The Decibel - How ‘black swan’ events are ruining economic predictions

Episode Date: May 8, 2023

Economic predictions can be incredibly important for companies and governments trying to plan for the months and year ahead – and that ripples down to individuals trying to budget as well. But some experts worry that the forecasts are becoming less and less accurate.Todd Hirsch has been an economist for around three decades. Today, he tells us why forecasts aren’t working as well as they used to, how so-called ‘black swan’ events are changing things, and what that means for our ability to plan for the future.Questions? Comments? Ideas? Email us at [e-mail]

Hi, it's Mainika. Your support means a lot to us here at The Decibel, and your feedback matters too. And we want to know what you think of the podcast. So we've set up a quick survey where you can tell us your thoughts. And as a thank you, there's a chance to win some prizes. We'll tell you how at the end of today's episode. Back in January 2020, the Bank of Canada was predicting slow but steady growth for the economy. But then, COVID hit. Everything shut down and the economy fell into chaos. Something similar happened last year, when Russia invaded Ukraine.

Economic forecasts can be incredibly important, especially for companies and governments, but some experts say they're becoming less and less accurate. Todd Hirsch has been an economist for around 30 years, and he spends a lot of time trying to predict the economy. Today, he tells us why this isn't working as well as it used to, how so-called black swan events are changing things, and what that means for how we plan for the future. I'm Aynika Raman-Wilms, and this is The Decibel from The Globe and Mail. Todd, thank you so much for joining me. Oh, it's a real pleasure to be here. Thank you.

So let's just start with the base question here, which is why do we have economic forecasts in the first place? Like, why are they important? when economists started to develop these econometric models where you would put in your assumptions and out would come sort of a forecast of what is likely to happen next. And it's all sort of based on historic patterns. So as economists got better and better at doing this and more mathematically inclined, a whole sort of industry grew up around economic forecasting to the point where we're at today, where people put a lot of emphasis on these forecasts.

Can you give me a sense of like, like who would be using these forecasts? Who finds this actually valuable? Well, they report these forecasts in the news. So a lot of, you know, Canadians just driving around, listening to the news, they'll hear, oh, you know, this bank is forecasting real GDP growth to be 4% this year or 1%. The Bank of Canada, of course, uses forecasts to give guidance to institutions, other governments, what the economy is likely to do or their best assumptions about what the economy is going to do.

It's used a lot by companies that are doing budgeting. And so it's important in that respect, that the company that is putting together their budget, or their projections on sales or for the next year that they have some idea what the economy is likely to do. Okay, so important for companies. I imagine it's also important for governments, though. So how would a government use an economic projection like a GDP projection? How would they use that, for revenue, they might expect in the coming year. If the economy is going to project it to do really well, or if it's projected to do poor, that will give the government a better sense of what they can expect for revenues and also spending.

When we think about employment insurance, or when we think about other benefits and supports that governments pay, that's a cost to governments. What the economy is going to do or likely to do will give them some guidance in how to plan around that. Okay. And so, Todd, you said these forecasts are essentially based on a set of assumptions about what might happen.

So how accurate are these forecasts? Well, we're getting worse and worse at doing it, and it's not a reflection of the mathematical abilities of the forecasters. It's not a problem with the model. The problem is that in the 21st century, we are bombarded by increased numbers of what we might call black swan events. These events that come out of nowhere, that almost by their definition are unpredictable, but they have a massive impact either, you know, plus or minus to the economic forecast. The very best example would be what happened in March of 2020 when the world was plunged into a pandemic. Nobody really saw that coming, even though, you know, people for years have been sort of warning of a pandemic coming.

No one knew the timing of it. So it really threw forecasts out the window. And it's not just COVID. It's all sorts of other things. It's Russia's invasion of Ukraine. It's more severe and more violent and destructive weather events. These things are impossible to predict with any accuracy. And as a result, economic forecasting has been worse. We've gotten these forecasts wrong more often. I want to get into how they are worse and, you know, what we can do about that. But maybe let's just stick with some of these examples that you've thrown out here, Todd. Can you walk me through one, I guess? Like, let's look at one of these black swan events, as you say. How exactly would it have messed up

economic forecasts? Well, we can take the Russian invasion of Ukraine. When that happened, when the tanks rolled across the border, almost instantly financial markets swooned because there's instability and there's questions about, you know, military conflict, a ground war in Europe. So, you know, investors were anxious, markets would fall. But the other thing that happened is commodity prices went through the roof. And that's because both Ukraine and Russia are really important producers and suppliers of agricultural commodities, and in the case of Russia, energy commodities. So we saw oil prices, you know, spike through the

roof. We saw wheat prices and canola prices. But it threw the forecast out the window because suddenly these commodity prices spiked, oil particularly. So for here in Alberta, when we're trying to make a forecast, energy prices are really important. Nobody really was able to factor in Russia's invasion of Ukraine and what that did to oil prices. What about an example like with natural disasters? You mentioned that a little bit earlier. So we're seeing obviously more natural disasters these days because of the effects of climate change. And I'm thinking about something like, you know, the BC floods that happened in 2021.

