Consumer credit usage in Canada during the coronavirus pandemic (2024)

  • Journal List
  • Wiley - PMC COVID-19 Collection
  • PMC9111881

As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsem*nt of, or agreement with, the contents by NLM or the National Institutes of Health.
Learn more: PMC Disclaimer | PMC Copyright Notice

Consumer credit usage in Canada during the coronavirus pandemic (1)

Link to Publisher's site

Can J Econ. 2022 Feb; 55(Suppl 1): 88–114.

Published online 2022 Feb 11. doi:10.1111/caje.12544

PMCID: PMC9111881

Language: English | French

Anson T. Y. Ho,1 Lealand Morin,2 Harry J. Paarsch,2 and Kim P. HuynhConsumer credit usage in Canada during the coronavirus pandemic (2)3

Author information Copyright and License information PMC Disclaimer

Associated Data

Supplementary Materials


The recent COVID‐19 pandemic has devastated economies worldwide. Using detailed, monthly data from a major consumer credit reporting agency in Canada, we have examined individuals’ use of credit cards and home‐equity lines of credit (HELOCs). We found a dramatic leftward shift in the distribution of credit card and HELOC outstanding balances, providing evidence for a widespread reduction in credit usage. Our findings suggest that, during the COVID‐19 recession, Canadian consumers were able to meet their financial needs without increasing their debt burdens. These results complement other findings concerning a decline in consumer spending and the results of government assistance programs, and imply that the economic consequences of this pandemic are very different from those in other recessions.


Recours au crédit à la consommation au Canada au cours de la pandémie de coronavirus. La récente pandémie de COVID‐19 a ravagé les économies mondiales. Grâce aux données mensuelles détaillées d'une des agences d'évaluation du crédit les plus importantes du pays, nous avons examiné le recours aux cartes de crédit et aux marges de crédit hypothécaires des Canadiens. Sur les figures illustrant la distribution des soldes impayés de cartes de crédit et de marges de crédit hypothécaires, nous constatons un important déplacement vers la gauche, ce qui reflète un fléchissem*nt généralisé du recours au crédit. Nos résultats suggèrent qu'au cours de la récession provoquée par la COVID‐19, les consommateurs canadiens ont pu satisfaire leurs besoins financiers sans alourdir leurs dettes. Ces conclusions complètent d'autres observations, notamment sur la baisse des dépenses de consommation ou sur l'effet des programmes d'aide du gouvernement, et laissent entendre que les conséquences économiques de cette pandémie sont très différentes de celles observées dans d'autres types de récession.

1. Introduction and motivation

The 2019 novel coronavirus (COVID‐19) pandemic has devastated economies worldwide.1 It has plunged the global economy into the deepest recession since the Second World War, with substantial economic uncertainty (World Bank2020, OECD2020b). Understanding the economic impacts of COVID‐19 on households’ financial health is imperative to economic policy‐making. Canada, as one of the economies with the highest levels of household debt, serves as an important case study. In 2019, Canadians’ household‐debt‐to‐disposable‐income ratio reached 181.2%, which was the highest among G7 countries and the third highest among G20 countries (OECD2020a).2 Elevated household indebtedness has been identified as a key vulnerability in the Canadian financial system (Cateau etal.2015).

The pandemic has significantly affected the Canadian economy, with the unemployment rate surging from 5.6% in February 2020 to 13.7% in May 2020. The Canadian government has responded with various aid programs, as outlined in the Economic Response Plan.3 For individuals, the Canada Emergency Response Benefit (CERB) was launched on March 15, 2020, to provide a taxable benefit of $2,000 every four weeks to employed and self‐employed workers whose employment was affected by COVID‐19.4 According to the CERB benefits report5 CERB has paid out about $81.6 billion to 8.9 million applicants, most of whom earned under $47,630 in 2019.6 Since April 2020, financial institutions, at their discretion, have also provided loan payment deferrals to clients who self‐declared to have been negatively impacted by the pandemic. The Financial Consumer Agency of Canada(2020) shows that the vast majority of the deferral requests were approved and that this accounted for a significant fraction of the outstanding loans; see Vallée(2020) and Allen etal.(2022).

Using anonymized monthly data from a major consumer credit reporting agency in Canada, we have investigated the financial positions of consumers—in terms of their outstanding loan balances relative to their pre‐pandemic positions. We did so by forecasting the credit usage of consumers using our estimated law of motion for debt balances, which we identified in the period before the pandemic, and then compared our forecasts with actual credit usage during the pandemic. Through this analysis, we have established important facts about consumers’ use of credit during the pandemic, without seeking to evaluate the effectiveness of any specific assistance program.

Our analysis focuses on two popular credit instruments: credit cards and home equity lines of credit (HELOCs). Both instruments provide individuals with fast and flexible access to credit. Account holders can use any credit amount within their pre‐approved credit limits, and interest is charged only on the portion that is actually used.7 Repayment is flexible: while the minimum payment for credit cards is usually 1% to 3% of the outstanding balance, the required monthly payment on a HELOC is usually the interest accrued on the outstanding balance. Interest rates on credit cards and HELOCs, however, sit at two extremes. Credits cards provide unsecured lending with interest rates of about 20% (Mintel2018), among the highest of all credit instruments. HELOCs, on the other hand, are secured by home equity, with interest rates of about 2.7% since the pandemic began.

We conducted our analysis using a variant of intervention analysis, developed by Boxand Tiao(1975), and model the law of motion of the distribution of loan balances with a first‐order Markov process.8 We used this framework to contrast what happened with what was predicted—interpreting the difference as the excess effect of the pandemic. Our framework is a flexible approach that accommodates features of the data, such as boundaries and gaps, that are otherwise problematic for standard time‐series methods.

Using this framework, we have found dramatic decreases in individuals’ credit card debt. Since the onset of the COVID‐19 pandemic, we found a much higher proportion of consumers with credit card balances lower than $1,000, relative to that predicted by our forecasts, and a declining proportion of consumers with credit card balances at all levels greater than $1,000. Of consumers with loan balances in the range from $4,000 to $10,000, we found an astounding decline of 25% in the proportion. The decline attenuated but persisted after the economy gradually reopened since late June 2020. We have found similar patterns in HELOC usage. The proportion of consumers with HELOC balances in the range of $5,000 to $10,000—the category most affected—declined 15%, relative to our forecasts. The decline tapers off to categories with loan balances up to $300,000, beyond which HELOCs are used mainly for home purchases in combined mortgage‐HELOC plans; see Al‐Mqbali etal.(2019).

Our results suggest that Canadian consumers were able to meet their financial needs without increasing their debt burdens. These findings complement other documented facts in Canada concerning job losses, CERB benefits and the pattern of decreases in consumer spending. Since the early stages of COVID‐19, the Canadian labour market had shifted to its maximum telework capacity at 39.1% (Deng etal.2020). Households with higher levels of education and earnings were more likely to keep their jobs and work remotely from home (Messacar etal.2020). Low‐income individuals, who were also at higher risk of unemployment and financially more vulnerable, have been significantly supported by government programs. The Bank of Canada(2020) estimates, in their Monetary Policy Report, that the CERB was able to replace a major fraction—in some cases over 100%— of wage earnings for individuals below the 2019 median income. The decline in consumption spending mainly occurred in categories not amenable to physical distancing, such as restaurants and recreational services, which is consistent with the substantial drop in credit card balances that we observed.

