Canadian International Trade

In April 2020, Canadian exports and imports fell respectively by 29.7 % and 25.1 %, compared to the previous month. Compared to the same period last year (i.e., April 2019), they declined respectively by 35.2 % and 30.6 %. These historic declines have more to do with the confinement than with the border closures during the COVID-19 pandemic. Border closures apply to the movement of people, but not to freight shipping. The confinement or the nationwide lockdown has affected the production and the sales of goods that were not deemed essential by the provincial governments. As a result, their supply and demand fell, which impacted on international trade.

Figure: Exports and Imports, Canada, 1997:M1-2020:M4

As it appears in the above figure, the trade balance of Canada (i.e., the difference between its exports and its imports) has become negative since December 2008. Even though, in April 2020, the decline in the exports or in the imports was unprecedented, the trade deficit occasioned was not. In April 2020, the trade deficit was 3 251 million versus 4 953 in December 2018.

There will not be any important fall in the Canadian gross domestic product (GDP) because of the decline in both exports and imports. There are two reasons for that. First, both exports and imports fell. Second, it is rather the balance of trade that contributes to the GDP.

The United States (US) are Canada's main trading partner (see the table below). In April 2019, 73.6 % of Canadian exports went to the US. In April this year, this share dropped to 68.4 %, as exports to the US decreased by 40.6 %. On the other hand, between these two periods, the share of China rose from 4.2 % to 6.3 % and the share of Japan rose from 2.2 % to 3.7 %. The increase in the share of China is mostly due to the decline in the exports to the US. Exports to China rose by 15.2 % between March and April, but they fell by 4.5 % compared to April 2019. On the other hand. exports to Japan rose by 8.1 % year-over-year and by 26.3 % between March and April.

Table: Canada's Main trading Partners by Shares of Exports and Imports, 1997:M1-2020:M4.

Country	Exports	Imports
United States	77.8 %	67.5 %
European Union	7.3 %	9.8 %
China	2.8 %	4.9 %
Mexico	1.4 %	2.6 %
Japan	2.4 %	2.6 %

Imports from the US, Mexico, and Japan also fell. Between April 2019 and April 2020, while the share of the US in Canadian imports was falling from 63.7 % to 56.1 %, the share of China rose from 7.4 % to 11.3 %. The imports from such other trading partners as Brazil, Peru. and Switzerland also rose. In April 2020, imports from these three countries respectively rose by 52.3 %, 36.9 %, and 24.1 %.

Does the Consumer Price Index Accurately Measure Changes in the Living Cost?

The consumer price index (CPI) in Canada declined by .66 %, in April this year compared to the previous month. On a year-over-year basis (i.e., compared to April 2019), it declined by .15 %. This is the first year-over-year decline in the CPI observed in the month of April, since 1992. The average year-over-year inflation rate (i.e., percentage change in the CPI) for the month of April is 1.74 %. I am wondering if the CPI or the inflation rate of April 2020 really makes sense as a measure for the living cost during the lockdown of the Canadian economy.

Basically, the CPI is a weighted average of the retail prices of the goods and services consumed by households. These goods and services are classified into eight product groups. The table below shows the averages and the values in April 2020 of the year-over-year percentage change in each of these eight product groups' CPI.

Table: Year-over-Year Percentage Change in CPI, Canada, 1992:M2-2020:M4.

Product group	Average	April 2020
Food	2.24 %	3.49 %
Shelter	1.87 %	1.32 %
Household operations, furnishings and equipment	1.26 %	.24 %
Clothing and footwear	.06 %	-4.40 %
Transportation	2.44 %	-4.39 %
Health and personal care	1.45 %	1.42 %
Recreation, education and reading	1.44 %	-.26 %
Alcoholic beverages, tobacco …	2.93 %	.41 %
All items	1.80 %	-.15 %

In April, the year-over-year percentage change in the CPI for food was 3.49 %, which is much higher than its historical average of 2.24 %. The year-over-year percentage change in the CPI for shelter was also high (1.32 %) but below its historical average of 1.87 %. These two product groups along with alcoholic beverages, tobacco products and recreational cannabis were mainly the goods deemed essential and mostly the only ones that were available to households, during the lockdown imposed by the federal and the provincial governments to stop the spread of the COVID-19.

The fact that their CPIs rose on a year-over-time basis and, at the same time, the year-over-year inflation rate fell casts a doubt on the use of the CPI for all items as measure of the cost of living. It is true that the CPI for clothing and footwear and that for transportation fell by 4.4 % on a year-over-year basis, which has dragged down the inflation rate, but these goods and services were not those people in Canada mainly purchased during the lockdown.

One also ends up at the same conclusion, looking instead at the monthly growth rates of the CPI for these product groups. In April 2020, the CPI for food grew by 1.12 % compared to March, the monthly CPI for alcoholic beverages, tobacco products and recreational cannabis grew by .12 %. Shelter was the only product group deemed essential whose CPI declined (-.34 %). One can still sustain that the .34 % decline in the CPI for shelter has caused a decline in the living cost during the lockdown, as both the year-over-year and the monthly inflation rates were suggesting. To rule out this possibility, I have computed the shares of each of the eight product groups in the monthly inflation rate. They are plotted in the pie chart below. (I have computed these shares by performing a linearly constrained optimization.)

Figure: Shares of Eight Product Groups in the Monthly Inflation Rate

It turns out that shelter is the product group that accounts for the largest share of the monthly inflation rate in Canada (27.5 %). Food accounts for 16.5 % while alcoholic beverages, tobacco products and recreational cannabis accounts for 5.7 %. Health and personal care accounts for the lowest share of the monthly inflation rate (4.7 %).

Even though shelter accounts for the largest share in the inflation rate, its contribution to the change in the living cost in April was only -.09 % (i.e., -.34 % x .275) whereas the contribution of food was .18 % (i.e., 1.12 % x .164) and that of alcoholic beverages, tobacco products and recreational cannabis was .01 (i.e., .12 % x .057). My conclusion is that the CPI is a good measure of the level of prices, but it cannot accurately measure the change in the living cost in periods of economic lockdown.