How would something like that change things? The flood in BC in 2021, Hurricane Fiona on the east coast of Canada in late 2022, even some of the ice storms that hit Canada and the major airports, these events, because they bring a lot of economic activity grinding to a halt very quickly, it can, depending on how long it goes on, it can have a very negative impact on the economy. The floods in BC, a great example, because that was weeks or even months before some of those transportation corridors were back up and running. So it's really disruptive to the economy activity, what that normally would be happening is not able to happen. So it leads to a big drop, a sudden drop in economic activity, at least in the short run. Okay, so this makes sense. I can understand how like a black swan event, as any of these things are, can kind of throw things into disarray. You talked about how forecasts are getting more inaccurate, Todd. So let's focus on that. Like how inaccurate are forecasts now? Can

we get into some of the numbers? Well, it's hard to pinpoint how accurate or inaccurate they are because we can always revise a forecast. And I think what's happening now are companies or there may be economics teams, they're being asked to revise forecasts more frequently as situations develop. And that's fine. But, you know, taken to the absurd extreme, we could be revising a forecast every hour with new information, and you're never really going to be wrong. But then it's not much of a forecast at all. It's just sort of a real-time read of what the economy is doing. And maybe that's what we need to think about rather than locking in a forecast, you know, and giving people this false sense of confidence that what the economy is going to do in several months or several years.

Let's focus a little bit then on how the economy has been behaving recently because you said these forecasts are based on how the economy has responded historically. So how is it behaving these days? Well, still with a lot of unpredictability. Another great example sort sort of a recent situation here in Canada, is late 2021, early 2022, the Bank of Canada was basically saying we're going to keep interest rates low. They did not expect any, you know, inflation pressures. But these inflation pressures in 2021 did sort of come out of nowhere. They caught everybody, including the Bank of

Canada, by surprise. And as a result, the Bank of Canada did have to start raising interest rates quite aggressively in early 2022 and right up until, you know, sort of they've leveled off now. But throughout 2022, we saw a big increase in interest rates by the Bank of Canada. And it's not that, you know, the Bank of Canada, I don't want to suggest they got it wrong. They reacted to new information when it was available. It's just that back in 2021, I think they left Canadians with a bit of, you know, again, false confidence, this idea that interest rates are going to stay long for an extended period of time. They spoke of that with a lot of confidence.

And a lot of Canadians sort of made decisions based on what the Bank of Canada was saying, that interest rates aren't going anywhere. But then they did. So what does that mean for people who are listening to these forecasts? How does that impact the people in Canada who are making these decisions for themselves? Well, this is where, you know, there's some really sad stories. I remember listening on the radio, a story of a woman, a young mother who had listened to the advice of the advisors around her. She had done all the right things.

She had sort of done the homework and listened to what the Bank of Canada was saying. And at the time in 2021, you know, go with a variable rate mortgage because you'll get a better rate and interest rates aren't going anywhere. But then they did. So in this case, this woman, even though she did her research, she was still caught off guard. So this is where, you know, I talk about leaving people with a sense of false confidence. We always have to remember anything could happen.

A forecast is really only based on the assumptions that we put in it. If there is an event like inflation spiking out of nowhere, like Russia's invasion of Ukraine, we have to remember that those kinds of events can throw these forecasts off. So we never really want to have all that much confidence in a forecast. We need some healthy skepticism. It's good to say, okay, here's what is likely to happen, but we always have to keep in mind anything could happen and we have to be prepared to react to that. We'll be back in a moment. So is the economy behaving the same way that it has in the past? Like,

is looking to the past, I guess, still helpful for us in this moment? Well, that's a whole other issue. When we look at the economy in the 21st century, it doesn't really resemble the economy of the mid-20th century in many ways at all. We could get into all of the new industries that we're seeing, for example, artificial intelligence, cybersecurity, cryptocurrencies. All of these things are having an impact on the global economy that didn't exist 50 years ago when econometric modeling and, you know, looking at the patterns of how an economy reacts when all of those were first established. The economy is not behaving as it did in the past, especially the labor market. And this has a lot of economists scratching our heads because when we look at the past, you know, certain patterns of the labour market are quite predictable. But this time, you know, and it might be because of COVID,

it might be some of the psychology of workers has changed fundamentally. There's also social and demographic things going on, but the labour market is not really behaving in the same way that it has previously. So recently we've had a lot of really good job numbers coming out in Canada. And this is sort of defying gravity because we've seen interest rates increase. This is supposed to be putting the brakes on the economy. At this point, we would probably expect fewer jobs being created. We'd expect the unemployment rate creaking higher. We're not seeing it at all.