This paper complements a rapidly developing literature concerned with the effects of COVID‐19 on consumer finance around the world. In the United States, Sandler and Ricks(2020) showed that consumers in the United States did not rely on credit card debt for financial liquidity in the early stages of the COVID‐19 pandemic because the average balance decreased by about 10%. Nagypál etal.(2020) found that credit card applications declined in regions with more unemployment insurance claims. These findings regarding credit card usage are consistent with other research showing a decline in spending. Chetty etal.(2020) and Baker etal.(2020) reported that consumers reduced their overall spending and liquidity‐constrained households responded rapidly to the fiscal stimulus payments from the 2020 CARES Act.9 Similar effects on consumer spending are also found in Denmark and Spain; see Andersen etal.(2020) and Carvalho etal.(2020), respectively. A common finding across these studies is a significant switch in spending categories; spending decreases were stronger in more restricted sectors, which may arise because of a short‐run aggregate supply shock from economic lock‐downs. In addition to spending changes, consumers also increased cash holdings; see Chen etal.(2020a). Our analysis contributes to this literature by not only providing evidence of reduced credit usage in Canada but also analyzing this change across the distribution of credit usage.

Our research is also related to a long‐standing literature on debt utilization in the presence of negative income shocks. Hurst and Stafford(2004), Chen etal.(2020b) and Agarwal etal.(2006) note that wealthy homeowners tend to borrow against their home equity through mortgage refinancing or HELOCs, but consumers with few assets smooth their unemployment shocks via unsecured debts. Agarwal and Qian(2017) reported that homeowners access credit cards when they experience unexpected reductions in home equity. Conversely, Ganong and Noel(2019) showed that households do not borrow more from credit cards during unemployment spells. The poorest often have no access to credit at all; see Sullivan(2008).

The remainder of the paper is organized as follows: In section2, we describe the source of our data, a major consumer credit reporting agency in Canada. In section3, we report our empirical results, introducing features of our empirical framework along the way. We also validate our method on a recent economic downturn—an unanticipated period of low oil prices that shocked residents of the province of Alberta, Canada, in 2015 and 2016.

2. Data from a major consumer credit reporting agency

We acquired access to anonymized consumer‐credit data maintained by TransUnion®, one of the major consumer‐credit reporting agencies that operate in Canada. The data are reported monthly, which admits a fine analysis of the effects of COVID‐19. Our dataset contains account‐level information concerning consumers’ outstanding balances, which we aggregated to the individual level. This represents the total amount owed by a consumer at the end of each month. For credit cards, it is the total credit used regardless of whether it is paid off or carried forward to the future. Henry etal.(2018) estimated that about 30% of Canadian credit card holders have positive revolving balances.

Various types of financial institutions are included in the dataset—ranging from chartered banks of different sizes to credit unions and credit card companies. Major loan types are included, such as mortgages, HELOCs, credit cards, automobile loans and other instalment loans.10 When cleaning the data, accounts that were in debt collection, being written off or included in consumer proposals or bankruptcy filings were excluded.11 Accounts that have not been updated for 90 days (one quarter) or that had missing information concerning outstanding balance were also dropped.12

Our dataset is representative of most of the Canadian population. The aggregate balances held at chartered banks in our data closely match with those from the regulatory filings collected by the Bank of Canada. Using 2016 Census data, we estimated that some 93.9% of the Canadian population aged 20 and older is represented in the data from the credit reporting agency. A detailed description of our data validation is provided in appendixA. We drew a 1%, random sample from the entire set of credit card holders, a sample of 292,854 unique account holders with 12,743,957 monthly observations. For HELOCs, we drew a 5%, random sample of HELOC holders to obtain a sample of 303,400 unique account holders with a total of 13,841,400 monthly observations.

The time series of individuals’ outstanding balances are reported in figure1. The solid line in figure1(a) depicts the mean individual credit card balances over the sample period from January2017 to August2020, while the dashed and dotted lines depict the inner quartiles. During the period January2017 to January2020, on average, the outstanding balance exhibited a year‐over‐year growth rate of around 2.7%. Since then, the average dropped to a new low in May 2020. The shift also appears in the median and the third quartile. A seasonal pattern exists: higher average credit card balances obtain in the fourth quarter of every year (the holiday season), and the aggregate balances decrease in the first quarter every year. The seasonal pattern is more pronounced for the third quartile, indicating that much of the activity is in the upper tail. The time series of HELOC balances are depicted in figure1b: a gradual decline in HELOC balances has occurred since the summer of 2017, but the decline was largest during the COVID‐19 pandemic, with the exception of a temporary increase in July 2020.

Open in a separate window


Consumers’ outstanding balances, 2017–2020

In order to model a first‐order Markov process, we assigned these balances to discrete balance categories, with one category for consumers having zero balances. Instead of using evenly‐spaced bins, individual's loan balances were organized into intervals of declining width to account for the lengthy tail of the distribution and to correspond (roughly) to typical categories of credit lines offered by banks. For credit cards, balance categories are defined in intervals of width $250 up to $1,500, intervals of $500 up to $6,000, $1,000 up to $10,000, $2,000 up to $20,000, $5,000 up to $30,000, one $30,000 to $40,000 category and a category of $40,000 and above in the tail. We defined balance categories for HELOCs in a similar fashion, collecting those with zero balances in the first category, and we defined the next two categories to have widths of $2,500, up to $5,000. We then divided the next three categories with intervals $5,000 wide, up to $20,000, and follow those with intervals of $10,000, up to $50,000. Next, we used intervals of $25,000 up to $100,000, intervals of $50,000 up to $300,000, one category from $300,000 to $400,000 and the remaining categories with width $200,000 up to $800,000 followed by a category with balances of $800,000 and above.

Any empirical framework designed to measure the impact of the COVID‐19 pandemic on credit card and HELOC balances must accommodate four main features of those data: first, the distribution is highly skewed and the changes in the distribution of balances is more pronounced in the tails. In figure1, there is little movement in the first quartile, but the third quartile hovers near the mean. During the pandemic, the third quartile of credit card balances dips below the mean. This suggests that we must use an empirical framework that focuses on the distribution, rather than the mean.

Second, a substantial fraction (over 10%) of individuals in the sample have zero balances. The case of credit cards is depicted in figure2(a) using data from January2017 to January2020. Compare this to the histogram for HELOCs, in figure2b: more than 30% of these credit lines have balances of zero, which is nearly double the proportion for credit cards.

