The US is now averaging close to 1 million tests per day. Based on the experience of other countries, the percent positive needs to be well under 5% to really push down new infections (probably close to 1%), so the US still needs to increase the number of tests per day significantly (or take actions to push down the number of new infections).

There were 1,096,494 test results reported over the last 24 hours.

There were 88,452 positive tests. This is a new record.

Almost 21,500 Americans deaths from COVID have been reported in October. See the graph on US Daily Deaths here.

Click on graph for larger image.

This data is from the COVID Tracking Project.

The percent positive over the last 24 hours was 8.1% (red line is 7 day average).

For the status of contact tracing by state, check out testandtrace.com.

And check out COVID Exit Strategy to see how each state is doing.

The second graph shows the 7 day average of positive tests reported.

The dashed line is the July high.

Note that there were very few tests available in March and April, and many cases were missed (the percent positive was very high - see first graph). By June, the percent positive had dropped below 5%.

This is a new record 7-day average for the USA.

Investment has been weak for some time, and slumped in Q1, and fell off a cliff in Q2 along with the overall economy.  Investment bounced back in Q3, especially for residential investment and investment in equipment - but not for non-residential structures.

The first graph below shows the contribution to GDP from residential investment, equipment and software, and nonresidential structures (3 quarter trailing average). This is important to follow because residential investment tends to lead the economy, equipment and software is generally coincident, and nonresidential structure investment trails the economy.

In the graph, red is residential, green is equipment and software, and blue is investment in non-residential structures. So the usual pattern - both into and out of recessions is - red, green, blue.

Of course - with the sudden economic stop due to COVID-19 - the usual pattern doesn't apply.

The dashed gray line is the contribution from the change in private inventories.

Click on graph for larger image.

Residential investment (RI) increased at a 59.3% annual rate in Q3.  Equipment investment increased at a 70.1% annual rate, and investment in non-residential structures decreased at a 14.6% annual rate.
The contribution to Q3 GDP from investment in private inventories was 6.6 percentage points.

On a 3 quarter trailing average basis, RI (red) is up solidly, equipment (green) is up, and nonresidential structures (blue) is down sharply.
I'll post more on the components of non-residential investment once the supplemental data is released.

The second graph shows residential investment as a percent of GDP.

Residential Investment as a percent of GDP increased in Q3.  RI as a percent of GDP is still close to previous lows, and I expected RI to continue to increase further in this cycle.

I'll break down Residential Investment into components after the GDP details are released.

Note: Residential investment (RI) includes new single family structures, multifamily structures, home improvement, broker's commissions, and a few minor categories.

The third graph shows non-residential investment in structures, equipment and "intellectual property products".  
Investment in non-residential structures declined in Q3, and will probably be weak for some time (hotel occupancy is low, office and mall vacancy rates are rising).

From HotelNewsNow.com: STR: US hotel results for week ending 24 October

U.S. weekly hotel occupancy fell back below the 50% mark, according to the latest data from STR through 24 October.

18-24 October 2020 (percentage change from comparable week in 2019):

• Occupancy: 48.0% (-31.7%)
• Average daily rate (ADR): US$95.49 (-29.4%)
• Revenue per available room (RevPAR): US$45.83 (-51.8%)
emphasis added

Since there is a seasonal pattern to the occupancy rate, we can track the year-over-year change in occupancy to look for any improvement. This table shows the year-over-year change since the week ending Sept 19, 2020:

Week EndingYoY Change, Occupancy Rate 9/19-31.9% 9/26-31.5% 10/3-29.6% 10/10-29.2% 10/17-30.7% 10/24-31.7%
This suggests no improvement over the last 6 weeks.

The following graph shows the seasonal pattern for the hotel occupancy rate using the four week average.

Click on graph for larger image.

The red line is for 2020, dash light blue is 2019, blue is the median, and black is for 2009 (the worst year since the Great Depression for hotels - before 2020).

Seasonally we'd expect the occupancy rate to start declining. Note that there was little pickup in business travel that usually happens in the Fall.

Note: Y-axis doesn't start at zero to better show the seasonal change.

