Tuesday, June 2, 2020

Pitfalls in Epidemiological Modelling


Tyler Cowen is sounding just a bit fed up with Twitter, in this Marginal Revolution post and in particular with the way Twitter has degenerated into a name calling equilibrium. Prisoners Dilemma with everyone choosing to betray (what’s the payoff structure that would support such an outcome?).  the point at issue is the explanation which one modeller has given for why his model overestimated the impact of COVID in Germany – his casual reference to Dark Matter prompted people to say that “Dark Matter” is shorthand for “my model doesn’t work, and I don’t know why”.

The sad thing is that if he’d said that it must mean that people were doing more social distancing than he’d anticipated, he would have placed himself squarely in the mainstream of COVID modelling.  At some point we’ll have to face up to the fact that the epi models for COVID have bee pretty much useless.  Most of the world fell into line behind the Imperial College London (ICL) model, despite its tendency to produce massively wide prediction bounds (you can’t call them confidence intervals since most modellers don’t assign probability weights to them).  The Babylon Bee, a satirical website, ran a headline a while ago saying that: “Latest Computer Model Predicts Between 0 And12.6 Billion New COVID-19 Deaths By Summer”  but when it comes to epi modelling, reality is well on its way to overtaking satire. 

COVID-19 and Liquidity Preference


Stephen Poloz made an observation, both in his University of Alberta lecture and also in his final appearance before the Senate National Finance Committee as Governor of the Bank that part of the reason the Bank was buying large quantities of government bonds from financial institutions (as distinct from the ones it’s buying from the government in the primary market) had to do with large firms drawing down their lines of credit.  There’s more (out of the US) on this phenomenon here as well as a VoxEU post on it here and a (pretty technical) National Bureau of Economic Research working paper on it here.

The basic idea is that one thing which a lot of firms learned in 2008 was that if you waited until they were in financial difficulty to all on their lines of credit with their banks, they could find their access truncated by banks which were worried about their own ability to support those loans.  So this time they decided to draw on their lines of credit as soon as the first signs of trouble, rather than wait and see how things played out.

Trudeau calls in the big guns from the Boston Consulting Group


A colleague of mine brings this story to my attention:  “Ottawa working with U.S consulting firm on early planning for economic recovery”.
According to the Globe and Mail report:

“One of the world’s biggest multinational consulting firms is advising Justin Trudeau’s cabinet on how Canada could emerge from its COVID-19 lockdown, as a growing number of federal ministers get involved in early discussions around an economic recovery strategy.
Canadian representatives of the Boston Consulting Group (BCG) met for several hours at the start of this week with roughly eight of Mr. Trudeau’s ministers, as part of a broader advisory role the U.S.-based consultancy is serving on an unpaid basis.”
………………………………………………………………………………………………………………………………………………………………….
“…According to the agenda for Monday’s meeting, the consultants provided an overview of the pandemic around the world, before ministers discussed “game-changing ideas” and “building blocks for recovery.” The document pointed out the government still had much work ahead.”

Friday, May 29, 2020

Measuring Inflation in Lockdown


Statistics Canada released the April CPI report on May 20.  The headline sentence was:

“The Consumer Price Index (CPI) fell 0.2% on a year-over-year basis in April, down from a 0.9% gain in March. This was the first year-over-year decline in the CPI since September 2009.”

If you take a look at Chart 1 in the Daily release  you’ll get a pretty dramatic illustration of what StatCan says next: 

“The CPI declined as energy prices fell as a result of the COVID-19 pandemic. Excluding energy, the CPI rose 1.6%.”

So when you take energy out of the CPI, inflation was 1.6%, with energy in the index it was -0.2%.

Thursday, May 28, 2020

The Bank of Canada's Balance Sheet


One expression which you come across pretty frequently these days refers to the Bank of Canada as having enlarged its balance sheet.  Given what the Bank is, that basically refers to its creating new money, but just saying that is being a bit too bald.  Fortunately, the Bank has an explanation of its balance sheet here, and while the text at that site was written in 2019, the charts are up to date.

The Bank’s balance sheet is a statement of the Bank’s assets and liabilities, where its liabilities are what make up the monetary base.  Since we refer to an increase in its liabilities as “printing money”, it’s interesting to note that back at the end of 2018 the largest single item among its liabilities was “Bank notes in circulation” at about $90 billion, followed by Government of Canada deposits at about $22 billion.  (Strictly speaking that’s liabilities and capital, but I’m going to be sloppy here.)  Deposits held at the Bank by members of Payments Canada came in at $251 Million, as is consistent with Canada’s having eliminated reserve requirements as we traditionally talked about them quite some time ago.  The asset side was dominated by Government of Canada bonds, at $79 billion, with Government of Canada Treasury Bills coming in second at $24 billion and repos at just over $10 billion. Total assets and liabilities at year end, 2018, came (each) to about $116 billion.

Sunday, May 24, 2020

Stephen Poloz's Optimism


Bank of Canada Governor Stephen Poloz has been sounding remarkably sanguine lately, and presumably not just because in a few days the current mess will be Tiff Macklem’s problem.  In comments to the press the other day he was saying that he thought that the pessimism he’d been hearing was overblown, and he was talking about how we’d shut off the economy and used fiscal and monetary policy to put supports under it, going on to say:

“So afterwards, when we turn the economy back on – which we’re just beginning to do – you should see a very rapid return of production,” he said. “If we’ve done a good job, it will have been cushioned well – the shock absorbers have been put in place, and people can pick up where they left off.”