Third, from month to month, considerable dependence exists over time; for example, if an account had a zero balance at the end of last month, then it is very likely (in general over 50% of the time) to have a zero balance at the end of this month. Figure3 depicts the joint empirical distribution of period (t−1) (denoted by “Past”) and period t (denoted by “Current”), using data from January2017 to January2020. Panel3(a) depicts the conditional histogram of current balance categories for credit cards, and panel 3(b) depicts that for HELOCs. Most of the transition probability mass is on the diagonal elements, indicating even stronger dependence on past values for credit lines of all sizes.

Open in a separate window


Conditional histograms of individuals’ balances

Fourth, not every consumer is observed in every period of the sample: consumers sometimes have gaps in their data.13 This would complicate the analysis of the data were we to use standard panel‐data techniques. Our framework is robust to this feature of the data: we can calculate estimates of the relevant transition matrices using a large sample of pairs of consecutive balances—whether or not the observations are recorded for all consumers every month.

3. Empirical results

Borrowing from the demography literature in which excess deaths are a measure of the effect of, say, a pandemic, we introduce the term “excess balances” for the notion that balances have increased, relative to previous experience, as a result of the pandemic; see Statistics Canada(2020). In the analysis that follows, we investigate excess balances from a number of perspectives. In the first, we compare the histograms of loan balances that occurred in 2020 against those in previous years. Because seasonality is observed in the aggregates series, we first compare histograms during the pandemic with the histograms from the same months of the previous three years. To make this comparison, we develop a statistic to measure any changes in the distributions. Next, we use this statistic to develop a graphical framework within which to view the changes in terms of the proportion of consumers in each balance category.

The analysis of histograms is only an approximation of the change, however, because it ignores any growth in balances year‐over‐year. For this reason, we apply an empirical framework to estimate transition matrices to describe the law of motion of the vector of proportions of consumers in each balance category. Using this information to condition on the state before the pandemic, a more precise measurement of the counterfactual distribution can be obtained. With this set of forecasted distributions as a benchmark, we can then re‐evaluate changes in the distributions of balances to measure changes in the high‐balance categories more accurately than before—in a way that captures the notion of excess balances. These complement the results from analyzing histograms alone, which we discuss next.

3.1. Comparisons of histograms

Let p~k,T+ denote the observed vector of proportions of consumers with a balance in category k at time (T+ℓ), which is ℓ months after some reference period T before the pandemic. Similarly, p^k,T+ denotes the observed vector of proportions of consumers in the benchmark distribution. In terms of the analysis of histograms, p~k,T+ is represented by the histograms for activity in 2020, while p^k,T+ is represented by the histograms calculated from the sample period from 2017 to 2019. In principle, p^k,T+ could be calculated as the forecast from any model, and the sample histogram is the first such model we consider.

As a summary measure of the difference between these histograms, we consider following the statistic:



which represents the percentage difference in cell k=0,1,…,K between what actually obtained during the pandemic relative to what appeared in the histograms in the sample period. This object can be plotted on the ordinate versus the various cells on the abscissa to provide the reader with a visual description of how balances (however measured) have changed relative to what would have been predicted before the pandemic struck. The statistic



measures the weighted average of the percentage change in the proportion of consumers in each balance category, weighted by the proportions p^k,T+, which are prescribed under the null hypothesis that the proportions are the same in both samples. The N in the definition of the statistic normalizes the random part to have a stable limiting (asymptotic) distribution.

This statistic, first proposed by Kullback and Leibler(1951) and commonly referred to as the Kullback–Leibler divergence criterion, was built on a concept of information introduced by Shannon(1948). As shown in Belov and Armstrong(2011), for a pair of continuous distributions, a version of this statistic has a limiting (asymptotic) χ2 distribution with one degree of freedom. Parkash and Mukesh(2013) investigated the case of discrete distributions, which corresponds to our application, and determined that the statistic has a limiting χ2 distribution, but with K degrees of freedom—that is, the number of categories minus one. Song(2002) demonstrated that the Kullback–Leibler divergence statistic is asymptotically equivalent to the likelihood‐ratio statistic for detecting a difference between distributions.

As a method for discriminating between distributions, the Kullback–Leiblerdivergence criterion is more sensitive to small changes in proportions in categories with low probability mass, which is important for comparing skewed distributions. Using simulation methods, Ho etal.(2020) demonstrated the validity of this variant of the Kullback–Leibler divergence statistic, showing that the statistic follows the χ2 distribution with K degrees of freedom, as expected. The asymptotic distribution obtains with sample sizes on the order of 10,000 observations, which is well within the size of our cross‐section of about 300,000 consumers.14

Using histograms from the sample period as the benchmark for comparison, we calculate the statistic in equation(2). The p‐value columns show the probability of observing a statistic more extreme from the χ2 distribution with 28 degrees of freedom for credit cards and 18 degrees of freedom for HELOCs. From the magnitude of the divergence statistics measured in table1, the distributions in 2020 are clearly different from those in the past. The change is apparent as early as January 2020, well before the pandemic hit Canada. During April, the first complete month of lock‐down, there appears a much larger difference in the distribution of balances for credit cards. This difference expands in May, recedes slightly in June and appears to level off over July and August, with a large difference still remaining by August.


Divergence from sample histograms

Credit cardsHELOCs
January 2020272.05<0.0000e‐16662.07<0.0000e‐16
February 2020286.49<0.0000e‐16656.63<0.0000e‐16
March 2020156.57<0.0000e‐16662.56<0.0000e‐16
April 20203,809.09<0.0000e‐16828.96<0.0000e‐16
May 20206,803.32<0.0000e‐161,158.38<0.0000e‐16
June 20204,376.69<0.0000e‐161,453.77<0.0000e‐16
July 20202,682.10<0.0000e‐161,706.15<0.0000e‐16
August 20202,289.10<0.0000e‐161,507.30<0.0000e‐16

Open in a separate window

The difference in April is less pronounced for HELOCs, with gradual changes through July, again, receding slightly by August 2020. This is in contrast to the abrupt jumps noted for credit card balances. For both types of loans, there exists a large difference continuing from May that persists until August. Having detected a statistically significant change in the distributions, we document the proportional changes in each of the balance categories.

3.1.1. Credit cards

The changes noted in the distributions of credit card balances are depicted in figure4, compared with balances in the same months during the three‐year sample drawn from the years 2017 to 2019. For February and March, the reduction in the proportion of zero balances, relative to the previous years, are clearly shown in the first bars on the left of each panel. This shows a 5% reduction in the proportion of consumers with a balance of zero on their credit card. On the other extreme, for balances in the tens of thousands, a substantial increase is evident for February of 2020, relative to the previous years. This pattern persists into March of 2020. The increase in the proportion of high‐balance consumers was not readily apparent in the histograms shown above and explains the large values of the divergence statistic in the first quarter of 2020.