From the NAR: Pending Home Sales Falter 2.2% in September

The Pending Home Sales Index (PHSI), a forward-looking indicator of home sales based on contract signings, fell 2.2% to 130.0 in September. Year-over-year, contract signings rose 20.5%. An index of 100 is equal to the level of contract activity in 2001.
...
The Northeast PHSI grew 2.0% to 119.4 in September, a 27.7% increase from a year ago. In the Midwest, the index slid 3.2% to 120.5 last month, up 18.5% from September 2019.

Pending home sales in the South decreased 3.0% to an index of 150.1 in September, up 19.6% from September 2019. The index in the West fell 2.6% in September to 116.8, up 19.3% from a year ago.
emphasis added

This was below expectations for this index. Note: Contract signings usually lead sales by about 45 to 60 days, so this would usually be for closed sales in October and November.

The DOL reported:

In the week ending October 24, the advance figure for seasonally adjusted initial claims was 751,000, a decrease of 40,000 from the previous week's revised level. The previous week's level was revised up by 4,000 from 787,000 to 791,000. The 4-week moving average was 787,750, a decrease of 24,500 from the previous week's revised average. The previous week's average was revised up by 1,000 from 811,250 to 812,250.
emphasis added

This does not include the 359,667 initial claims for Pandemic Unemployment Assistance (PUA) that was up from 344,905 the previous week. (There are some questions on PUA numbers).

The following graph shows the 4-week moving average of weekly claims since 1971.

Click on graph for larger image.

The dashed line on the graph is the current 4-week average. The four-week average of weekly unemployment claims decreased to 811,250.

The previous week was revised up.

The second graph shows seasonally adjust continued claims since 1967 (lags initial by one week).

At the worst of the Great Recession, continued claims peaked at 6.635 million, but then steadily declined.

Continued claims decreased to 7,756,000 (SA) from 8,465,000 (SA) last week and will likely stay at a high level until the crisis abates.

Note: There are an additional 10,324,779 receiving Pandemic Unemployment Assistance (PUA) that increased from 10,152,753 the previous week (there are questions about these numbers). This is a special program for business owners, self-employed, independent contractors or gig workers not receiving other unemployment insurance. 
An additional 3,683,496 are receiving Pandemic Emergency Unemployment Compensation (PEUC).

From the BEA: Gross Domestic Product, Third Quarter 2020 (Advance Estimate)

Real gross domestic product (GDP) increased at an annual rate of 33.1 percent in the third quarter of 2020, according to the "advance" estimate released by the Bureau of Economic Analysis. In the second quarter, real GDP decreased 31.4 percent.

The GDP estimate released today is based on source data that are incomplete or subject to further revision by the source agency. The "second" estimate for the third quarter, based on more complete data, will be released on November 25, 2020.
...
The increase in real GDP reflected increases in personal consumption expenditures (PCE), private inventory investment, exports, nonresidential fixed investment, and residential fixed investment that were partly offset by decreases in federal government spending (reflecting fewer fees paid to administer the Paycheck Protection Program loans) and state and local government spending. Imports, which are a subtraction in the calculation of GDP, increased.

The increase in PCE reflected increases in services (led by health care as well as food services and accommodations) and goods (led by motor vehicles and parts as well as clothing and footwear). The increase in private inventory investment primarily reflected an increase in retail trade (led by motor vehicle dealers). The increase in exports primarily reflected an increase in goods (led by automotive vehicles, engines, and parts as well as capital goods). The increase in nonresidential fixed investment primarily reflected an increase in equipment (led by transportation equipment). The increase in residential fixed investment primarily reflected an increase in brokers' commissions and other ownership transfer costs.
emphasis added

Click on graph for larger image.

This graph shows the percent decline in real GDP from the previous peak (currently the previous peak was in Q4 2019).

This graph is through Q3 2020, and real GDP is currently off 3.5% from the previous peak.  For comparison, at the depth of the Great Recession, real GDP was down 4.0% from the previous peak.

The advance Q3 GDP report, at 33.1% annualized, was close to expectations.

Personal consumption expenditures (PCE) increased at 40.7% annualized rate in Q3, up from 33.2% decrease in Q2.  Residential investment (RI) increased at a 59.3% rate in Q3. Equipment investment increased at a 70.1% annualized rate, and investment in non-residential structures decreased at a 14.6% pace.

I'll have more later ...

