    2. Great question Eric. The U.S. is not in a recession, but is slowing down. And in my view, it is tied to this tightening. The slowdown can be seen in the growth of NGDP. Since 2014Q3 has been trending down.See https://research.stlouisfed.org/fred2/graph/?g=3xFO So yes, the U.S. economy is not suffering as much as the dollar bloc countries from this tightening.

      Delete
    3. It is slowing down? Really? I don't see it. Stop following government bean counters. They revise persistently for decades. The 80's got revised down before it got revised up for example.

      Delete
  2. It's hard to tell from growth rate plots, since we're really interested in a level-based target. For instance, from the plots you linked, you would then have to say that 2012Q1-2013Q2 saw an even greater tightening than the more recent episode, but with no trend in Oil prices over the former period. If you go back and look at a plot of the level (log NGDP) for the US, it's really hard to make the case that there's anything measurably tight about the 2014-15 period; see:

    https://drive.google.com/file/d/0B_VIvX5tmpwCd1hEakFObVN0cEU/view?usp=sharing

    ReplyDelete
    Replies
    1. Eric, in terms of a level target yes, it is tough to know exactly where we are on some trend. First, we have to be sure we are using the right trend. I've actually attempted to do estimate where NGDP should be using various methods. It is part of a paper that hopefully one day will be done. I put the various measures up on website here: http://davidbeckworth.wix.com/ngdp-watch Second, even if we are on trend it is still the case the growth of NGDP has been slowing down. It it continues then at some point it will fall below its optimal path.

      Delete
  3. It´s the fallacy of reasoning from a price change!
    https://thefaintofheart.wordpress.com/2015/12/11/reasoning-from-a-price-change-will-set-you-on-the-wrong-path/

    ReplyDelete
  4. It´s the fallacy of reasoning from a price change!
    https://thefaintofheart.wordpress.com/2015/12/11/reasoning-from-a-price-change-will-set-you-on-the-wrong-path/

    ReplyDelete
  5. Is the world economy as fragile as the feelings of today's college students who insist that the campus must be a "safe place" where "everybody knows your name"? It must be according to Professor Beckworth; after all, we aren't talking about a Paul Volcker level rate increase here. I would also point out that a few of Professor Beckworth's colleagues at the Mercatus Center are of the view that it isn't the Fed's job to support asset prices, and that falling asset prices have a silver lining. Blaming the Fed for the financial crisis when the mother of all asset bubbles burst and for the current slowdown in a recovery that's lasted almost eight years has come to define Beckworth's career. Somebody has to take the blame.

    ReplyDelete
    Replies
    1. Rayward, I hate to break it to you but this is a non-linear world. The economy today is very differnt than the one in the early 1980s with different relationships among economic variables. I fear you're fighting the last war.

      Also, I would encourage you to see my book where I do layout the Fed's role in the housing boom. http://www.independent.org/store/book.asp?id=100

      Delete
    2. You and I actually agree on the fragility of the economy, although I'm not totally convinced that the Fed is primarily responsible. My view is that we have come to expect too much of the Fed, putting both the economy and the Fed's credibility at risk. And it's my view that the growth and size of the financial sector are what's at the heart of the fragility. Now what's responsible for the growth and size of the financial sector is another, and the most important, issue. But as long as the financial sector looms so large, we are at the mercy of small changes in monetary aggregates having large consequences to the economy. Keep up the good work.

      Delete
  6. On the other hand, global oil consumption rose by something like 1.6 - 1.8 MMbbl/day in 2015. Maybe that was all an elastic price response and the demand curve really did shift to the left. Or maybe the fact that supply growth still outpaced consumption growth suggests the rise of Saudi and Iraqi production, along with the modest decline in U.S. output, means this is mostly a supply situation.

    ReplyDelete
    Replies
    1. The Orioles Way, observed oil consumption is a misleading way to assess demand on several levels. First, the observed amount is a confluence of supply and demand conditions. Second, one also needs to know what consumption would have been given a healthier economy and compare that to actual consumption to know the demand shock.

      Delete
    2. Oil consumption grew in 2015, but at half the price of 2014.

      I think the easiest way to understand supply-demand dynamics in oil market is to assume vertical short run supply curve down to a price around usd7-10/bbl, which is roughly MC for most producers. Most fluctuations in oil price is due global ngdp shocks or precautionary demand shocks on back of geopolitical risk. If you want to observe true supply shock have a look at 1985-86 following the Saudi decision to raise output by 2mb/d in Oct 1985. This is exactly the event which proves David's analysis is correct.

      Delete
    3. It got cut in half because China stopped stocking. David's analysis completely misses this. I mean, completely.

      China drove a artificial bubble in oil prices. Nothing more or less. Production overamped globally in response.

      Delete
    4. Anonymous, but if you follow David's links to his China post there he shows the Fed's tightening was key to China's slowdown. That is why China stopped stocking.

      Delete
    5. Sorry, but China's stocking was the problem. If the Fed stopped their stocking, GOOD FOR THEM FINALLY!!!! The surge in oil prices from 2010-2014 was artificially driven any way you can describe it.

      David is a mess. His points are a mess. He doesn't want to admit the obvious.

      Delete
    6. You're missing the point. The IMF, Bernanke, Hamilton, Stephen King, and others say weak global demand is causing oil prices to fall. (David's only novel twist, I think, is that he attempts to trace the weak global demand to Fed policy.) Acting like China's stocking is some isolated event that operates in a vacuum independent of these developments is not very convincing.

      Delete
  7. Thanks so much for the link to my post! Btw, It would be nice to investigate whether the Fed is or has been the "worlds central bank". I dont want to infer from this post, so I ask you directly: Do you think the Fed is powerful enough as to destabilize global aggregate demand (nominal spending) and cause a global resesion given the current current link with China?

    ReplyDelete
    Replies
    1. Andres, I do think it is able and have coauthored an article awhile back on it. Here is the link: http://people.wku.edu/david.beckworth/savinglut.pdf This idea was further developed by Colin Gray in this article: https://ideas.repec.org/a/kap/atlecj/v41y2013i2p173-184.html How much power it has varies, though,depending on how many countries are pegging to it.

      Delete
  8. Where can I get my hands on a copy of Federal Funds Future rate data? Or does this have to be derived from a historical series? If so, please give me a shove in the right direction!

    Thanks!

    ReplyDelete