Open in a separate window


Deviations from histograms (credit cards)

The effects of the pandemic emerged during the month of April: figure4(c) depicts the change in the proportion of consumers in each balance category during the month of April 2020. A larger fraction of the population exists with balances below $1,000 and a smaller fraction exists with balances between $1,000 and $25,000 than before. A substantial increase in the proportion of consumers with credit card balances in excess of $40,000 also exists. In this simple comparison, the change in distributions is confounded with the rising incidence of very high balances over the previous years: a pattern that continues from February before the pandemic.

The changes noted in April 2020 persist through May and June: An increase in the proportion of consumers with credit card balances below $1,000 and a decrease in the proportions for higher balance categories also exist. If the increase in the proportion of consumers with balances in the tens of thousands had persisted from January beyond March 2020, then the reduction in the proportion of consumers with high credit card balances may be understated for April and May.

3.1.2. Home‐equity lines of credit

An investigation of home‐equity lines of credit (HELOCs) exposes another source of complexity, namely, a pattern of changes in balances that continues through the months of February to May 2020 (figure5). An increase in the proportion of HELOCs with balances of zero, a reduction in the proportions for positive balances below $400,000, and an increase in higher balance categories occurred. From March to May, the differences from the sample histograms closely match those in February. This suggests that the month‐to‐month variation during the pandemic is obscured by the year‐over‐year changes that arise from comparing histograms measured during different years. This combination of activity makes it difficult to discern any pattern of changes connected to the COVID‐19 pandemic. As in the investigation of credit card balances, the balances of HELOCs must be modelled more accurately in order to characterize adequately the changes attributable to the pandemic.

Open in a separate window


Deviations from histograms (HELOCs)

3.2. Comparisons with forecasts

Although the comparison with histograms from the sample period does account for seasonality, it does not take into account the trend of increasing balances, nor does it condition on the state of these credit markets before the pandemic. Obtaining an accurate comparison requires that we estimate a forecast of the proportions of consumers in each balance category.

How do we distinguish between dependence in the data and the effect of the pandemic on excess balances? We estimate transition matrices P^ for each calendar month during the sample period. Because the months of the year are represented by mutually exclusive indicator variables, the transition matrices are easily concentrated out and estimated separately. Furthermore, the kth rows of P^ are also estimated separately, using only the observations with past balances in category k, which is implemented by calculating the histograms of balances across the categories in the subsequent month, using the relevant observations selected by the previous balance category.

Following this approach, we estimate these transition matrices up to the date before the pandemic—in this case, January2020. Then, based on those estimates, we calculate the one‐, two‐, three‐, up to the seven‐month ahead forecast distributions according to the Markov model:



We then compare the distributions generated by equation(3) with what actually obtained: in our notation, p~T+1, p~T+2, p~T+3, p~T+4 and so on, using the statistic in equation(2).

The only change from equation(2) above is the estimation of the transition probability matrix, the rows of which are consistently estimated by the sample histograms. Consequently, the forecast distribution, for which the transitions are left‐multiplied several times, is also consistently estimated, assuming the correct model specification for the transition matrices. With the large sample that we have, the difference from sampling variation is minimal.

Each transition is calculated using the matrix P^ that corresponds to the particular month. Were it not for the strong seasonality in the distribution of balances, one could use a simpler model in which the transitions are calculated using a single, fixed transition matrix throughout the year. Such a fixed transition matrix would have 28×29=812 parameters, with 29 categories of credit card balances in the discretized model. In contrast, the model that we estimate, with monthly transition matrices, has 12 times as many—some 9,744 parameters.

To detect whether a difference exists between the observed distributions and the forecasted ones, we perform another series of tests. In the first set of tests, we compare the benchmark with an ℓ‐step‐ahead forecast. We initialize the forecasts, p~T, with the proportion of consumers in each balance category that is observed on 1January2020, reflecting activity recorded during the month of December 2019.15 We then calculate a series of forecasts p^T+ by left‐multiplying the vector with the transition matrices for each month, P^. The forecast for the next month (ℓ+1) is calculated in the same way, using the forecast from the period before: p^T++1=P^+1p^T+.

The results of this series of comparisons for the model are collected in table2. As before, the statistic is the Kullback–Leibler divergence statistic, except this time, the statistic is used to compare the ℓ‐step‐ahead forecasted distribution with the observed sample distribution, as defined in equation(2). The p‐value columns show the probability of observing a statistic more extreme from the χ2 distribution with 28 degrees of freedom for credit cards and 18 degrees of freedom for HELOCs.


Divergence from ℓ‐step‐ahead forecasts

Credit cardsHELOCs
February 202044.392.5411e‐0225.701.0681e‐01
March 202071.181.2710e‐0585.808.1259e‐11
April 20204,436.14<0.0000e‐16221.97<0.0000e‐16
May 20207,773.38<0.0000e‐16534.41<0.0000e‐16
June 20205,456.78<0.0000e‐16926.60<0.0000e‐16
July 20203,846.25<0.0000e‐161,281.55<0.0000e‐16
August 20203,274.21<0.0000e‐161,352.55<0.0000e‐16

Open in a separate window

The divergence statistics up to March do not wander far from those expected under the null hypothesis of no change, especially considering that they are calculated using tens of millions of observations. At the end of April, a substantial change in the distribution of credit card balances exists; this difference is even more pronounced for balances at the end of May. The divergence declines slightly from June to August, but settles at a level that suggests the distribution is still far from that forecasted from before the pandemic. For HELOCs, the change in the distribution occurs gradually; it begins to shift in April and continues through August, although the change is not as extreme as that for credit cards.

3.2.1. Credit cards

Using forecasts calculated with separate monthly transition matrices, we show the deviations in probability mass in figure6, where each bar in the figure is provided in equation(1). In April, we find a much higher proportion of consumers with credit card balances less than $1,000, relative to that predicted by our forecasts. The deviations from forecast with May balances are even more extreme: we find more than 20% more consumers than forecast with balances from zero to $250. Conversely, we find a declining proportion of consumers with credit card balances at all levels greater than $1,000, and a remarkable 25% decline in the proportion of consumers with balances in the range from $4,000 to $10,000. These changes persisted through July and August, even after the economy gradually reopened in late June, with a decline of roughly 20% in all balance categories above $6,000.

Open in a separate window


Deviations from forecasted credit card balances

3.2.2. HELOCs

For HELOCs, smaller percentage changes exist in the distribution from forecast when compared with those for credit card balances, as shown in figure7. The category with the greatest proportional change is represented by the fourth bars with balances between $5,000 and $10,000. The proportion of consumers in this category dropped by 10% in May, 13% in June, 16% in July and 15% in August. In the remaining four of the lowest six categories, balances between $10,000 and $20,000, in the second, third, fifth and sixth bars, the proportion decreased by about 7% in May and 10% in June, remaining in this range until August. The proportion of consumers with balances in most of the remaining categories, from $20,000 to $350,000, decreased by a smaller percentage, between 5 and 10%.

