Energy Prices and Feedback Loops

As I am writing this (6/23/2011), energy prices are falling dramatically. This has been coming for some time now. There were a number of signs that this day would come. But you would never have known this based on the way the recent run-up in commodity prices has been covered in the media.

Let's start by considering an obvious relationship between energy prices and economic growth illustrated in the chart below (click to enlarge).

Everyone knows this. We witnessed this first hand in early 2008 and again this year. Higher energy prices affect both supply and demand. As for demand, higher energy prices reduce the purchasing power of consumers in terms of their real (inflation adjusted) income, diminishing their ability to purchase goods and services. This cuts into discretionary spending, spending we don't really have to do, but like to do. Industries that thrive on discretionary income are then hurt (ex: restaurants). On the supply side, higher energy costs increase input costs to businesses, raising their overall costs, potentially hurting their profits. How much will business costs rise? That depends on how much a particular business or industry uses the now-more-expensive energy inputs.

As this was occurring, media pundits did something they almost always do, attempted to predict the future entirely in terms of what was happening at that time. The basis for their forecasts were therefore predicated on the use of rulers, not economic theory or underlying economic signals. The result was forecasts like the one below (click to enlarge):

So, for those naive enough to rely on the media, panic naturally ensued. Would energy prices ever fall? How would we deal with the onslaught of permanently higher energy prices?

Of course all of this overlooked a very important saying about energy prices that has more than passed the test of time: "The best cure for higher energy prices is higher energy prices." In other words, you must take into account not only the entire set of shorter-term effects associated with rising energy prices outlined above but the longer-term impacts as well.

There are several effects of weakening demand. First and foremost, less demand hurts business activity overall, making it more difficult for firms to hire workers, possibly causing some layoffs. In other words, firms will tend to purchase fewer inputs overall than if activity were stronger. And as they cut back on purchases, their suppliers are also dragged into the weakness as this process continues (economists refer to this as the multiplier). Second, with diminished demand, even though costs are higher (as shown above) the ability of firms to pass along these higher costs is reduced, intensifying the overall weakness as profits weaken. These are important macroeconomic effects. Over the longer-term, the production of energy tends to rise, further helping to limit energy price increases.

Combining all of these factors, what occurs is a "feedback loop," where slower (and slowing) economic growth hurts the demand for energy, moderating or reversing earlier energy price increases. This is illustrated below (click to enlarge):

This is what we are beginning to witness now. The Federal Reserve and a host of other forecasters have been adjusting their growth projections for 2011 lower, largely the result of their presumption of continued higher energy prices. So, contrary to being able to forecast energy prices using a ruler, it is necessary to consider all of the effects I have outlined above. What will energy prices look like in the future when we correctly allow for this feedback loop? The chart below shows this (click to enlarge):

What all of this shows is that energy prices and the level of economic activity are simultaneously determined. There is not a one-way-only causation. I wish it were, that would make economic forecasting much easier!

From this, it should now be fairly easy to grasp the fact that lower energy prices are not necessarily an unmixed blessing. It is necessary to first consider why energy prices are falling. Since the emergence of OPEC in the early 1970s, energy prices often end up declining as a result of economic weakness they cause to occur. This does not mean that economic activity is about to explode. It will improve, but only after imploding for a bit longer (multipliers taking things lower for a while). Reiterating that last point, while higher energy prices don't instantly lower economic activity (it takes time for the effects of everyone paying the extra money to be felt), lower energy prices won't instantly restore the losses in economic activity. All of this takes time. That's why lags are so important in forecasting and economic analysis.

EPILOGUE: I created all the graphics for this blog post during my office hours in April, when I predicted to my classes on numerous occasions that we were near a short-run top in energy prices and nearing some sort of stock market correction. What was my basis for these predictions?

I used intermarket analysis, which is based on the fact that key asset markets (stocks, bonds, currencies, and commodities) are all linked together in today's global economy (actually I have created a course in this that I was teaching this spring).  I showed my classes that commodity prices were indeed rising (we all knew that). But, as commodity prices were rising, the inflation "alarm bell," interest rates, were actually declining not rising as higher expected inflation would suggest. Along with this, one commodity that is particularly sensitive to economic growth, copper, was also declining. Finally, the defensive sectors of the stock market (consumer staples, health care, and utilities) were outperforming the overall stock market. So, if future commodity prices were best depicted by a "ruler-based forecast" this intermarket pattern would either not have emerged, or it would have been temporary at best. The rest, as they say, is history.

I didn't make this post in April because, quite honestly, everyone would have thought that I was crazy. Gee that would be something new!