Cheap Oil, the Dollar, and Deflation

I want to shift gears a little today and open a discussion on some of the wider economic implications of a lower oil price.  The motivation behind this post is a discussion I had with a close friend, a former investment banker who currently works as a corporate strategist for an international oil company.  I mentioned to him how I believed that a lower oil price would almost certainly have deflationary, or at least have disinflationary consequences.  My friend, who is an inflationist, countered with the counterintuitive argument that if lower energy prices stimulate spending in other areas, that spending could result in demand pull inflation.  At the time I wasn't in a position to argue, but his point was not lost on me, so I went home to examine whatever data I could find on the subject.

Correlation between oil price and deflation in past oil price spikes and crashes:

Last week's oil price crash was not unprecedented, and examining past oil price crashes is likely to provide some historical insight.  Recent oil price drops occurred in 1981, 1985, 1997, 2000, and of course 2008.  The price drops of 1997, 2000, and 2008 can be attributed to demand shocks associated with the Asian financial crisis, bursting of the technology bubble, and global financial crisis respectively, whereas the 1985 oil price crash was due to a supply glut after Saudi Arabia flooded the global oil market.  Finally, the 1981 price crash was a combination of oversupply and the early 1980's recession triggered by major interest rate hikes specifically targeting inflation.

In all of these cases lower oil prices led to lower inflation, and even deflation, in the following years as seen in the figures below:

Long term real oil price (www.macrotrends.net)

Long term United States CPI (www.inflation.eu)

When looking at these figures, it's extremely important to acknowledge that in cases where the oil price drops were brought about by recessions it is impossible to attribute disinflationary or deflationary effects to the fall in the oil price.  This leaves the oil price crash of 1985 as the only comparable precedent.  Interestingly, it can be observed in this case that disinflationary effects were short lived.

Disinflationary mechanisms associated with a lower oil price

There are a number of fundamental reasons that we should expect a lower oil price to lead to deflation, at least in the short term.  Chief among those is the impact of the price of oil on the US dollar, as a strong US dollar reduces the price of imported goods and services.  In spite of the recent resurgence in US oil production, the US is still a significant petroleum importer.  As such, the US runs a large petroleum trade deficit, which is US dollar negative.  However, as shown in the following figures, this deficit is shrinking and will continue to shrink, and perhaps even turn to a surplus, due to both lower oil prices and increased US oil production.  This is immensely bullish for the US dollar.

Furthermore, energy, and particularly oil, is an important input to cost of many goods and services.  If the price of oil remains low, we should expect to see some cost savings work it's way back to the US consumer.  In fact, we are already seeing the impact on prices as illustrated by the CPI, and my preferred method of measuring inflation, the BPP index (Billion prices project).  The BPP is perhaps the broadest and most unbiased measure of inflation available, as it measures prices by scraping the internet for prices from hundreds of online retailers.

BPP Daily Index (bpp.mit.edu)

BPP Monthly Inflation (bpp.mit.edu)

Demand Pull Inflation

The crux of my friends argument was that lower oil prices would stimulate spending in other areas, leading to demand pull inflation.  Based on the 1985 precedent, where inflation rebounded quickly after initially falling, the data may actually be on his side.  The arguments against demand pull inflation revolve around consumer deleveraging and economic slack.  If consumers choose to pay back debt instead of spend their petroleum savings, we would not expect demand pull inflation.  Furthermore, if there is enough economic slack to absorb additional demand, the demand pull inflation argument also becomes less compelling.

Measures of household debt appear to be returning to early 2000 levels seen in the chart below.  Although household leverage appears to be stabilizing, there is still some room for leverage levels to decrease further.

Household Debt as a % of Disposable Income (Seeking Alpha via Economics Fanatic)

On the other hand, indicators of economic slack appear to be in line with long term historical levels.  Most of the slack created during the 2008 recession appears to have already been absorbed as shown below:

Indicators of Economic Slack (Wall Street Journal)

With data on household deleveraging mostly inconclusive and data on economic slack showing historically normal levels of slack, it appears that demand pull inflation may in fact be a possibility if consumers spend their petroleum savings.

Conclusions and Investment Implications

It appears my friend has a compelling case for demand pull inflation in the medium-long term. However, in the short term, the US dollar is likely to continue to strengthen and long dated US treasuries may outperform as they will benefit from the current deflationary environment.  That deflationary environment may be short lived due the possibility of "demand pull" inflation, where US consumers spend their petroleum savings on other goods and services.  If this is the case, US treasuries may not be such a good investment in the longer term. It is important therefore to keep a close eye on indicators such as the US trade deficit, household debt, and measures of economic slack.  Personally, I find it difficult to take a position here either way, as timing bond and currency markets is not my strong suit.