Open in a separate window


Deviations from forecasted HELOC balances

Much of the activity in the distribution of HELOC balances is concentrated in the low and high balances: many consumers paid off their HELOC loan entirely and there exists substantial variation in the highest balance categories from month to month. This suggests increased activity in the flows into and out of the HELOC market. The observed decline in HELOC balances coincides with reduced consumer spending during the lockdown period, as confirmed by the Financial Consumer Agency of Canada(2018) report, showing that 49% of HELOC holders used their accounts to pay for home renovations, with a median amount of $10,000. In contrast, a sharp increase in borrowing activity is observed in the three categories with balances above $400,000 in July. This spike is due to a surge in home sales—a 30.5% increase in home sales—the highest recorded increase in the last 40 years, supported by historically low financing costs, resilient incomes for higher‐earning households, and pent‐up demand from delayed spring activity; see Bank of Canada(2020). This episode was short‐lived, however, with decreases in all non‐zero balances categories by August.

3.3. Comparison with the oil price shock in Alberta, 2014–2015

As a point of comparison, we conduct a similar analysis with a sample from the petroleum‐rich province of Alberta during the oil price shock in 2015. During this episode, in late 2014 and all of 2015, oil prices declined sharply and pushed the province into a recession. In late 2014, oil prices dropped from over $100 per barrel in July to nearly $45 by January 2015. Prices then declined to a low of just under $30 by the beginning of the next year, 2016. Consequently, the unemployment rate in Alberta increased from 4.2% in December 2014 to 6.4% in December 2015.

Consumers in Alberta were pushed into financial stress; a notable increase in credit card balances occurred in the province. Figure8(a) illustrates that, following the oil price decline, individuals’ credit card balances on average increased by about 10.1% from $4,215 at the end of 2014 to $4,657 at the end of 2015. We use this example as a benchmark—to compare what happened during the COVID‐19 pandemic. In contrast, a different scenario was unfolding in HELOC usage in Alberta over the same period. Figure8(b) illustrates that the aggregate HELOC balances were in decline during the oil price shock.16

Open in a separate window


Consumers’ outstanding balances, Alberta, 2012–2016

To place a proper event window around the initial period of declining oil prices, we restricted the in‐sample period to the year 2013. This allowed for a balanced sample size for estimating each of the monthly transition matrices. The year 2015 was reserved as the out‐of‐sample period over which we measure the impact of the oil price shock in Alberta, a horizon similar to that of the COVID‐19 pandemic.

Our sample was drawn randomly from the years 2013 to 2015—an in‐sample period of 24 months over the years 2013 to 2014. For credit cards, we drew a 10% sample in Alberta that includes 366,794 unique consumers with at least one opened credit card account, collectively the sample contains 18,286,464 monthly observations. For HELOCs, we included 40% of the HELOC holders in Alberta, which contains 330,019 unique consumers with a total of 18,427,526 monthly HELOC observations.

Loan balances were organized into the same intervals as in our analysis of the period during the COVID‐19 pandemic. We estimated the transition matrix by estimating a histogram for each row, conditioning on the balance category in the previous month, using the sample period from January 2013 to December 2013. We then calculated the ℓ‐step‐ahead forecast of the discrete distribution for the subsequent months by applying repeated multiplication of the transition matrix, starting from the discrete distribution in December 2014. We calculated the Kullback–Leibler divergence statistic to measure the distance between the forecast probability vector and the actual one observed for each month. The test results were collected in table3. For both credit cards and HELOCs, the distributions gradually shifted away from the distributions in December 2014. The divergence of the HELOC distribution is similar in magnitude to that during the COVID‐19 pandemic. In contrast, the shift in the distribution of credit card balances was less pronounced than that during the pandemic, but still a significant divergence from the distribution at the beginning of the year.


Divergence from ℓ‐step‐ahead forecasts in Alberta, 2015

Credit cardsHELOCs
January 201571.701.0733e‐0573.771.0296e‐08
February 2015147.90<0.0000e‐16213.80<0.0000e‐16
March 2015163.82<0.0000e‐16346.98<0.0000e‐16
April 2015308.18<0.0000e‐16642.17<0.0000e‐16
May 2015845.04<0.0000e‐16959.92<0.0000e‐16
June 2015709.63<0.0000e‐161,300.12<0.0000e‐16
July 2015850.45<0.0000e‐161,661.28<0.0000e‐16
August 2015748.95<0.0000e‐161,955.30<0.0000e‐16
September 2015698.63<0.0000e‐162,257.07<0.0000e‐16
October 2015714.21<0.0000e‐162,595.07<0.0000e‐16
November 2015669.75<0.0000e‐163,264.34<0.0000e‐16

Open in a separate window

The difference between the forecast and observed distributions of balances for October 2015 is illustrated in figure9(a). Each bar quantifies the percentage change in the proportion for a bin. A substantial reduction exists in the proportion of consumers with a balance of zero. Conversely, most of the categories of balances greater than $2,000 have increased membership relative to the forecast proportions. This demonstrates a pronounced change in credit usage. It suggests that many consumers may have resorted to borrowing from credit cards to cover expenditures, while they would otherwise not do so.

Open in a separate window


Deviations from forecasted balances in Alberta, October 2015

In contrast, the distribution of HELOC balances moved in the opposite direction. As illustrated in figure9(b), balances of HELOCs declined during the oil‐price shock, in roughly the same proportions as observed during the COVID‐19 pandemic. Our results provide further evidence that the majority of HELOC holders did not tap into their home equity to access liquidity. These findings are consistent with Financial Consumer Agency of Canada(2017) in that, from 1999 to 2010, 40% of HELOC borrowers used their accounts to pay for consumption and home renovations, 36% used their accounts for investment, and only 26% used them for debt consolidation. The oil‐related economic downturn resulted in stagnant housing price growth and reduced consumer spending through HELOC usage.

Our analysis provides an appropriate yard stick with which to measure the effect of the COVID‐19 pandemic on the same variables. During COVID‐19, credit card balances decreased during the COVID‐19 pandemic, to a degree roughly twice the proportion of the increase in indebtedness that resulted from the oil price shock. HELOC balances have declined during both COVID‐19 and the Alberta price shock, suggesting that obtaining financial liquidity is not a major purpose of HELOC usage.

4. Summary and conclusions

We have presented an empirical framework to describe the law of motion of consumers’ credit usage and used it to provide a plausible forecast of the distribution of credit card and HELOC balances under the counterfactual state in which the COVID‐19 pandemic had not taken place. We found a significant downward shift in consumer credit usage in Canada, slashing billions of dollars off credit card and HELOC balances across the country, to a level not seen in years. This followed a long period of increasing debt levels in Canada, which has been a concern over the prospects for the Canadian economy.

The changes in financial position materialized through decreased proportions of consumers in high‐balance categories. Although the percentage changes in balance categories were lower for HELOCs than for credit cards, the dollar values of the changes are measured in thousands and tens of thousands. That we found a leftward shift in the distribution of HELOC loan balances demonstrates that the reduction in borrowing is not restricted to high‐interest, credit card debt: the changes that we documented are more widespread than typically perceived. Finally, the two types of accounts are typically used differently by consumers. Credit card usage represents both monthly expenditures and borrowing because many consumers will pay off their balances soon after the statement. In contrast, HELOC loans exclusively represent debt, so the reduction of the balances of HELOC loans represents an unambiguous reduction in debt for consumers in Canada.