Thursday:
• At 8:30 AM ET, Gross Domestic Product, 3rd quarter 2020 (advance estimate). The consensus is that real GDP increased 31.9% annualized in Q3, up from negative 31.4% in Q2.

• At 8:30 AM, The initial weekly unemployment claims report will be released. The consensus is initial claims increased to 800 thousand from 787 thousand last week.

• At 10:00 AM, Pending Home Sales Index for September. The consensus is 4.5% increase in the index.

The US is now averaging close to 1 million tests per day. Based on the experience of other countries, the percent positive needs to be well under 5% to really push down new infections (probably close to 1%), so the US still needs to increase the number of tests per day significantly (or take actions to push down the number of new infections).

There were 875,738 test results reported over the last 24 hours.

There were 78,661 positive tests.

Almost 20,500 Americans deaths from COVID have been reported in October. See the graph on US Daily Deaths here.

Click on graph for larger image.

This data is from the COVID Tracking Project.

The percent positive over the last 24 hours was 9.0% (red line is 7 day average).

For the status of contact tracing by state, check out testandtrace.com.

And check out COVID Exit Strategy to see how each state is doing.

The second graph shows the 7 day average of positive tests reported.

The dashed line is the July high.

Note that there were very few tests available in March and April, and many cases were missed (the percent positive was very high - see first graph). By June, the percent positive had dropped below 5%.

This is a new record 7-day average for the USA.

Freddie Mac reported that the Single-Family serious delinquency rate in September was 3.04%, down from 3.17% in August. Freddie's rate is up from 0.61% in August 2019.

Freddie's serious delinquency rate peaked in February 2010 at 4.20%.

These are mortgage loans that are "three monthly payments or more past due or in foreclosure".

Click on graph for larger image

Mortgages in forbearance are being counted as delinquent in this monthly report, but they will not be reported to the credit bureaus.

This is very different from the increase in delinquencies following the housing bubble.   Lending standards have been fairly solid over the last decade, and most of these homeowners have equity in their homes - and they will be able to restructure their loans once (if) they are employed.

Note: Fannie Mae will report for September soon.

From Wards: U.S. Light Vehicle Sales & Inventory Forecast, October 2020 (pay content)

Click on graph for larger image.

This graph shows actual sales from the BEA (Blue), and Wards forecast for October (Red).

Sales have bounced back from the April low, and will likely be unchanged year-over-year in October.

The Wards forecast of 16.8 million SAAR, would be up about 3% from September.

This would put sales in 2020, through October, down about 17% compared to the same period in 2019.

The Case-Shiller house price indexes for August were released yesterday. Zillow forecasts Case-Shiller a month early, and I like to check the Zillow forecasts since they have been pretty close.

From Matthew Speakman at Zillow: August Case-Shiller Results and September Forecast: No Signs of Cooling

The remarkable surge in home prices continued into August as prices showed no signs of cooling down heading into the fall.
...
By some measures, home prices are rising at a faster pace than they ever have – an incredible feat considering the market is rising from an already elevated level. The supply of for-sale homes, already extremely tight, has only become more constrained in recent months, and historically low mortgage rates continue to encourage many buyers to enter the market. This heightened competition for the few homes on the market has placed consistent, firm pressure on home prices for months now, and there are few signs that this will relent any time soon. While the path of the overall economy is likely to be most directly dictated by coronavirus-related and political developments in the coming months, recent trends suggest that the housing market – which has basically withstood every pandemic-related challenge to this point – will continue its strong momentum in the months to come.

Annual growth in September as reported by Case-Shiller is expected to accelerate in all three main indices. S&P Dow Jones Indices is expected to release data for the September S&P CoreLogic Case-Shiller Indices on Tuesday, November 24.
emphasis added

The Zillow forecast is for the year-over-year change for the Case-Shiller National index to be at 6.6% in September, up from 5.7% in August.

The Zillow forecast is for the 20-City index to be up 6.2% YoY in September from 5.2% in August, and for the 10-City index to increase to be up 5.7% YoY compared to 4.7% YoY in August.

From the MBA: Mortgage Applications Increase in Latest MBA Weekly Survey

Mortgage applications increased 1.7 percent from one week earlier, according to data from the Mortgage Bankers Association’s (MBA) Weekly Mortgage Applications Survey for the week ending October 23, 2020.