And I think there are some reasons for that. I think there's some reasons why consumers have not been responding to higher interest rates like they would have in the past. In the past, they would very quickly have stopped buying homes or bought fewer cars or consumer spending would be dampened by the increase in interest rates. But we didn't really see that coming out of COVID, not in the same way, not in the predictable way that we would have seen it in the past. So it's left us a little bit confused and a little bit curious as to why the economy, especially the labor market, is behaving in a different way.

So Todd, you're saying you're kind of we're in an almost like an unprecedented state here, especially the labor market, is behaving in a different way. Yeah. So, Todd, you're saying you're kind of we're in an almost like an unprecedented state here. But I guess I want to push back on this a little bit because there have always been wars and health scares and bank collapses. So what's different about the way that things are happening now? Yeah, you're absolutely right. None of the things that we're seeing today are, you know, especially new. Maybe COVID. I mean, we haven't been through anything like that since, you know, 1918. Not in the same way, not in the same global pandemic at which these disruptive events can disrupt a global economy. We are more connected globally than we were 20 years ago or 50 years ago for sure. We trade more in commodities than we did 50 years ago. So when Russian oil exports are cut off, that has a really immediate impact on oil prices around the world.

It happens literally in hours. These events, because we are more connected electronically, we're connected through trade and global commerce. It's not that they're new, but they're having a bigger impact on the global economy and a more immediate impact. So we feel these shockwaves more suddenly and maybe more extreme. I mean, I understand the connectedness factor that seems fairly straightforward, but can you just explain the immediacy a little bit to me? Like how has learning about these events so quickly really affected things? Well, take for example, the collapse of Silicon Valley Bank. It was a Friday morning that happened. By Friday afternoon, the entire world knew about this, and everyone was starting to look at their bank a little bit differently and wondering, boy, I wonder if my bank is going to be the next one to go through this. So, say, 50 years ago, we would have seen bank collapses, but especially because it happened on a Friday, people wouldn't have even heard about this.

They might have read about it in a newspaper over the weekend. But because of the connectedness and the immediacy of global media, everyone, at least everyone sort of in the banking and finance world, everyone knew about the collapse of Silicon Valley Bank within a few hours, and it had a more immediate and a bigger impact on the financial markets. So because of all of these things, you know, like we said, there's always been wars, there's always been bank failures. But today, these Black Swan events, these unforeseeable events, not only are they making forecasting more difficult, but they are presenting companies and individuals with a sort of a wider array of possible scenarios that could play out. And this is where I'm thinking a role for an economist, we might be better serving people if in fact, rather than trying to come up with our pinpoint forecast for GDP,

but rather help individuals and help companies think their way through any number of scenarios that could befall them and how they are going to react to that. Is that something that's attainable that predictors could actually start doing that? Well, in some ways, a lot of economists, we do do this. And when I was the chief economist at a financial institution out here in Alberta, we always did. Our forecast contained a low, a medium, and a high case scenario, recognizing that we have not a lot of confidence that these different events could happen. The problem with that is people still want to say, well, what's the base case?

Give us a number. They still sort of want to default to what is the number, even if we do try to say, well, we have to be conscious there is a possibility of a low case or a high case scenario. So in a lot of ways, we do try already to present a bunch of different scenarios that could play out. But maybe it's in the wording that we need to repackage this a bit. So as not to leave people with this, again, the sense of false confidence that it's going to be the medium case, the base case scenario, which is what we call it sometimes. So how do you think things could be done differently going forward? What's something that we could actually kind of tangibly do here to make it better? I think maybe what we could do is look

at some of the black swan events in recent history, the geopolitics, for example, and maybe extrapolate that into the future and say, well, what does it look like for this industry or for this company if we are to see intensified geopolitical tensions around the world? What is the scenario that could play out without trying to be predictive and saying, here's what is going to happen, but rather say, well, here are some possible scenarios. Can we think our way through as an industry or a company or a government? Can we think our way through as an industry or a company or a government can we think our way through what would be our response if this happens or if that happens and instead of just one forecast number look at a broad array of different scenarios and sort of focus on what is our response to each of those and i think that's the gist of what I'm talking about lately is not that we shouldn't try to look into the future,

but we should try to be less confident or we should be less confident about these pinpoint forecasts that these econometric models produce because they're not likely to be right. We need to look at all of the possible scenarios and be able to react to any of those. Todd, this is really interesting. Thank you so much for joining me today. It's been a real pleasure. Thank you for having me. Thanks for listening today. So about that survey on the decibel, we want to know about you and what you want to hear on the show. Just go to thedecibelsurvey.ca, fill it out, and as a thank you, you'll be automatically entered to win one of three

grocery gift cards for Loblaws worth $100. Again, that's thedecibelsurvey.ca. We can't wait to hear from you. That's it for today. I'm Mainika Raman-Wilms. Our interns are Wafa L. Reyes, Andrew Hines, and Tracy Thomas. Our producers are Madeline White, Cheryl Sutherland, and Rachel Levy-McLaughlin. David Crosby edits the show. Adrian Chung is our senior producer, and Angela Pachenza is our executive editor. Thanks so much for listening, and I'll talk to you tomorrow.