When compared with changes during a recent recession in Canada, the magnitude of the changes in credit card balances is nearly twice as large, except that the distributions of balances shift in the opposite direction. Instead of increasing their use of high‐interest debt, consumers are paying down their balances—something not usually seen during a recession.

Supporting information

Supplementary material accompanies the online version of this article.

Click here for additional data file.(322K, zip)

Appendix A. Data appendix

Given the ongoing concern surrounding delayed credit reporting during the COVID‐19 pandemic (Sandler and Ricks2020), we provide detailed comparisons of our dataset obtained from the credit reporting agency with other available data sources. Overall, we conclude that our dataset is representative of most of the Canadian population, both in the cross‐section and over time.

We have validated the database from the consumer reporting agency by comparing against the regulatory filings reported by chartered banks and maintained by the Bank of Canada. We included this comparison because regulatory filings are required to follow specific accounting regulations to report the status of a financial institute, while credit reports are constructed to reflect individuals’ creditworthiness through their historical loan performance. Although the information is largely derived from the same sources, different reporting conventions and data processing may result in discrepancies, which our results suggest to be small.

FigureA1 depicts the monthly time series of aggregate balances from January2017 to September2020. In the left panel, the series collected by the Bank of Canada is shown with the solid line. The dashed line represents the analogue from the database held by the credit reporting agency. It is the sum of all credit card balances held by chartered banks for all consumers in the database. Although slightly higher than that reported by the chartered banks, the monthly changes are very similar, indicating that the database held by the credit reporting agency is representative of the population of the customers of chartered banks. The dotted line shows the aggregate credit card balances from all financial institutions in Canada, recorded in the database from the credit agency. Loan balances at institutions other than chartered banks, such as credit unions and credit card companies, account for about 12% of the total outstanding credit card balances.


Consumer credit usage in Canada during the coronavirus pandemic (12)

Open in a separate window

Time series of aggregate credit card balances

For another method of data validation, we examined the Nilson Report(2020). This report contains information concerning a wider range of credit card issuers in Canada. TableA1 shows a comparison of credit card statistics between the Nilson Report and the database held by the consumer credit agency. This table shows the number of accounts and the number of those accounts that are active, in millions. The numbers of accounts are similar, with less than a 4% difference between the numbers from the Nilson Report and the credit reporting agency. This similarity holds in spite of the fact that the data‐cleaning steps may well have been different in each set of statistics. The total balances for these accounts differ by no more than 4%.


Comparison of accounts at the credit agency with nation‐wide totals in the Nilson Report

Nilson ReportCredit agency

Open in a separate window

In terms of consumer coverage, we also compared the number of credit‐active consumers in the credit reporting agency with the population reported in the 2016 Canadian Census. For this comparison, we only considered the population for Canadians aged 20 and above. FigureA2 illustrates the number of individuals with any form of credit divided by the population in each province of Canada. The first bar shows the ratio of coverage for the entire country of Canada (CA) and it shows that 93.9% of consumers are represented in the database. The rest of the bars show the same ratio for each province, ordered by population. For the provinces of Ontario (ON), British Columbia (BC) and Alberta (AB), the coverage rate is even higher than the national average. For Quebec (QC), the Atlantic provinces (ATL), Manitoba (MB) and Saskatchewan (SK), the coverage rate is slightly lower but still greater than 88%. Coverage is lowest, near 80%, for the Canadian territories (CAT), in the northern part of Canada, where the population is the lowest. Overall, we conclude that our data set correctly represents the population of the entire country.


Consumer credit usage in Canada during the coronavirus pandemic (13)

Open in a separate window

Credit data coverage for adults in Canada, by province


The views expressed in this paper are those of the authors. No responsibility for these views should be attributed to the Bank of Canada. Any errors are the responsibility of the authors. We thank Jason Allen, Brian Peterson, James Chapman, the Co‐Editor, two anonymous referees and participants of a brown bag lunch seminar at the Bank of Canada for many useful comments and suggestions. We also thank the HPC team at the Bank of Canada for their excellent assistance with EDITH 2.0 High Performance Cluster. Anson T. Y. Ho is an external academic consultant to the Financial System Research Centre at the Bank of Canada. The authors have no conflicts of interest to disclose. This research did not receive any specific grant from funding agencies in the public, commercial or not‐for‐profit sectors.


1The full scientific name of the virus is severe acute respiratory syndrome coronavirus 2, abbreviated as SARS‐CoV‐2.

2Putting these numbers into context, the household‐debt‐to‐disposal‐income ratio in the United States peaked at 143.6% in 2007 prior to the global financial crisis.


4CERB expired on September 27, 2020, and transitioned to a revised employment insurance program as well as the new Canada Recovery Benefit (CRB) program.


6The Canadian government also introduced the Canada Emergency Business Account (CEBA) to provide interest‐free loans to small businesses and the Canada Emergency Wage Subsidy (CEWS) program, which allows eligible employers to receive a 75% subsidy on their employees’ wages, up to $58,700.

7At origination, the amount borrowed through a HELOC cannot exceed 65% of the market value of the underlying residential property. Furthermore, the HELOC limit combined with the outstanding mortgage balance cannot exceed 80% of the market value.

8Ho etal.(2020) provided a detailed description of the empirical methodology; we have applied a special case of this framework.

9In the United States, the Coronavirus Aid, Relief, and Economic Security (CARES) Act provided a one‐time cash payment of $1,200 to most adults and $500 per child and unemployment benefits of $600 per week over an additional 13 weeks of benefits.

10Because we analyzed only individuals’ use of credit cards and HELOCs, other loan types are not considered in our dataset.

11We focused on the outstanding balance instead of delinquent loans because the unprecedented availability of loan deferral programs allowed many consumers, who might otherwise become delinquent, to remain current on their payments. An analysis focusing on delinquency or bad debt may provide a misleading picture of consumers’ financial status. Indeed, 90‐day credit card delinquency decreased from 1.11% in March 2020 to 0.63% in September 2020.

12Less than 0.1% of accounts in our sample are dropped with this restriction. The most common reason for accounts not being updated is that these accounts were inactive with zero balance. Those account may also be sold, transferred or paid off without being reported in a timely manner. Some lenders may cease reporting severely delinquent accounts that were eventually written off. Our way of handling outdated accounts is consistent with the method applied to the Consumer Credit Panelat the Federal Reserve Bank of New York (Lee and van der Klaauw2010).

13Although none of consumers in our nation‐wide sample have time gaps, there exists roughly 1% of consumers with gaps in the sample drawn from the province of Alberta. This fraction of consumers is consistent with the notion that a small subset of people have migrated both in and out of Alberta during the sample period. Our econometric technique allows for this possibility.