Dataset and Code

The Relationship between Unemployment Rate and Economic Growth across Canada

Among the ten provinces of Canada, Quebec showed the highest unemployment rate, in April 2020 (17 %). In January and February, the province of Quebec along with Manitoba and British Columbia recorded the lowest unemployment rate in Canada. In February 2020, the level of the unemployment rate in Quebec was 4.5 %, the lowest since 1976. The extent of the spread of the COVID-19 explains the high unemployment rate in Quebec. As a matter of fact, Quebec accounts for more than half of the confirmed cases of COVID-19 in Canada (44 197 cases out of 79 502, as of May 20). Most of these cases (about half) are in the region of Montreal that generates the third of Quebec's gross domestic product (GDP).

In the Prairie Provinces (i.e. Manitoba, Saskatchewan, and Alberta), the unemployment rates in April 2020 nearly doubled, compared to their historical averages (see Figure 1). For example, in Alberta, the unemployment rate was 13.4 % in April, versus an historical average of 6.4 %. In April, the unemployment rate in Alberta was the highest in the Prairies and also a record high in the history of this province.

Figure 1: Unemployment Rates across Canada.

The picture of the situation is different in Atlantic Canada (i.e. Newfoundland and Labrador, Prince Edward Island, Nova Scotia, and New Brunswick). In Newfoundland and Labrador, the unemployment rate in April 2020 (15.9 %) was very close to its historical average (16 %). This reminds that unemployment rate has often been very high in this province. (See my blogpost Unemployment Rate in Newfoundland and Labrador.) Prince Edward Island is the only province in Canada where the unemployment rate during the outbreak of COVID-19 has been lower than the historical average. As of May 20, only 27 cases of COVID-19 have been found in this province.

Throughout Canada, three economic sectors have been mainly affected by the lockdown restrictions: (1) accommodation and food services, (2) construction, (3) forestry, fishing, mining, quarrying, oil and gas. In the sector of accommodation and food services, between January and April 2020, the unemployment rate rose from 9 % to 44.7 % in Quebec, from 2 % to 30.9 % in Manitoba, from 4.7 % to 35 % in Saskatchewan, and from 5.9 % to 38.9 % in British Columbia.

In the sector of forestry, fishing, mining, quarrying, oil and gas, between January and April 2020, the unemployment rate rose from 8.7 % to 27.9 %, in Quebec. In the construction sector, while in Quebec the unemployment rate rose from 11.3 % in January to 40.3 % in April, it only rose from 5.4 % to 14 %, in Ontario.

In a previous post [here], I have used the Okun's law to predict GDP growth from changes in unemployment rates using data for Canada as a whole. In this post, I take things a bit further by estimating the Okun's law using instead data of the ten provinces (panel data). I have kept the short-run effect of unemployment on real GDP unchanged across the provinces and allow for heterogeneity only in the intercept. (This is referred to as a least squares dummy variable model, in econometrics.)

Figure 2 shows as scatter plot the changes in the annual unemployment rates and the corresponding GDP growth rates in the ten provinces, between 1981 and 2018. The blue line in this figure represents the predictions from the Okun's law. The estimate of the short-run effect of unemployment on real GDP is -1.45, which means a 1 percentage point increase in unemployment rate causes a 1.45 % decrease in real GDP. This estimate is close to the one obtained previously using aggregate data for Canada (-1.4). But, the explanatory power of this panel data model is much higher. It explains 58.5 % of the variability in the data versus 35 % for the model that uses aggregate Canadian data.

Figure 2: Okun's Law: Percentage Point Change in Unemployment Rate and Real GDP Growth Rate, Canadian Provinces, 1981-2018.

The table below reports the predictions of the real GDP growth rates for the ten provinces for the first quarter of 2020 and April 2020.

Table: Predictions of real GDP Growth Rates for the Provinces of Canada based on a Panel Data Model.

Province	First Quarter of 2020	April 2020
Newfoundland and Labrador (NL)	1.4 %	-4.1 %
Prince Edward Island (PE)	2.3 %	-.9 %
Nova Scotia (NS)	1.6 %	-2.6 %
New Brunswick (NB)	1.9 %	-4.7 %
Quebec (QC)	.9 %	-11.2 %
Ontario (ON)	1.5 %	-2.8 %
Manitoba (MB)	1.8 %	-5.2 %
Saskatchewan (SK)	.7 %	-3.7 %
Alberta (AB)	1.9 %	-3.8 %
British Columbia (BC)	1.5 %	-3.8 %

The Okun's model that uses panel data predicts a positive real GDP growth across Canada for the first quarter of 2020. This contrasts with the predictions I made previously using the aggregate Canadian data. On the other hand, it predicts negative growth rates across Canada for April 2020. However, it is worth pointing that these forecasts do not take into account the effects of the various economic stimulus programs initiated by the federal and the provincial governments in response to COVID-19.

Dataset and Code

The Relationship between Unemployment Rate and Economic Growth in Canada

The unemployment rate in Canada rose from 7.8 % in March to 13 % in April, due to the lockdown restrictions imposed by the federal and the provincial governments to stop the spread of the COVID-19. The average unemployment rate in Canada is 8.2 %. The highest ever recorded unemployment rate in five decades (13.1 %) was in December 1982.

In the early 1980s, unemployment, inflation, and interest rates were simultaneously very high in the most developed countries. In Canada, the unemployment rate was steadily above 10 %, between May 1982 and December 1985 (on average, 11.4 %). The second episode of high unemployment period Canada experienced was in the early 1990s. Particularly, between February 1991 and October 1994, unemployment rate rose, as a result of the restrictive monetary policy that aimed at curbing the high inflation inherited from the 1980s.

Since May 11, several economies have started easing the lockdown restrictions. As a result, one can expect unemployment to decline in May and over the coming months. More and more people are returning to work, as some businesses are allowed to reopen. Unfortunately, it is not all the layoff employees that are returning to work. Some businesses failed, due to the COVID-19 pandemic. The travel bans and the physical distancing rules in effect keep affecting the sector of accommodation and food services, where the unemployment rate rose from 18.4 % in March to 34.3 % in April, this year. The unemployment rate in this sector was 6.1 % in February.