... The Refinance Index increased 3 percent from the previous week and was 80 percent higher than the same week one year ago. The seasonally adjusted Purchase Index increased 0.2 percent from one week earlier. The unadjusted Purchase Index decreased 0.3 percent compared with the previous week and was 24 percent higher than the same week one year ago.

“Mortgage applications to buy a home were flat compared to the prior week, but overall activity remains strong this fall. Applications jumped 24 percent compared to last year, and the average loan size reached another record high at $372,600. These results highlight just how strong the upper end of the market is right now, with outsized growth rates in the higher loan size categories. Furthermore, housing inventory shortages have pushed national home prices considerably higher on an annual basis,” said Joel Kan, MBA’s Associative Vice President of Economic and Industry Forecasting. “Refinance activity has been somewhat volatile over the past few months but did increase almost 3 percent last week. With the 30-year fixed rate at MBA’s all-time survey low of 3.00 percent, conventional refinances rose 5 percent. However, the government refinance index decreased for the first time in a month, driven by a slowdown in VA refinance activity.”
...
The average contract interest rate for 30-year fixed-rate mortgages with conforming loan balances ($510,400 or less) decreased to 3.00 percent from 3.02 percent, with points decreasing to 0.35 from 0.36 (including the origination fee) for 80 percent loan-to-value ratio (LTV) loans.
emphasis added

Click on graph for larger image.


The first graph shows the refinance index since 1990.

The refinance index has been very volatile recently depending on rates and liquidity.

But with record low rates, the index remains up significantly from last year.

The second graph shows the MBA mortgage purchase index

According to the MBA, purchase activity is up 24% year-over-year unadjusted.

Note: Red is a four-week average (blue is weekly).

The US is now averaging close to 1 million tests per day. Based on the experience of other countries, the percent positive needs to be well under 5% to really push down new infections (probably close to 1%), so the US still needs to increase the number of tests per day significantly (or take actions to push down the number of new infections).

There were 892,015 test results reported over the last 24 hours.

There were 73,096 positive tests.

Almost 19,500 Americans deaths from COVID have been reported in October. See the graph on US Daily Deaths here.

Click on graph for larger image.

This data is from the COVID Tracking Project.

The percent positive over the last 24 hours was 8.2% (red line is 7 day average).

For the status of contact tracing by state, check out testandtrace.com.

And check out COVID Exit Strategy to see how each state is doing.

The second graph shows the 7 day average of positive tests reported.

The dashed line is the July high.

Note that there were very few tests available in March and April, and many cases were missed (the percent positive was very high - see first graph). By June, the percent positive had dropped below 5%.

This is a new record 7-day average for the USA.

Earlier from the Richmond Fed: Fifth District manufacturing activity strengthened in October

Fifth District manufacturing activity strengthened in October, according to the most recent survey from the Richmond Fed. The composite index rose from 21 in September to 29 in October, its highest reading on record, buoyed by increases in the shipments and new orders indexes, while the third component—the employment index—was unchanged. Firms reported improving business conditions and growing backlogs of orders, overall. Manufacturers were optimistic that conditions would continue to improve in the coming months.

Survey results indicated that many manufacturers continued to increase employment and wages in October.
emphasis added

This was the last of the regional Fed surveys for October.

Here is a graph comparing the regional Fed surveys and the ISM manufacturing index:

Click on graph for larger image.

The New York and Philly Fed surveys are averaged together (yellow, through October), and five Fed surveys are averaged (blue, through October) including New York, Philly, Richmond, Dallas and Kansas City. The Institute for Supply Management (ISM) PMI (red) is through September (right axis).

The ISM manufacturing index for October will be released on Monday, November 2nd. Based on these regional surveys, the ISM manufacturing index will likely increase in October from the September level.

Note that these are diffusion indexes, so readings above 0 (or 50 for the ISM) means activity is increasing (it does not mean that activity is back to pre-crisis levels).

The Census Bureau released an update today on 2020 Census Paid Temporary Workers

As of the September reference week, there were 246,801 decennial Census temporary workers. As of the October reference week, October 11th - 17th, there were 99,490 temp workers.

That is a decrease of 147,311 temporary jobs.

