A few of us journalists got Bank of Canada Governor Stephen Poloz talking about forecasting over the weekend. The International Monetary Fund earlier had revised its outlook for global economic growth higher, instead of lower, which had been its habit for the previous six years. It seemed important, but Poloz’s enthusiasm was muted. “If you balance the risks in your forecast, you know there is a 50-50 chance just through normal data that you are going to have a positive one,” he said. “And so it’s not necessarily as significant as you make it sound.”
Poloz’s reluctance to fully embrace the shift in economic prospects has become a thing. Bay Street is cheering the rebound. “Although it is still early days and risks abound, signs are pointing to an economy that looks increasingly poised to shake off the setback of recent years,” Brian DePratto, an economist at Toronto-Dominion Bank, said last month when Statistics Canada reported gross domestic product grew at an annual rate of 2.6 percent in the fourth quarter, the fastest in years. The Bank of Canada held back. “While the recent rebound in GDP is encouraging, it is too early to conclude that the economy is on a sustainable growth path,” the central bank said in its latest policy statement on April 12.
Poloz isn’t the only joykill. Maurice Obstfeld, the University of California, Berkeley economist who currently is serving as the IMF’s head of research, also emphasized the negatives when he released his new forecast. At a press conference, Obstfeld noted that confidence readings appeared “excessive.” That’s a euphemism for irrational. Obstfeld appeared to be warning of an enthusiasm bubble that could pop at any moment.
Anyone who pays even a little attention to stock markets knows that investors have been deliriously happy since Donald Trump won the presidential election in November. Executives and consumers have been telling pollsters that they suddenly feel great about the prospects, probably because Trump and the Republican leaders of the House of Representatives and the Senate have promised big tax cuts and fewer regulations. The Conference Board’s index of consumer confidence slipped in April, but still is higher than at any point during in pre-crisis expansion, according to CIBC World Markets.
Economists call these surveys “soft” data because they are based on what fallible humans tell other fallible humans over the phone. They can herald shifts in the trend, but they say nothing about how many goods were actually ordered or services that were actually sold. That’s why economists rely on “hard” data such as retail sales, industrial production, and exports and imports to estimate what really is going on the economy.
In recent weeks, observers have begun to notice that the soft data and the hard data are telling different stories, especially in the United States. A closely watched survey of consumer sentiment published by the University of Michigan is at its highest levels since 2000. Yet the most recent measures of retail sales, manufacturing, housing starts and inflation are all negative. The Federal Reserve Bank of New York’s Nowcast, which incorporates soft data to make real-time forecasts, predicts the U.S. economy grew at an annual rate of 2.7 per cent in the first quarter. The Atlanta Fed’s GDPNow forecast, which uses only hard data, predicts first-quarter growth of only 0.5 percent.
Most sentiment surveys are superficial. They tend to ask yes-or-no-type questions, and then the yes’s and no’s are tallied to create an index. Many of these indices have decent records of predicting the economy’s path, but they are subject to sharp movements. Consider the weekly Canadian consumer sentiment survey by Bloomberg and Nanos Research. The report for the week ending April 21 shows that expectations that home prices will keep rising is driving overall confidence. What happens if prices correct? Confidence would plunge.
The Bank of Canada values a quarterly survey it does of 100 companies. On the surface, the most recent Business Outlook Survey suggests investment is picking up after an unexpectedly long period of decline. But the central bank doesn’t stop at the yes-or-no questions. It probes deeper to make a more qualitative assessment of business intentions. By doing so, it discovered that most companies were planning only to spend money on maintenance. That aligned with Statistics Canada’s more thorough survey of investment plans, which found at the start of the year that executives planned to increase spending by only 0.8 percent in 2017 after three years of decline.
“There is definitely an improvement in sentiment around investment,” said Poloz. “And yet StatsCan’s very thorough survey of companies shows that the planned level of investment really has not gone up: very, very modest growth. So it doesn’t correlate with the sudden move upward in the soft data. The soft data are to help you find turning points when they happen. They can move quite suddenly, by a lot, because more people said yes versus saying no. It’s a correlated thing with the underlying process, but it’s not 100 per cent or anything close to it.”
On business investment in Canada, Poloz added: “It’s not expansionary investment, which is what we are actually watching for. I’m not saying nobody is expanding because of course you can always come up with examples that are, but as a macro phenomenon, it’s not really there.”
Canada’s central bank chief seemed frustrated with all the criticism he’s had to endure from the more faddish forecasters on Bay Street. “There is an advantage, I guess, to being able to do a new forecast pretty much every week, but we don’t,” he said. I told him that sounded like a dig at the faddishness of private forecasters. “No,” he said. “It’s a compliment. It’s an advantage. We just do the four of them. I ask you to bare some of those time horizons in mind because we don’t do running commentary on the data.”
Forecasters also are human. They rely to a significant degree on mathematics and algorithms, but every prediction also contains a certain amount of judgment. Institutional forecasters such as the Bank of Canada and the IMF have struggled for years to get a handle on the post-crisis economy, which has behaved much differently than their models predicted. Poloz conceded that being consistently wrong can start to play with your mind.
“If you had suffered serial disappointment, which is what we did suffer, you revise down time and time and time … then at some point forecasters are going to say, `Gosh, I guess that’s a long-term trend.’ They begin to embody that,” he said. “Serial disappointment teaches you a lesson. For sure it teaches you to be cautious.”
Poloz’s reticence has brought accusations that he is playing games with currency traders. If he denies the economy’s strength, investors will be forced to assume that he intends to leave interest rates low, which will keep downward pressure on the exchange rate. That’s unlikely. Bay Street economists have very little at stake when they make forecasts. Some even have an incentive to be aggressive because getting ahead of the pack brings attention to themselves and their employers. Reputations are made on being the first to get a call right; oddly, blown forecasts are quickly forgotten on Bay Street.
But mistakes aren’t forgotten in Ottawa. If the Bank of Canada moves too soon, it could derail the rebound it has been trying so hard to inspire. He has no incentive to rush, especially since the central bank has been so often wrong in recent years. “We have our ways of ensuring that every time we are doing a new forecast we take in everything that’s visible to us into account, that’s as honestly balanced as our methods will allow us,” he said. “If it’s a mistake that’s been made, it’s in good faith. It’s not purposefully biased.”
Watch: Stephen Poloz’s historical ties to the Canadian Pacific Railway