The data on the Gross Domestic Product (GDP) of Canada (i.e., the value of the wealth created by Canadian residents) over the first quarter of 2020 are not released yet. But, it is certain that GDP will fall over the first quarter of this year. To predict the extent of this decline, one can use an economic relationship known as Okun's law. Okun's law predicts a consistent relationship between changes in the unemployment rate and the real GDP growth rate. In the US, a 1 percentage point increase in the unemployment rate is said to result in a 2 % decline in the real GDP. Given that the data for the unemployment rate in Canada are already available for the first quarter of 2020 and even for the month of April, I can use them to predict the decline in the real GDP by estimating the linear relationship suggested by Okun.

Even though there has been some deviations from the Okun's law over the years, I use it to predict the changes to expect the real GDP because of its simplicity. The scatter plot in the figure below shows the percentage point change in the unemployment rate and the corresponding real GDP growth rate in Canada. The data points associated with the periods of economic expansion are in green and those associated with the periods of recession are in red. During periods of expansion and recession, the expected quarterly GDP growth rates are respectively .74 % and -1.07 %.

Figure: Okun's Law: Percentage Point Change in Unemployment Rate and Real GDP Growth Rate, Canada, 1976:Q1-2019:Q4

In the above figure, the line in black represents the predictions of the Okun's law that make no distinction between periods of expansion and recession. The short-run effect of unemployment on real GDP from this linear model is -1.2, i.e., a 1 percentage point increase in the unemployment rate results in a 1.2 % decline in the real GDP. This linear model predicts a -.14 % decline in the real GDP during the first quarter of 2020.

In the above figure, the blue segment lines represent the predictions of the Okun's law conditional on the state of the business cycle. The conditional short-run effects of unemployment on real GDP are respectively -.66 and -.29 during periods of expansion and recession. This conditional model (referred to as Markov-switching model) predicts that the real GDP fell by .99 % in Canada, over the first quarter of 2020.

The table below summarizes the predictions of the real GDP growth rate based on the Okun's law.

Table: Predictions of real GDP Growth Rates for Canada based on the Okun's Law.

Stock Exchange	First Quarter of 2020	April 2020
Linear Model	-.14 %	-5.84 %
Markov-Switching Model	-.99 %	-2.34 %

Given the various financial assistance programmes initiated by the Liberal government of Justin Trudeau (which includes the Canada emergency response benefit that provides for a maximum of 16 weeks a weekly pay of $ 500 to layoff employees), the prediction of a decline of 2.34 % in real GDP in April is more realistic.

The updates on the global financial turbulence score will now be available in the tab "The Financial Barometers" of this blog.

Dataset and Code

The impacts of the coronavirus on the global economy: Part VIII: The stock markets

Some stock exchanges are recovering faster than others from the financial crisis caused by the outbreak of the coronavirus disease. On April 28, the NASDAQ composite index and the SIX Swiss exchange mid-cap index were respectively only 5.1 % and 7.3 % below their levels of January 6. On the other hand, the year-to-date return of the Brazil stock exchange index was -30.43 %. Those of the London Stock Exchange FTSE All Share, the Euronext N150, the Bombay Stock Exchange sensitive index, and the Australia Securities Exchange index were about -22 %.

Between January 6 and April 28, capital loss on the Hong Kong Stock Exchange went as low as -25.3 % (this value is the percentage change between the lowest and the highest values of the benchmark index). On the NASDAQ, the range of the capital los was 30.1 %, but this exchange is recovering faster than the Hong Kong Stock Exchange. Why capital loss has been more important on some stock exchanges than the others and why some exchanges have recovered faster than the others? There are two possible explanations. The first one is the sensitivity of the exchange to factors affecting the global economy (the systematic risk) and the second one is the structure or the composition of the exchange.

In the table below, it appears that the systematic risks on the New York Stock Exchange (NYSE) and the Brazil Stock Exchange (Bovespa) are very high during turbulent periods (actually, they are greater than 1). This means that these two exchanges are more exposed to global risk than the other major exchanges. This explains why the year-to-date returns of their benchmark indices are very low. The systematic risk on the Hong Kong Stock Exchange is only .55 and the year-to-date decrease in its benchmark index is less than those on the NYSE composite and the Bovespa index.

Table: Year-to-Date Returns on Apr 28, 2020 and Systematic Risk during Turbulent Periods of some Stock Exchanges.

Stock Exchange	Year-to-Date Return	Systematic Risk
NYSE	-18.81 %	1.11
NASDAQ	-5.11 %	1.11
Tokyo Stock Exchange	-14.80 %	.55
London Stock Exchange	-21.99 %	.83
Hong Kong Stock Exchange	-12.93 %	.51
Euronext	-21.79 %	.88
Toronto Stock Exchange	-13.49 %	.87
Bombay Stock Exchange	-21.05 %	.54
Frankfurt Stock Exchange	-17.76 %	.98
Australian Securities Exchange	-21.12 %	.59
SIX Swiss exchange	-7.28 %	.72
Brazil Stock Exchange BOVESPA,	-30.43 %	1.06

The systematic risk on the NYSE is the same as on the NASDAQ, but the latter exchange is recovering faster than the former. This means that the systematic risk is not the only factor explaining returns on the exchanges. The activity sector and the performance of the main companies in the benchmark indices also explain their year-to-date returns. Half of the companies in the NASDAQ composite operate in the technology sector and 11 % in the healthcare sector. As I show in my previous post [here], these are the two sectors that are performing better during this crisis. Likewise, more than half of the components of the Swiss exchange mid-cap index operate in the healthcare, the technology or the telecommunication sector.

On April 24 the global financial turbulence score rose from 4.38 to 4.62 (a 5.3 % rise). After keeping falling since March 20, The VIX (the implied volatility index) rose by 3.5 % to 37.19, on April 24.

Figure : Global Financial Turbulence Scores and VIX, Jan 8, 2000 - Apr 24, 2020

Dataset and Code

---

The Latest Global Financial Turbulence Scores.

Date	Score
Feb 14, 2020	1.36
Feb 21, 2020	6.23
Feb 28, 2020	3.68
Mar 6, 2020	9.74
Mar 13, 2020	12.89
Mar 20, 2020	9.37
Mar 27, 2020	8.03
Apr 3, 2020	7.70
Apr 10, 2020	4.10
Apr 17, 2020	4.38
Apr 24, 2020	4.62

The components indices of the global financial turbulence score

(1) NYA: the New York Stock Exchange composite index, (2) IXIC: the NASDAQ composite, (3) N225, the Tokyo Stock Exchange average index, (4) FTAS, the London Stock Exchange FTSE all share, (5) HSI, the Hong Kong Stock Exchange index, (6) N150, the Euronext Next 150 index, (7) GSPTSE, the Toronto Stock Exchange composite index, (8) BSESN, the Bombay Stock Exchange sensitive index, (9) GDAXI, the Frankfurt Stock Exchange performance index, (10) AXJO, the Australian Securities Exchange S&P 200, (11) SSMI, the SIX Swiss exchange mid-cap index, and (12) IBOVESPA, the Brazil Stock Exchange index.