He said that it might take us a year to get back onto the old growth trend line, but he is quite explicitly saying that we will return to the pre-COVID growth trend.

Poloz has said that he tends to think of the current episode as being akin to the effects of a natural disaster.  I’m not entirely convinced of the analogy, since a natural disaster generally involves some destruction of productive capacity, but I’m probably misreading the analogy.  In any event, Poloz is giving a video talk on “Monetary Policy in Unknowable Times” to the University of Alberta on Monday May 25,  with the video to be posted on the Bank’s website.

The rapid bounceback argument does rest heavily on the view that the current situation is best thought of as a major Aggregate Demand event.  Thinking about how to model it, I’d be inclined to set it up as a forced saving variant of the Paradox of Thrift story.

Thursday, May 21, 2020

The Nature of the COVID-19 Economic Downturn


On May 5, the  Macdonald-Laurier Institute’s Philip Cross released the latest run of the MLI’s Leading Economic Indicator, in a post which was headlined  WORSE THAN A RECESSION,CANADA’S ECONOMY IS HEADED TOWARD FREEFALL.  That’s a judgement that it’s hard to dispute.  Further, it’s important to remember that the index is a Leading Economic Indicator, so it’s not just looking at what’s happening today, it’s looking at what today’s occurrences say about what’s likely to happen in the next few weeks.  Overall Philip Cross’s index is pretty accurate, so it’s worth taking seriously.  We can expect to see the Aggregate Demand Curve continuing to shift to the left, even as lockdown eases, in part because people who have been maintaining their spending by running down their assets are going to be running out of those assets.  Government transfers will help cushion that a bit, but we can still expect to see AD remaining soft even after the recovery begins, as people try to restore their depleted stocks of financial assets.  (Cross enumerates the elements of the damage here.)

Sunday, April 19, 2020

The Macdonald-Laurier Institute’s April Economic Report


The Macdonald-Laurier Institute has just released Philip Cross’s April 2020 Economic Report, which focuses on the hammering the Canadian labour market took in March and on why Philip Cross argues that the unemployment rate is no longer a good indicator of the state of the Canadian labour market.

Cross’s starting point is the fact that, according to the March Labour Force Survey (LFS) unemployment jumped from 5.5% to 7.8%.  Cross makes the point that, while this is the largest ever jump in the unemployment rate in a single month, it’s not even a good measure of what’s happening to measured unemployment, since the LFS is done in a single week, so it doesn’t cover the entire month. 

Beyond the fact that over a million jobs disappeared in March, another 1.3 million people weren’t able to work, despite still being employed and almost a further million reported reduced working hours.  Cross notes that while employment dropped by 5.3%, total hours worked dropped by a full 15.1%.

This drop is obviously going to carry through to GDP, and since GDP is a measure of income generated in the country in a year, it will also carry through to income.  It all works through the aggregate production function.   

Saturday, April 18, 2020

The Bank of Canada’s April 2020 Monetary Policy Report


The Bank of Canada released its April Monetary Policy Report on the 15th and, as you would expect, it does not make for cheerful reading.  The focus is on how deep the recession’s going to be and how long it’s going to last.  Normally in the MPR the Bank reports what it calls the base case forecast – what it sees as the most likely path for the economy to follow over the next few months, then goes on to discuss upside and downside risks – factors which could make the economy grow faster than the base case projects (a risk because it could lead to unexpected inflationary pressure) and factors which could cause growth to fall short of the base case projection.  In the April report they don’t do that - there’s so much uncertainty about the next few months that the Bank doesn’t even report a base case projection.  Instead it pools all of the simulations it’s run, each of which would have incorporated different assumptions about the behaviour of key economic variables, into something which is almost, but not quite, a fan chart for GDP (Chart 11) and another for inflation (Chart 12).  Not quite, because a fan chart usually indicates the probability of different paths by some kind of colour coding, with the darkest range being the most likely path.  The April MPR doesn’t draw any such distinctions – instead its graphs include shaded areas which it labels as the range of the scenarios, and there’s no variation in shading within the range, suggesting that all of the possible trajectories are equally likely (or that there’s no basis for thinking one is more likely than the others).

Tuesday, April 14, 2020

COVID-19 And Demand and Supply Side Effects

John Palmer has a post at his EclectEcon blog on how he's handling the COVID-19 situation.  Down towards the bottom he remarks:
 And I further resent the use of the term "stimulus". There's nothing to stimulate! This is a major supply-side shock to the economy, and until people can start producing goods, and especially services, we can't help but experience a major drop in GDP.
The whole issue of whether the current shock is a supply-side or demand-side episode informs not just what's happening right now, but it also plays into what the recovery will look like when restrictions are eventually eased.  Whether, as John Cochrane puts it in this Grumpy Economist blog post we're going to have a V- shaped, a U-shaped or an L-shaped recovery.

A V-shaped recovery would mean a deep downturn while restrictions are in place, followed by a large and rapid recovery as soon as they're lifted.  That's what a lot of the early commentators said would happen, and while it's not impossible, it's looking increasingly unlikely.