14The Kullback–Leibler divergence statistic also permits other restrictions one might seek to impose on the data in order to reduce sampling variability. Because we have a large amount of data concerning a major portion of the Canadian population, sampling variability is not a major concern.

15Although this may be early, it avoids initializing our forecasts while the pandemic was active in Canada. Also, it allows for a few months of out‐of‐sample tests of the modelling framework in the months before March 2020.

16HELOC usage had already exhibited a decreasing trend since 2013 because of a policy change at the end of 2012, reducing the maximum loan‐to‐value ratio of HELOCs from 80% to 65%; see the Residential Mortgage Underwriting Practices and Procedures Guideline (B‐20), in OSFI(2012).

17Financial institutions are required to report their aggregate financial status to the Bank of Canada and OSFI. A detailed list of regulations is publicly available from www.osfi‐ Notable items include the balance sheet, income statement, mortgage and non‐mortgage loan balances and reports of new lending.

18This ratio includes not only those consumers with credit cards and HELOCs but also those with mortgages, instalment loans, other lines of credit and even telecommunication accounts.


  • Agarwal, S., Ambrose B. W., and Liu C. (2006) “Credit lines and credit utilization,” Journal of Money, Credit and Banking38, 1–22 [Google Scholar]
  • Agarwal, S., and Qian W. (2017) “Access to home equity and consumption: Evidence from a policy experiment,” Review of Economics and Statistics99, 40–52 [Google Scholar]
  • Al‐Mqbali, L., Bilyk O., Caputo S., and Younker J. (2019) “Reassessing the growth of HELOCs in Canada using new regulatory data,” Bank of Canada Staff Analytical Note2019(14), 1–5 [Google Scholar]
  • Allen, J., Clark R., Li S., and Vincent N. (2022) “Debt‐relief programs and money left on the table: Evidence from Canada's response to COVID‐19,” Canadian Journal of Economics, forthcoming [Google Scholar]
  • Andersen, A.L., Hansen E. T., Johannesen N., and Sheridan A. (2020) “Consumer responses to the COVID‐19 crisis: Evidence from bank account transaction data,” Covid Economics7, 88–114 [Google Scholar]
  • Baker, S.R., Farrokhnia R. A., Meyer S., Pagel M., and Yannelis C. (2020) “Income, liquidity, and the consumption response to the 2020 economic stimulus payments,” NBER working paper no. 27097
  • Bank of Canada(2020) Bank of Canada Monetary Policy Report – October 2020, Bank of Canada; [Google Scholar]
  • Belov, D.I., and Armstrong R. D. (2011) “Distributions of the Kullback–Leibler divergence with applications,” British Journal of Mathematical and Statistical Psychology64, 291–309 [PubMed] [Google Scholar]
  • Box, G.E.P., and Tiao G. C. (1975) “Intervention analysis with applications to economic and environmental problems,” Journal of the American Statistical Association70, 70–79 [Google Scholar]
  • Carvalho, V.M., García J. R., Hansen S., Ortiz A., Rodrigo T., Mora S. R., and Ruiz P. (2020) “Tracking the COVID‐19 crisis with high‐resolution transaction data,” CEPR discussion paper no. DP14642 [PMC free article] [PubMed]
  • Cateau, G., Roberts T., and Zhou J. (2015) “Indebted households and potential vulnerabilities for the Canadian financial system: A microdata analysis,” Bank of Canada Financial System Review2015, 49–58 [Google Scholar]
  • Chen, H., Engert W., Huynh K. P., Nicholls G., Nicholson M., and Zhu J. (2020a) “Cash and COVID‐19: The impact of the pandemic on demand for and use of cash,” Bank of Canada Staff Discussion Papers, no. 2020‐6
  • Chen, H., Michaux M., and Roussanov N. (2020b) “Houses as ATMs: Mortgage refinancing and macroeconomic uncertainty,” Journal of Finance75, 323–75 [Google Scholar]
  • Chetty, R., Friedman J. N., Hendren N., Stepner M., and Team T. O. I. (2020) “The economic impacts of COVID‐19: Evidence from a new public database built using private sector data,” NBER working paper no. 27431
  • Deng, Z., Morissette R., and Messacar D. (2020) “Running the economy remotely: Potential for working from home during and after COVID‐19,” StatCan COVID‐19: Data to Insights for a Better Canada, catalogue no. 45280001
  • Financial Consumer Agency of Canada(2017) “Home equity lines of credit: Market trends and consumer issues,” technical report, Government of Canada; [Google Scholar]
  • Financial Consumer Agency of Canada(2018) “Home equity lines of credit: Consumer knowledge and behaviour,” technical report, Government of Canada; [Google Scholar]
  • Financial Consumer Agency of Canada(2020) “COVID‐19: FCAC tracking of bank relief measures,” technical report, Government of Canada; [Google Scholar]
  • Ganong, P., and Noel P. (2019) “Consumer spending during unemployment: Positive and normative implications,” American Economic Review109(7), 2383–424 [Google Scholar]
  • Henry, C., Huynh K. P., and Welte A. (2018) “2017 Methods‐of‐Payment Survey Report,” Bank of Canada discussion paper no. 18‐17
  • Ho, A.T.Y., Morin L., Paarsch H. J., and Huynh K. P. (2020) “A flexible framework for intervention analysis applied to credit‐card usage during the coronavirus pandemic,” mimeo; [PMC free article] [PubMed] [Google Scholar]
  • Hurst, E., and Stafford F. (2004) “Home is where the equity is: Mortgage refinancing and household consumption,” Journal of Money, Credit and Banking36, 985–1014 [Google Scholar]
  • Kullback, S., and Leibler R. A. (1951) “On information and sufficiency,” Annals of Mathematical Statistics22, 79–86 [Google Scholar]
  • Lee, D., and vander Klaauw W. (2010) “An introduction to the New York Fed Consumer Credit Panel,” Federal Reserve Bank of New York Staff Report, no.479
  • Messacar, D., Morissette R., and Deng Z. (2020) “Inequality in the feasibility of working from home during and after COVID‐19,” StatCan COVID‐19: Data to Insights for a Better Canada, catalogue no. 45280001
  • Mintel(2018) “Canada unsecured lending market report.” London, UK: Mintel Group Ltd. [Google Scholar]
  • Nilson Report(2020) April 2020, Issue 1173
  • Nagypál, É., Gibbs C., and Fulford S. (2020) “The early effects of the COVID‐19 pandemic on credit applications,” CFPB Office of Research Special Issue Brief
  • OECD(2020a) “Household debt (indicator),” OECD National Accounts Statistics Database
  • OECD(2020b) OECD Economic Outlook, Volume 2020, Issue 1. Paris: OECD Publishing; [Google Scholar]
  • OSFI(2012) “Residential mortgage underwriting practices and procedures guideline (B‐20),” Office of the Superintendent of Financial Institutions. Ottawa: Government of Canada; [Google Scholar]
  • Parkash, O., and Mukesh S. (2013) “Relation between information measures and chi‐square statistic,” International Journal of Pure and Applied Mathematics84(5), 517–24 [Google Scholar]
  • Sandler, R., and Ricks J. (2020) “The early effects of the COVID‐19 pandemic on consumer credit,” CFPB Office of Research Special Issue Brief
  • Shannon, C.E. (1948) “A mathematical theory of communications,” Bell Systems Technical Journal27, 379–423 [Google Scholar]
  • Song, K.‐S. (2002) “Goodness‐of‐fit tests based on Kullback–Leibler discrimination information,” IEEE Transactions on Information Theory48(5), 1103–17 [Google Scholar]
  • Statistics Canada(2020) “Provisional death counts and excess mortality,” January to September 2020,” The Daily. November 26, 2021. Available at‐quotidien/201126/dq201126c‐eng.htm
  • Sullivan, J.X. (2008) “Borrowing during unemployment: Unsecured debt as a safety net,” Journal of Human Resources43, 383–412 [Google Scholar]
  • Vallée, G. (2020) “Monitoring payment deferrals during the COVID‐19 pandemic,” technical report, Bank of Canada Financial System Hub; [Google Scholar]
  • World Bank(2020) Global Economic Prospects, June 2020, Washington: World Bank Group; [Google Scholar]