Formula

$$d_t² = (r_t - μ ) Σ ^-1 (r_t - μ )',

where $d$ denotes the turbulence score, the vector $r$_t lists the current growth rates of the benchmark indices, the vector $$ μ their historical averages, and $$ Σ designates their variance-covariance matrix. For further details, see Mark Kritzman and Li Yuanzhen (2010).

The impacts of the coronavirus on the global economy: Part VII The sectors of the stock markets

To what extend the outbreak of the coronavirus disease (the COVID-19) has affected investments in the various sectors of stock markets? To find this out, I have computed the year-to-date returns of two exchange-traded funds (ETF) that track the performance of the various sectors of the United States (US) and the global stock markets. The year-to-date return of a fund is the percentage change in its market value between the first trading day of the current year and the current date. The two ETFs that I am using to proxy the performance of the sectors of stock markets are: the select sector Standard and Poor's Depository Receipt (SPDR) funds and the iShares Standard and Poor's (S&P) global. The SPDR tracks the sectors within the S&P 500 (which consists of companies based in the US) and the iShares S&P global tracks the S&P global 1200 index (which consists of companies based in 31 countries).

Energy companies followed by the financial and the industrial companies turn out to be the three sectors that are most affected by the current crisis. In the US and the global markets, energy stocks lost respectively 41.6 % and 39.7 % of their values, between January 2 and April 27 (see the table below). This situation is explained by the dramatic drop in the price of the crude oil. As a matter of fact, over this time period, the spot price of a barrel of the West Texas Intermediate (WTI) crude oil fell from US$ 61.17 to $ 12.17 (which represents an 80.1 % decrease). The price of the Brent crude oil plunged from US$ 67.05 to $ 15.17 (which represents a 77.4 % decrease). On April 20, the WTI turned negative.

Table: Year-to-Date Returns of ETFs, Jan 2, 2020 - Apr 27, 2020.

Sector	Select Sector SPDR	iShares S&P Global
Consumer Discretionary	-9.90 %	-16.61 %
Consumer Staples	-5.76 %	-7.92 %
Energy	-41.61 %	-39.67 %
Financials	-27.67 %	-29.55 %
Health Care	-.54 %	.13 %
Industrials	-23.99 %	-24.05 %
Information Technology	-4.28 %	-5.48 %
Materials	-15.67 %	-18.18 %
Telecommunication Services	-4.45 %	-9.28 %
Utilities	-7.65 %	-9.54 %

The lockdown of economies and the layoffs that followed also considerably harmed the financial sector (personal, commercial, corporate and investment banking, transaction processing services, wealth management, …) and the industrial sector (manufacturers of capital goods, …) in the US and the other markets across the globe.

Health care (pharmaceuticals; health care providers, health care equipment and supplies, …) is the only sector that has recorded a capital gain, during this pandemic. The consumer staples, the information technology, and the telecommunication services are the three other sectors where investors incurred less losses. The reasons are that: (1) consumer staples (food, beverages, home and personal care, alcohol, and tobacco, …) are essential goods and services, (2) the services provided by information technology and telecommunication companies are ways of breaking isolation and loneliness during the lockdown. As an example, during the first quarter of this year, 15,8 million new people subscribed to the movie streaming services of Netflix.

Utilities, (gas, electricity and water distribution), which are known as a defensive sector, poorly performed, as many households waiting for employment insurance benefits had to postpone the payment of their bills.

The year-to-date returns of the select sector SPDR are similar to those obtained using such other major ETFs as the Vanguard and the Fidelity index funds that rather track the MSCI US index. Unlike stock markets, the year-to-date yields of bonds are positive. The year-to-date yield of the vanguard total bond market index fund is 3.92 % and that of the Fidelity total bond ETF is 2.03 %.

After plunging to 4.10 on April 10, the global financial turbulence score rose to 4.38 on April 17 (see the figure below). As I pointed out in my previous post [here], this means that the financial crisis caused by the outbreak of the coronavirus is not over yet. Unlike, the turbulence score, the VIX (the implied volatility index) keeps falling. It went down from 38.15 to 35.93, on April 17 (which represents a 5.8 % decrease). This means that despite the fact that volatility on stock markets starts rising again, investors are less pessimistic about the future.

Figure: Global Financial Turbulence Scores, Jan 8, 2000 - Apr 17, 2020.

Dataset and Code

---

The Latest Global Financial Turbulence Scores.

Date	Score
Feb 14, 2020	1.36
Feb 21, 2020	6.23
Feb 28, 2020	3.68
Mar 6, 2020	9.74
Mar 13, 2020	12.89
Mar 20, 2020	9.37
Mar 27, 2020	8.03
Apr 3, 2020	7.70
Apr 10, 2020	4.10
Apr 17, 2020	4.38

The components indices of the global financial turbulence score

(1) NYA: the New York Stock Exchange composite index, (2) IXIC: the NASDAQ composite, (3) N225, the Tokyo Stock Exchange average index, (4) FTAS, the London Stock Exchange FTSE all share, (5) HSI, the Hong Kong Stock Exchange index, (6) N150, the Euronext Next 150 index, (7) GSPTSE, the Toronto Stock Exchange composite index, (8) BSESN, the Bombay Stock Exchange sensitive index, (9) GDAXI, the Frankfurt Stock Exchange performance index, (10) AXJO, the Australian Securities Exchange S&P 200, (11) SSMI, the SIX Swiss exchange mid-cap index, and (12) IBOVESPA, the Brazil Stock Exchange index.

Formula

$$d_t² = (r_t - μ ) Σ ^-1 (r_t - μ )',

where $d$ denotes the turbulence score, the vector $r$_t lists the current growth rates of the benchmark indices, the vector $$ μ their historical averages, and $$ Σ designates their variance-covariance matrix. For further details, see Mark Kritzman and Li Yuanzhen (2010).

