The surprises and complaints surrounding the Bank of Canada’s (BoC) monetary policy decisions earlier this year largely boil down to one word: oil. Both the rate cut announced in January 2015 and the decision not to change rates in April 2015 can be attributed to the volatility of the oil price and its impact on the Canadian economy. One has to look no further than the “risks” section of the last two Monetary Policy Reports (MPRs) to see what Governor Poloz and his team were thinking. Oil went from being a secondary, two-sided risk to the inflation forecast in the January 2015 MPR to the number one downside risk in the April 2015 MPR.[1] Subsequently, Governor Poloz has spoken at length about the price of oil during public appearances and Parliamentary testimony.
This article sets out how oil prices affect the Canadian economy; whether oil prices are forecastable; and how the BoC factors oil price uncertainty into its monetary policy decisions. It concludes by explaining which oil prices to watch and how oil prices may influence the path of interest rates.
How important is oil for the Canadian economy? Very important. In a December 2014 speech, Deputy Governor Timothy Lane noted that oil extraction accounts for about 3 percent of Canada’s GDP and crude oil about 14 percent of Canada exports.[2] My Ivey colleagues Adam Fremeth and Guy Holburn put the figure higher at around 7 percent of GDP – double oil’s importance vs. a decade and a half earlier.[3] By 2012 Canada was the 5th largest producer of crude oil producer and dry natural gas. Canadian oil exports accounted for more than 30% of U.S. oil imports in 2013 – the single largest source of imports.
How do lower oil prices affect the Canadian economy? There are numerous effects, some positive and some negative, which are handily summarized in an Appendix to the April 2015 MRP (p.25-26). First, lower oil prices reduce total CPI due to lower gasoline, fuel oil, and transportation costs. But the BoC ignores this effect by focusing on core CPI, which excludes volatile energy prices. Lower gas and utility bills, however, give households more disposable income, encouraging consumption. Second, oil producers earn lower revenues, pay lower taxes and royalties to governments, lay off workers, and suspend or cancel capital expenditures (i.e. investments). House prices in places like Alberta likely fall due to lower incomes and higher unemployment. Stock prices of big oil producers decline, reducing the wealth of households. Third, net exports to the U.S. fall, even though the Canadian dollar (“CAD”) cushions some of this effect by depreciating against the U.S. dollar. Economists speak of a deterioration in the “terms of trade”, which means Canada exports less relative to our imports. A decline in the terms of trade makes Canadians feel poorer, reducing aggregate demand and GDP.
There are some positive effects, as companies that use oil as an input benefit from the lower cost (such as airlines or utilities). The fall in the CAD may also boost other exports. And most importantly, a fall in the price of oil may boost growth in oil-importing countries, increasing global demand.
There are two important caveats. First, the net effect of an oil price decline depends on whether the change in prices is temporary or permanent. If oil prices rebound quickly, then the impact may be muted. Second, any forecasts of the impact assume that “other things remain unchanged”, including monetary policy. One obvious response is to ease monetary policy to counteract any negative effects. But other risks and shocks may materialize, throwing any economic forecast off course.
Given their importance, how does the BoC forecast oil prices? Unfortunately oil prices cannot be forecast with any accuracy. A large body of research – much of it conducted by the Bank of Canada – concludes that:
“while futures prices tend to produce forecasts that are correct on average, such forecasts are also highly volatile relative to no-change forecasts. Therefore, futures-based forecasts may be very inaccurate at a given point in time.”[4]
Given the conclusion that the oil price is a random variable, the BoC assumes oil prices will remain near their recent levels over the projection horizon. In other words, the price used in the BoC’s projection model is the average price observed since the last MPR. While this assumption may seem overly simplistic, it turns out that even oil-industry experts do not feel they can forecast oil prices.
How does this uncertainty affect monetary policy? As any executive knows, the future is uncertain but business decisions still need to be made. The key is to take this uncertainty into account when evaluating your options. The BoC explicitly checks the sensitivity of its economic forecast to various factors or “risks” that it does not control. For example, the Appendix to the January 2015 MPR evaluated what would happen if oil prices were to fall relative to the base case scenario:
“Lower oil prices reduce overall demand in the Canadian economy, leaving output lower by 1 to 1.8 per cent and putting downward pressure on core inflation”.
The implication is that if oil prices fall, there is less need to raise interest rates; in fact the case can be made to lower them further. As it turns out, oil prices did fall between January and April 2015, and the BoC decided not to change its overnight rate target in April. Unfortunately, it appears Bay street economists did not read the Appendix to the January MPR, which explains why so many of them were surprised by this decision.
What is an observer of monetary policy supposed to do? The answer is to monitor oil prices carefully and to understand how they will affect the output gap and inflation. Investors need to monitor three crude oil benchmarks: Brent for global crude oil, West Texas Intermediate (WTI) for light crude oil, and Western Canada Select (WCS) for heavy crude oil. The BoC estimates that Canada’s oil production by value is 10 percent Brent, 55 percent WTI, and 35 percent WCS. While Brent is the most important for inflation, WTI and WCS are the main benchmarks for Western Canadian producers, and by extension industry profits, net exports, and GDP growth.
Below Figure 1 and Table 1 chart these benchmarks and provide descriptive statistics. First, notice that the average price per barrel of Brent is the highest and WCS is the lowest. Differences in the U.S. dollar prices across benchmarks reflect supply and demand dynamics in different regions, and appear to be persistent. Second, changes in WCS are correlated above 90% with WTI, but below 90% with Brent. So WTI is the better benchmark. Third, WCS consistently trades at a discount to WTI of around $17. This discount reflects US restrictions on the export of US oil that boost the WTI price, and limited refining capacity for heavier oil plus transportation constraints that reduce the price of WCS.
Figure 1: Average Monthly Prices for Crude Oil (US dollars per barrel) (click to enlarge)
Table 1: Descriptive Statistics on Crude Oil Prices (click to enlarge)
In summary, the oil price is an important but random variable influencing Canada’s economic forecast and the future path of interest rates. The BoC does not forecast the oil price, but assumes the price over the forecast horizon will be the average price since the last MPR. If oil prices rise, the output gap will close more quickly leading to greater inflationary pressures, which will lead rates to rise sooner. But if oil prices decline, the opposite effect occurs. These forecasts are subject to considerable uncertainty, with the BoC signaling the degree of uncertainty in the “Risks” section of the MPR. Make sure to read this document all the way to the end.
Not even oil producers know what will happen to oil prices. So the best an observer can do is to understand how oil affects the economy, monitor the relevant benchmarks, and then reach a conclusion of what the likely path for interest rates will be. Just like Governor Poloz.
Michael R. King is a Professor at the Ivey Business School, Western University
[1] See: Bank of Canada Monetary Policy Report, January 2015, p.23; Bank of Canada Monetary Policy Report, April 2015, p.29.
[2] T. Lane, “Drilling Down – Understanding Oil Prices and Their Economic Impact”, Remarks to the Madison International Trade Association, Madison, Wisconsin, 13 January 2014.
[3] A. Fremeth and G. Holburn, “Developing Global Market Access for Canada’s Oil and Gas Industry”, Ivey Energy Policy and Management Centre – Policy Brief, November 2014. See: http://www.ivey.uwo.ca/energycentre/research/policy-briefs/
[4] R. Alquist and E. Arbatli, “Crude Oil Futures: A Crystal Ball?”, Bank of Canada Review (Spring 2010), p.8. See also: C. Baumeister, “The Art and Science of Forecasting the Real Price of Oil”, Bank of Canada Review (Spring 2014), p.21-31.