Focus
March 18, 2022 | 13:42
Inflation: Where Next?
Economics
Economics
Economics
Fixed Income Strategy
Economics
Economics
Economics
Economics
Economics
Economics
Fixed Income Strategy
Economics
Economics
Economics
Inflation: Where Next? |
|
“We have an expectation that inflation will peak and begin to come down this year. To the extent inflation comes in higher or is more persistently high… we would be prepared to move more aggressively by raising the federal funds rate by more than 25 basis points at a meeting or meetings.” — Fed Chair Powell, March 2, 2022 |
High inflation poses a major threat to the economic expansion, perhaps the largest since the 1970s. In the U.S., it’s hard to keep track of the many forces driving inflation to 40-year peaks—from surging fuel and food costs to rising wages and house prices, to name a few (Chart 1). The Russia-Ukraine war has only turned up the heat by juicing resource prices and snarling supply chains further. To wade through the fog, we employ our in-house equation for U.S. consumer prices to assess what the major forces mean for the outlook. Details on the equation are in the appendix. Spoiler alert: while inflation will likely retreat, it should also stay high enough to make the Fed sweat. Chart 2 shows the simulation results. The equation predicts inflation of around 6% currently, or about 2 ppts below the official rate. The large undershoot—about five times the standard error—could imply less underlying price pressure than commonly believed. Then again, like any simple model, ours doesn’t capture every driver of inflation. For example, it can’t account for a behavioral shift that boosted goods prices during the pandemic, or for the effect of extra household savings on the ability to pay higher prices. It likely misses the recent increased ability of businesses to pass input costs to consumers, as the equation is estimated mostly in a period when business pricing power was held in check by globalization. Nor does it properly capture base effects stemming from price weakness in early 2021. The CPI is up a less heated 4.7% annualized from two years ago, though it’s also sprinting at a near 9% rate in the past five months, so it’s clearly gained momentum. |
|
Before using our equation to forecast inflation, we need to outline paths for the explanatory variables. BMO Economics expects commodity prices to peak soon and fall 30% by the end of 2023, with oil prices averaging $100 this year and $85 next year. The dollar is expected to rise only slightly further in response to tighter Fed policy. Rising interest rates should corral house prices from the recent blistering pace of 19% to a more sustainable 3% rate in late 2023. Based on our slack measure, the U.S. labour market has never been tighter in the past quarter century, and we see some modest further pressure on wages as the unemployment rate slips to new half-century lows later this year. Perhaps the hardest assumption to make relates to global supply chains. As shown in Chart 3, supply pressures have eased a bit from record highs, but the Russia-Ukraine war and partial lockdowns in China could reverse this progress. We expect some backup in the index in the near term, before supply chains return to normal in late 2023 as global COVID restrictions end. |
Using these assumptions for the explanatory variables, a simulation of the equation starting in March 2022 (i.e., using the current 7.9% rate as the starting point) predicts the annual CPI rate will peak at below 9% soon, before falling to around 5% late this year and then to 3% late next year (Chart 4). While the peak of this base-case forecast lines up with Economics’ official call, the equation sees inflation trending about one percentage point below the department’s year-end forecast of near 6%. Recall, however, that it may be understating some price pressures. Importantly, the equation predicts that inflation will settle at around 3% late next year, or modestly above Economics’ current forecast. That extra persistence is due to the tight labour market. The equation tells us that a long list of things need to go right for inflation to fall even to 3% by the end of 2023. What if our assumptions for the explanatory variables are too sanguine? For example, what if commodity prices (and oil in particular) don’t fall? In this case, the equation expects inflation will run just over one percentage point above the base-case prediction by late this year, or a little over 6%. If house prices fail to slow, inflation could be one percentage point higher, and if supply chains fail to improve, inflation could be a quarter percentage point higher next year. In the event that all three shocks persist, inflation could be more than two percentage points above the base-case prediction by late 2023, or around 5%. And this assumes labour markets don’t tighten much further. |
|
Bottom Line: Our CPI equation predicts inflation will ease as commodity prices retreat, house prices cool, supply chains improve, the dollar firms, and the jobless rate steadies just below current levels. However, it also expects inflation to hover near 3% by late next year. This is largely in line with the Federal Open Market Committee’s median forecast for the consumer expenditures deflator (2.7% y/y in 2023Q4), likely spurring it to stay the course of a fairly aggressive tightening plan. |
Appendix: U.S. Inflation EquationThe equation regresses yearly percentage changes in the monthly CPI on similar changes in commodity prices (CRB spot index), the trade-weighted U.S. dollar, and national home prices (Case-Shiller), as well as on the level of pressures stemming from supply chains and labour markets. The New York Fed’s Global Supply Chain Pressure Index begins in 1997 and incorporates data on freight rates as well as manufacturing delivery times, order backlogs, and inventories. Our in-house index of slack in U.S. labour markets is derived from a comparison of current and historical levels of 16 variables. The equation is estimated from November 1997 to February 2022, and explains 93% of the variation in inflation over this period. The standard error of 0.4 ppts compares favourably against a sizeable standard deviation of inflation (1.4 ppts) over the estimation period. The explanatory variables are significant at the 3% level or better. The constant (of 1.9) is also significant, likely reflecting the anchoring effect of the Fed’s 2% target on price expectations. Commodity prices enter the equation contemporaneously, while the dollar and slack are lagged one month, supply chain pressures two months, and house prices six months. |