The impacts of the coronavirus on the global economy: Part VI The oil market

On January 24, 2020, a barrel of the West Texas Intermediate (WTI) crude oil for delivery on May 20 was traded at US$ 55.54, on the New York Mercantile Exchange (which is the largest physical commodity futures exchange in the world). At that time, the spot price of the WTI crude oil was US$ 54.09. This means, the cost of carry (i.e., the interest rate plus the storage cost minus the convenience yield) of the crude oil was 8.3 % per annum. One month before the delivery date (i.e. on April 20), the closing price of the May WTI crude oil fell to $ -2.60 and its spot price went further down to $ -36.98, which is unprecedented (see Figure 1).

Figure 1: Daily Spot and May Futures Prices of the WTI Crude Oil, Jan 24, 2020 - April 21, 2020.

First, the spot price of the WTI and its futures price both became negative, on April 20, and, second, they diverged suddenly. The law of supply and demand explains the decrease of the spot price of WTI crude oil into negative territory. The overproduction of crude oil (i.e., the increase in its supply) and the simultaneous drop in its demand due to the lockdown of economies worldwide result in the drop of its spot price. When he spot price of the WTI was $ -36.98, its futures price went as low as $ -39.44. The reasons for this important drop are: (1) there was no longer a convenience yield from holding inventories of crude oil compared to holding its futures contracts and (2) the storage cost of this commodity increased due to its overproduction. The divergence between the closing price of the May WTI crude oil and its spot price simply resulted from the fact that traders anticipated that the situation was temporary since the spot price of the alternative Brent crude oil was US$ 17.36 that day. This then caused the futures price of the WTI to rise.

Stock markets keep recovering from the crisis caused by the outbreak of the coronavirus. Between April 3 and April 10, the global financial turbulence score fell again, going from 7.7 to 4.1 (a 46.8 % decrease). As for the VIX, the implied volatility index, it went down from 41.67 to 38.15 (an 8 % decrease). The global financial turbulence score has been falling since the peak of March 13. One can wonder if the market bottom is reached, with this important fall.

Figure 2: Global Financial Turbulence Scores, Jan 8, 2000 - Apr 10, 2020.

Is the financial crisis over?

It is true that the high turbulence characterizing a financial crisis went down considerably. In my first post dedicated to the impacts of the coronavirus on the global economy [here], I predicted that the probability of a high turbulence in stock markets across the globe would decrease to 32 % by May 29. This probability remains unchanged, given the new available data. As one could see in Figure 2, the current level of the global financial turbulence score is still well above 3.4, which is the level expected during a turning point (represented by the green dotted line). By the end of the month of May, the probability of exiting the financial crisis would be 21.4 % and the probability of returning into it after a short recovery would be 33.7 %. Thus, the financial crisis is not over yet!

Dataset and Code

The Latest Global Financial Turbulence Scores.

Date	Score
Feb 14, 2020	1.36
Feb 21, 2020	6.23
Feb 28, 2020	3.68
Mar 6, 2020	9.74
Mar 13, 2020	12.89
Mar 20, 2020	9.37
Mar 27, 2020	8.03
Apr 3, 2020	7.70
Apr 10, 2020	4.10

The components indices of the global financial turbulence score

(1) NYA: the New York Stock Exchange composite index, (2) IXIC: the NASDAQ composite, (3) N225, the Tokyo Stock Exchange average index, (4) FTAS, the London Stock Exchange FTSE all share, (5) HSI, the Hong Kong Stock Exchange index, (6) N150, the Euronext Next 150 index, (7) GSPTSE, the Toronto Stock Exchange composite index, (8) BSESN, the Bombay Stock Exchange sensitive index, (9) GDAXI, the Frankfurt Stock Exchange performance index, (10) AXJO, the Australian Securities Exchange S&P 200, (11) SSMI, the SIX Swiss exchange mid-cap index, and (12) IBOVESPA, the Brazil Stock Exchange index.

Formula

$$d_t² = (r_t - μ ) Σ ^-1 (r_t - μ )',

where $d$ denotes the turbulence score, the vector $r$_t lists the current growth rates of the benchmark indices, the vector $$ μ their historical averages, and $$ Σ designates their variance-covariance matrix. For further details, see Mark Kritzman and Li Yuanzhen (2010).

<

The impacts of the coronavirus on the global economy: Part V Is a comparison with the Great Depression possible?

More and more commentators and economists are comparing the current economic crisis to the 1929 depression (also known as the Great Depression). For some of them, the current crisis is even worse. Are they right? In my humble opinion, they are wrong.

First of all, the Great Depression was caused by a stock market crash (that started on October 24, 1929). On the other hand, the current economic crisis is voluntarily induced by governments as a policy response to a global health crisis. Thus, the turbulence we are observing on stock exchanges around the globe is the consequence but not the cause of the current economic situation. The causes of the stock market crash of October 24, 1929 were mainly consumerism, easy credit, and speculation. Its consequences were mainly bank failures, hoarding, massive job losses, and poverty. Second, if the causes of the Great Depression and those of the current crisis are not the same, their consequences are not similar either. Currently, investors have fear, but they have not lost confidence in financial institutions. Millions of people are not currently employed, but they have not definitely lost their jobs. Third, the Great Depression lasted almost a decade (from 1929 till the outbreak of World War II), whereas the current crisis is hopefully coming to an end, since governments have started preparing plans to exit the lockdown restrictions they imposed some months ago.

It is true that the stock markets have been very turbulent since the outbreak of the coronavirus disease. On March 13, the situation on the major exchanges was the worst record over the past two decades (see Figure 1). Since then, it has been gradually improving. Between March 27 and April 3, the global financial turbulence score fell from 8.03 to 7.70 (a 4% decrease). The decrease in the VIX, the fear index, was much higher. This latter index went down from 46.8 to 41.67 (an 11% decrease). This situation is explained by the rise in the major benchmark indices. On the New York Stock Exchange, the NASDAQ, the Bombay Stock Exchange, the Toronto Stock Exchange, and the Brazil Stock Exchange, the benchmark indices increased by more than 10%.