Articles from The Canadian Journal of Economics. Revue Canadienne D'Economique are provided here courtesy of Wiley-Blackwell

Consumer credit usage in Canada during the coronavirus pandemic (2024)


Consumer credit usage in Canada during the coronavirus pandemic? ›

We found a dramatic leftward shift in the distribution of credit card and HELOC outstanding balances, providing evidence for a widespread reduction in credit usage. Our findings suggest that, during the COVID‐19 recession, Canadian consumers were able to meet their financial needs without increasing their debt burdens.

What is the consumer credit debt in Canada? ›

According to the report, the total consumer debt in Canada is $2.45 trillion, up by 3.2 per cent from the previous year.

How much consumer credit card debt is there in Canada today? ›

As credit cards are the most popular form of payment in Canada, it's not surprising that as of May 2023, the total credit card debt in Canada was $96.9 billion, and 42% of Canadians have credit card debts that are not paid off.

How did Covid 19 impact credit scores? ›

A recent CFPB analysis found that the distribution of credit scores shifted upward during the pandemic, suggesting that pandemic-era mortgage forbearances, the federal student loan repayment pauses, and federal cash transfers that improved some consumers' financial wellbeing drove the overall increases in credit scores ...

How does the pandemic affect consumer lending? ›

The unprecedented government transfers, coupled with changes in consumer spending and saving behaviors, led to a change in consumer lending trends. Consumer loan volumes declined in 2020 and remained below pre-pandemic levels through third quarter 2021.

What is the debt of Canada right now? ›

Adjusted for current inflation, Canada's total provincial and federal debt increased from $1.1 trillion to $2.1 trillion between 2007-2008 and 2022-2023, according to a study by the right-wing think tank Fraser Institute(opens in a new tab).

What is the current debt in Canada? ›

Federal government net debt, at $910 billion, or 36.3% of GDP, was above the pre-pandemic level, but was down from 42.7% of GDP in the previous year. As of March 2022, Canada's DBRS AAA federal credit rating was maintained.

What percentage of Canadians are in credit card debt? ›

Edited By. Forty-three percent of Canadians currently have credit card debt, according to NerdWallet's 2023 Consumer Credit Card report. Among them, 40% think it will take them six months or longer to pay off all of that debt, and 11% are not at all sure how long it will take them to pay it off.

Why are so many Canadians in debt? ›

The impact of higher interest rates and cost of living created increased vulnerability among Canadians,” said Matt Fabian, director of financial services research and consulting at TransUnion Canada.

What is the average credit card limit in Canada? ›

While exact figures on credit limits in Canada are somewhat elusive, high-limit credit cards typically carry limits that exceed $10,000 or $20,000. For comparison, a recent Equifax Report indicates the average credit limit in Canada is around $5,800.

How does COVID-19 affect debt? ›

The COVID-19 pandemic has greatly lengthened the list of developing and emerging market economies in debt distress. For some, a crisis is imminent. For many more, only exceptionally low global interest rates may be delaying a reckoning. Default rates are rising, and the need for debt restructuring is growing.

How much financial damage did COVID cause? ›

The economic toll of the COVID-19 pandemic in the U.S. will reach US$14 trillion by the end of 2023, our team of economists, public policy researchers and other experts have estimated.

Did credit score go up during pandemic? ›

The increase during the COVID-19 pandemic years was slightly less than 5 percentage points between late 2019 and early 2022. The largest year-over-year change reached 3.17 percentage points between the first quarter of 2020 and the first quarter of 2021.

How did the Covid-19 pandemic affect consumers? ›

The pandemic has also resulted in increased consumption of certain products which were either consumed in lesser quantities or not consumed at all before the event (Kirk & Rifkin, 2020). Such effects have led to significant upward shifts in the market demand for these products. We refer to such shifts as 'new demand'.

How does COVID-19 shock affect consumer credit? ›

First, we document a sharp drop in consumer credit use around mid-March, when the pandemic was declared a national emergency. Credit card spending across all borrowers decreased by 50 percent year over year by the end of April and subsequently started to recover in May, especially for the least creditworthy borrowers.

How did COVID affect the finance industry? ›

COVID-19 has adversely affected the stock market in uncertainty and reduced stock return worldwide, reducing capital flows. This decline due to stock market uncertainty ultimately created obstacles in the availability of liquidity and investment in the global financial system (Padhan and Prabheesh, 2021).

What is consumer credit debt? ›

Consumer debt consists of those loans used for personal consumption as opposed to debts incurred by businesses or through government activities. Consumer debt may be segmented into revolving debt, which is paid monthly and may have a variable rate; and non-revolving debt, paid as a fixed rate.

What is the average consumer credit debt? ›

Average Credit Card Balance by Generation
GenerationAverage Credit Card Debt
Generation Z$3,262
Generation X$9,123
Baby boomers$6,642
1 more row
Mar 12, 2024

What is the average consumer debt? ›

The average debt an American owes is $104,215 across mortgage loans, home equity lines of credit, auto loans, credit card debt, student loan debt, and other debts like personal loans. Data from Experian breaks down the average debt a consumer holds based on type, age, credit score, and state.

Top Articles
Latest Posts
Article information

Author: Fr. Dewey Fisher

Last Updated:

Views: 5469

Rating: 4.1 / 5 (42 voted)

Reviews: 89% of readers found this page helpful

Author information

Name: Fr. Dewey Fisher

Birthday: 1993-03-26

Address: 917 Hyun Views, Rogahnmouth, KY 91013-8827

Phone: +5938540192553

Job: Administration Developer

Hobby: Embroidery, Horseback riding, Juggling, Urban exploration, Skiing, Cycling, Handball

Introduction: My name is Fr. Dewey Fisher, I am a powerful, open, faithful, combative, spotless, faithful, fair person who loves writing and wants to share my knowledge and understanding with you.