Figure 1: Global Financial Turbulence Scores and VIX, Jan 8, 2000 - Apr 3, 2020

Oil markets are also suffering severely from the on-going crisis. On March 13, in the heat of the financial turbulence, the spot price of the West Texas Intermediate (WTI) and the Brent fell by 28.9% and 33.5%, respectively. (The WTI and the Brent are both sweet light crude oil serving as benchmark in pricing.) The following week they further declined by 25.3% and 23.9%, respectively. As one can see in Figure 2, the prices of these crude oil keep falling. This price crash is due to thee overproduction of oil and the decrease in its demand after the lockdown of several economies. This situation is affecting not only oil producing firms and the industries depending directly on this activity, but also public finance. In Canada, the provinces of Alberta and Newfoundland and Labrador are the most affected by this unexpected oil price crash.

Weekly Cushing, Oklahoma WTI and Europe Brent Spot Prices, Jan 1, 2000 - Apr 3, 2020

Dataset and Code

The Latest Global Financial Turbulence Scores.

Date	Score
Feb 14, 2020	1.36
Feb 21, 2020	6.23
Feb 28, 2020	3.68
Mar 6, 2020	9.74
Mar 13, 2020	12.89
Mar 20, 2020	9.37
Mar 27, 2020	8.03
Apr 3, 2020	7.70

The components indices of the global financial turbulence score

(1) NYA: the New York Stock Exchange composite index, (2) IXIC: the NASDAQ composite, (3) N225, the Tokyo Stock Exchange average index, (4) FTAS, the London Stock Exchange FTSE all share, (5) HSI, the Hong Kong Stock Exchange index, (6) N150, the Euronext Next 150 index, (7) GSPTSE, the Toronto Stock Exchange composite index, (8) BSESN, the Bombay Stock Exchange sensitive index, (9) GDAXI, the Frankfurt Stock Exchange performance index, (10) AXJO, the Australian Securities Exchange S&P 200, (11) SSMI, the SIX Swiss exchange mid-cap index, and (12) IBOVESPA, the Brazil Stock Exchange index.

Formula

$$d_t² = (r_t - μ ) Σ ^-1 (r_t - μ )',

where $d$ denotes the turbulence score, the vector $r$_t lists the current growth rates of the benchmark indices, the vector $$ μ their historical averages, and $$ Σ designates their variance-covariance matrix. For further details, see Mark Kritzman and Li Yuanzhen (2010).

The impacts of the coronavirus on the global economy: Part IV

Between March 20 and March 27, the global financial turbulence score dropped from 9.4 to 8, which represents a 14.3% decrease. At the same time, the implied volatility index, VIX, also fell from 65.54 to 57.08 (a 12.9% decrease). This means that stock markets across the globe have actually become less volatile and investors have started fearing less about the future. The major stock markets, except Bombay Stock Exchange, are recovering. Could this mean the global financial crisis caused by the outbreak of the COVID-19 is nearing an end? i do not think so and here are the reasons.

First, the levels of the global financial turbulence score and the VIX are still high. As it appears in Figure 1, they are above the red horizontal lines, which represent their expected values during high volatility periods. These expected values are respectively 3.4 and 27.8. Second, most benchmark indices are still, at least, 14% below their levels of February 14, 2020. For example, on May 27, the New York Stock Exchange (NYSE) composite index was 29.3% below its level of February 14, the Next 150 index was 31.6% below, and the Bovespa Index was 38.8% below. Third, stopping the spread of the COVID-19 has necessitated inducing a recession by closing borders and some businesses, which government are not yet ready to reopen.

Figure 1: Global Financial Turbulence Scores and VIX, Jan 8, 2000 - Mar 28, 2020

Are there signs of flight to safety?

Quite often, high uncertainty in financial markets leads investors seeking less risk to prefer government bonds to stocks. The move of investment funds out of stocks into bonds that follows is referred to as flight-to-safety. The current financial crisis has not triggered any flight to safety from stocks to bonds. The reason is that both bond yields and stock prices have been falling. The 13-week treasury bill rate went down from 1.54% on February 14 to .06% on April 2. In these conditions, acquiring bonds does not appear to be a safer alternative investment opportunity.

Figure 2: NYSE Composite Index and the 13-Week Treasury Bill Rate, Feb 14, 2020 - April 2, 2020

Dataset and Code

---

The Latest Global Financial Turbulence Scores.

Date	Score
Feb 14, 2020	1.36
Feb 21, 2020	6.23
Feb 28, 2020	3.68
Mar 6, 2020	9.74
Mar 13, 2020	12.89
Mar 20, 2020	9.37
Mar 27, 2020	8.03

The components indices of the global financial turbulence score

(1) NYA: the New York Stock Exchange composite index, (2) IXIC: the NASDAQ composite, (3) N225, the Tokyo Stock Exchange average index, (4) FTAS, the London Stock Exchange FTSE all share, (5) HSI, the Hong Kong Stock Exchange index, (6) N150, the Euronext Next 150 index, (7) GSPTSE, the Toronto Stock Exchange composite index, (8) BSESN, the Bombay Stock Exchange sensitive index, (9) GDAXI, the Frankfurt Stock Exchange performance index, (10) AXJO, the Australian Securities Exchange S&P 200, (11) SSMI, the SIX Swiss exchange mid-cap index, and (12) IBOVESPA, the Brazil Stock Exchange index.

Formula

$$d_t² = (r_t - μ ) Σ ^-1 (r_t - μ )',

where $d$ denotes the turbulence score, the vector $r$_t lists the current growth rates of the benchmark indices, the vector $$ μ their historical averages, and $$ Σ designates their variance-covariance matrix. For further details, see Mark Kritzman and Li Yuanzhen (2010).

The impacts of the coronavirus on the global economy: Part III The Global Financial Turbulence Score and the VIX

I have taken another pulse of the stock markets by recalculating the global financial turbulence scores using the latest weekly benchmark indices and compared them to the VIX, a volatility index, published by the Chicago Board Options Exchange (CBOE). An option is a contract that gives its owner the right to buy (in the case of a call option) or to sell (in the case of a put option) an underlying asset (e.g., a stock) at a specified price and date. The CBOE is the world's largest options exchange. Figure 1, below, plots both the turbulence scores and the VIX. The VIX is also known fear index.

Figure 1: Global Financial Turbulence Scores and VIX, Jan 8, 2000 - Mar 21, 2020

The global financial turbulence score is backward-looking, as it is based on historical benchmark indices. On the other hand, the VIX is an implied volatility index, as it is the market expectation of the next 30-day fluctuations in the S&P 500 that results from solving numerically an option pricing model. While the VIX has the advantage of being forward-looking, it has the disadvantage of being derived from a theoretical model that might not always hold true empirically.

At the end of last week (i.e., on March 20), both the turbulence score and the fear index went down, respectively, from 12.9 to 9.4 and from 66.04 to 61.59. This means that, even if the global financial crisis caused by the outbreak of the COVID-19 is still going on, stock markets became less volatile and investors that were betting on the future evolution of stock prices by trading options also became somewhat less pessimistic. Since the outbreak of the COVID-19, the highest values of both the turbulence score and the fear index were recorded on March 13. (Recall that trading paused on March 9 and 12 on the New York Stock Exchange, as its benchmark S&P 500 plunged below the 7% threshold of the market-wide circuit breakers.) Both the turbulence score and the fear index date the on-going financial crisis back to February 21, where their values rose sharply, respectively, from 1.4 to 6.2 and from 17.08 to 40.11.

It appears in Figure 1 that there is a co--movement between the turbulence score and the fear index. In general, when stock markets are very turbulent, investors are also very pessimistic about the future. For example, during the 2007-08 financial crisis and the on-going health crisis, both time series reach a peak. However, during the oil price crash, the turbulence index quadrupled going from 2.74 to 11.57 on January 9, 2015, whereas the fear index only rose from 17.55 to 20.95 (which represents a 19.3% increase). On November 10, 2017, whereas the turbulence score was decreasing, the fear index rose. Figure 2 shows the scatter plot of both time series.

Figure 2: Scatter Plot of the Global Financial Turbulence Scores and the VIX, Jan 8, 2000 - Mar 21, 2020

The correlation coefficient between the turbulence score and the fear index is .64. However, as it appears in Figure 1, during quiet periods, the relationship between these two time series is not as strong as it is during turbulent periods.

Dataset and Code

---

The Latest Global Financial Turbulence Scores.

Date	Score
Feb 14, 2020	1.36
Feb 21, 2020	6.23
Feb 28, 2020	3.68
Mar 6, 2020	9.74
Mar 13, 2020	12.89
Mar 20, 2020	9.37

The components indices of the global financial turbulence score

(1) NYA: the New York Stock Exchange composite index, (2) IXIC: the NASDAQ composite, (3) N225, the Tokyo Stock Exchange average index, (4) FTAS, the London Stock Exchange FTSE all share, (5) HSI, the Hong Kong Stock Exchange index, (6) N150, the Euronext Next 150 index, (7) GSPTSE, the Toronto Stock Exchange composite index, (8) BSESN, the Bombay Stock Exchange sensitive index, (9) GDAXI, the Frankfurt Stock Exchange performance index, (10) AXJO, the Australian Securities Exchange S&P 200, (11) SSMI, the SIX Swiss exchange mid-cap index, and (12) IBOVESPA, the Brazil Stock Exchange index.

Formula

$$d_t² = (r_t - μ ) Σ ^-1 (r_t - μ )',

where $d$ denotes the turbulence score, the vector $r$_t lists the current growth rates of the benchmark indices, the vector $$ μ their historical averages, and $$ Σ designates their variance-covariance matrix. For further details, see Mark Kritzman and Li Yuanzhen (2010).

The impacts of the coronavirus on the global economy: Part II

On March 11, 2020, the World Health Organization declared the outbreak of the COVID-19 (i.e., the coronavirus disease) a pandemic, which means this epidemic has spread worldwide. Since then, the United States (US) suspended for a month all flights from mainland Europe and declared the state of emergency. Canada closed its borders to foreign nationals, except its permanent residents, diplomats, and US citizens. The European Union also locked down for a month its borders to all non-member countries. Throughout the world, schools and universities are closed, and mass gathering (including religious celebrations) are called off. The various measures taken to put an end to this global health crisis and the panic caused by the situation are affecting the global economy.

On March 9, trading on the New York Stock Exchange (NYSE) paused for 15 minutes, after an initial 7% decline in its benchmark S&P 500. (This halt is the first level of the market-wide circuit breakers, which are a set of three emergency mechanisms aiming at curbing rapid and massive panic selling of securities.) On March 12, the plunge of many benchmark indices reached levels unobserved since the Black Monday (i.e., October 19, 1987). For a second time, trading on both the NYSE and the Toronto Stock Exchange (TSX) paused temporarily, as the S&P 500 and the S&P/TSX composite fell by 9.5% and 12.3%, respectively. On March 16, these two benchmarks respectively fell by 12% and 9.9%, which triggered the first tier of the circuit breakers for the third time in eight days.

To measure the turmoil on stock exchanges, I suggested, in my post The Impacts of the Coronavirus on the Global Economy, the use of the financial turbulence score, which is a multivariate distance measure in standard units proposed by Mark Kritzman and Li Yuanzhen (2010). (In Statistics, the square root of this measure is known as Mahalanobis distance.) I will now be referring to the time series I produced in the above-mentioned post, as global financial turbulence score as it consists of capital gains/losses computed using the benchmark indices of 12 of the 20 largest exchanges in the world. In this post, I update this time series in order to keep following the situation.

The figure below plots the square root of the global financial turbulence weekly time series. Last week, due to the fact that the circuit breaker halted twice stock trading on the NYSE and the TSX, turbulence on the major exchanges was higher than the week before. On March 13, the level of the global financial turbulence score was 12.9, versus 9.7 during the week ending on March 6.

Global Financial Turbulence Scores, Jan 1, 2000 - Mar 14, 2020

Last week, the turbulence score on the major stock exchanges far exceeded 4.3, which is the level expected during high volatility periods. This is the highest score recorded over the reference period.

Many central banks (including the Federal Reserve Bank, the Bank of Canada, and the Bank of England) cut their key interest rates, to stimulate their economies. These emergency measures have not yet succeeded to eliminate panic from financial markets, since the economic activity is still paralyzed by the border restrictions and the imposition of self-isolation (or social distancing). On March 18, the NYSE halted stock trading, for a fourth time in two weeks. As a matter of fact, travel agencies and tour operators, the transportation and warehousing sector, the arts, entertainment and recreation sector, and the accommodation and food services sector are suffering severely from the restrictions imposed to stop the spread of the coronavirus. The stocks of listed companies operating in these sectors will keep losing value as long as investors are not seeing any prospect of profit.

Dataset and Code

---

The Latest Global Financial Turbulence Scores.

Date	Score
Feb 14, 2020	1.36
Feb 21, 2020	6.23
Feb 28, 2020	3.68
Mar 6, 2020	9.74
Mar 13, 2020	12.89

The components indices of the global financial turbulence score

(1) NYA: the New York Stock Exchange composite index, (2) IXIC: the NASDAQ composite, (3) N225, the Tokyo Stock Exchange average index, (4) FTAS, the London Stock Exchange FTSE all share, (5) HSI, the Hong Kong Stock Exchange index, (6) N150, the Euronext Next 150 index, (7) GSPTSE, the Toronto Stock Exchange composite index, (8) BSESN, the Bombay Stock Exchange sensitive index, (9) GDAXI, the Frankfurt Stock Exchange performance index, (10) AXJO, the Australian Securities Exchange S&P 200, (11) SSMI, the SIX Swiss exchange mid-cap index, and (12) IBOVESPA, the Brazil Stock Exchange index.

Formula

$$d_t² = (r_t - μ ) Σ ^-1 (r_t - μ )',

where $d$ denotes the turbulence score, the vector $r$_t lists the current growth rates of the benchmark indices, the vector $$ μ their historical averages, and $$ Σ designates their variance-covariance matrix. For further details, see Mark Kritzman and Li Yuanzhen (2010).

The Impacts of the Coronavirus on the Global Economy

The outbreak in December 2019 of the coronavirus disease (COVID-19) is affecting now the global economy. Concerns about this disease, which was first identified in the Chinese province of Hubei, and the measures taken to stop its spread are affecting travel agencies and tour operators, the transportation and warehousing sector, financial markets, the arts, entertainment and recreation sector, and public finance. As a matter of fact, while some businesses and administrations are facing stock shortages as a result of the restrictions on exports from China, the world's factory, other businesses are coping with multiple cancellations of reservations and events. Stock and oil prices are falling. Governments are revising downward the forecasts of their economies' growth and the estimates of their budget revenue. Some commentators are already talking about a recession or worse a depression. How serious is the situation?

In this post, I take a look at the current situation in some major financial markets and compare it to historical data in order to find out if, actually, there are reasons to fear the worst. Figure 1 plots 12 benchmark indices of the following markets: (1) the New York Stock Exchange, (2) the NASDAQ, (3) the Tokyo Stock Exchange, (4) the London Stock Exchange, (5) the Hong Kong Stock Exchange, (6) the Euronext, (7) the Toronto Stock Exchange, (8) the Bombay Stock Exchange, (9) the Frankfurt Stock Exchange, (10) the Australian Securities Exchange, (11) the SIX Swiss exchange, and (12) the Brazil Stock Exchange.

Figure 1: Natural Logarithm of Some Weekly Stock Market Benchmark Indices, Jan 1, 2000 - Mar 7, 2020

How serious is the situation?

It appears clearly in Figure 1 that stock prices have been declining over the past weeks, in all these 12 major exchanges. Particularly, on January 24 and on February 21 of this year, these benchmark indices went down simultaneously, when the markets were closing. To take a measure of the situation using a single summary statistic instead of looking at 12 time series individually, I have computed financial turbulence scores following Mark Kritzman and Li Yuanzhen (2010). This statistics is given by the following relation

$$d_t² = (r_t - μ ) Σ ^-1 (r_t - μ )',

where the vector $r$_t lists the current growth rates of the benchmark indices (the capital gains or losses), the vector $$ μ their historical averages, and $$Σ designates their variance-covariance matrix.

Figure 2 plots the square root of the turbulence score (i.e., $the\; statistic$d_t ) computed using the benchmark indices of the 12 major exchanges listed above. The two horizontal lines on this figure are thresholds defining three regions: (1) in the region below 2.5 (demarcated by the green line), the global financial market is in a quiet state, (2) in the region above 4.1 (demarcated by the red line), the market is very turbulent, (3) in-between, there is an unconditional probability of 82% that the market be in a quiet state. For information, these estimates have been produced fitting a Markov-switching model to the turbulence statistics.

Figure 2: Financial Turbulence Score Based on 12 Major World Indices, Jan 7, 2000 - Mar 7, 2020

The two highest turbulence scores observed since the outbreak of the coronavirus are 6.2 (on February 21) and 9.7 (on March 6). As one could see in Figure 1, this indicates that the global financial market is currently in turbulence. But, contrary to what we were led to think, this only started on February 21. Furthermore, the situation is not comparable to the episodes of turbulence experienced during the burst of the dot-com bubble in the early 2000s, the financial crisis of 2007-08, or the oil crisis of 2015.

What to expect in the coming weeks?

An accurate answer depends on how the COVID-19 will evolve. While encouraging signs are coming from the province of Hubei in China where the virus was first identified, warning signs are coming from Italy where this mortal virus is spreading. Using the historical data, I can predict that there is a very high probability that the current state of the global financial market remain the same in the next 12 weeks. However, this probability decreases progressively going from 88% for the week ending on March 13 to 32% for the week ending on May 29.

Dataset and Code

---

The Latest Global Financial Turbulence Scores.

Date	Score
Feb 14, 2020	1.36
Feb 21, 2020	6.23
Feb 28, 2020	3.68
Mar 6, 2020	9.74

The components indices of the global financial turbulence score

(1) NYA: the New York Stock Exchange composite index, (2) IXIC: the NASDAQ composite, (3) N225, the Tokyo Stock Exchange average index, (4) FTAS, the London Stock Exchange FTSE all share, (5) HSI, the Hong Kong Stock Exchange index, (6) N150, the Euronext Next 150 index, (7) GSPTSE, the Toronto Stock Exchange composite index, (8) BSESN, the Bombay Stock Exchange sensitive index, (9) GDAXI, the Frankfurt Stock Exchange performance index, (10) AXJO, the Australian Securities Exchange S&P 200, (11) SSMI, the SIX Swiss exchange mid-cap index, and (12) IBOVESPA, the Brazil Stock Exchange index.