In August, the employment report showed a gain of 238,000 temporary 2020 Census workers, boosting the headline number.

In September, the employment report showed a decrease of 41,000 temporary 2020 Census workers, reducing the headline number.

The October employment report will show a 147,311 decrease in temporary Census employment.

The Census Bureau released the Residential Vacancies and Homeownership report for Q3 2020.

It is likely the results of this survey were significantly distorted by the pandemic.  See note from Census below.

This report is frequently mentioned by analysts and the media to track household formation, the homeownership rate, and the homeowner and rental vacancy rates.  However, there are serious questions about the accuracy of this survey.

This survey might show the trend, but I wouldn't rely on the absolute numbers. he Census Bureau is investigating the differences between the HVS, ACS and decennial Census, and analysts probably shouldn't use the HVS to estimate the excess vacant supply or household formation, or rely on the homeownership rate, except as a guide to the trend.

National vacancy rates in the third quarter 2020 were 6.4 percent for rental housing and 0.9 percent for homeowner housing. The rental vacancy rate of 6.4 percent was 0.4 percentage points lower than the rate in the third quarter 2019 (6.8 percent) and 0.7 percentage points higher than the rate in the second quarter 2020 (5.7 percent). The homeowner vacancy rate of 0.9 percent was 0.5 percentage points lower than the rate in the third quarter 2019 (1.4 percent) and virtually unchanged from the rate in the second quarter 2020 (0.9 percent).

The homeownership rate of 67.4 percent was 2.6 percentage points higher than the rate in the third quarter 2019 (64.8 percent) and not statistically different from the rate in the second quarter 2020 (67.9 percent). "

Click on graph for larger image.

The Red dots are the decennial Census homeownership rates for April 1st 1990, 2000 and 2010. The HVS homeownership rate decreased to 67.4% in Q3, from 67.9% in Q2.

I'd put more weight on the decennial Census numbers.   It is likely the results in Q2 and Q3 were distorted by the pandemic.

The HVS homeowner vacancy was unchanged at 0.9%.

Once again - this probably shows the general trend, but I wouldn't rely on the absolute numbers.

From Census:

Due to the coronavirus pandemic (COVID-19), data collection operations for the CPS/HVS were affected during the third quarter of 2020, as in-person interviews were only allowed for portions of the sample in July (41 percent), August (53 percent), and September (100 percent). The remaining interviews were conducted over the telephone. If the Field Representative was unable to get contact information on the sample unit, the unit was made a Type A noninterview (no one home, refusal, etc). We are unable to determine the extent to which this data collection change affected our estimates.

The rental vacancy rate increased to 6.4% in Q3.

The quarterly HVS is the most timely survey on households, but there are many questions about the accuracy of this survey.

S&P/Case-Shiller released the monthly Home Price Indices for August ("August" is a 3 month average of June, July and August prices).

This release includes prices for 20 individual cities, two composite indices (for 10 cities and 20 cities) and the monthly National index.

From S&P: S&P CoreLogic Case-Shiller Index Shows Annual Home Price Gains Increased to 5.7% in August

The S&P CoreLogic Case-Shiller U.S. National Home Price NSA Index, covering all nine U.S. census divisions, reported a 5.7% annual gain in August, up from 4.8% in the previous month. The 10-City Composite annual increase came in at 4.7%, up from 3.5% in the previous month. The 20-City Composite posted a 5.2% year-over-year gain, up from 4.1% in the previous month.

Phoenix, Seattle and San Diego reported the highest year-over-year gains among the 19 cities (excluding Detroit) in August. Phoenix led the way with a 9.9% year-over-year price increase, followed by Seattle with an 8.5% increase and San Diego with a 7.6% increase. All 19 cities reported higher price increases in the year ending August 2020 versus the year ending July 2020.
...
The National Index posted a 1.1% month-over-month increase, while the 10-City and 20-City Composites both posted increases of 1.1% before seasonal adjustment in August. After seasonal adjustment, the National Index posted a month-over-month increase of 1.0%, while the 10-City and 20- City Composites both posted increases of 0.5%. In August, all 19 cities (excluding Detroit) reported increases before seasonal adjustment, while 17 of the 19 cities reported increases after seasonal adjustment.

“Housing prices were strong in August,” says Craig J. Lazzara, Managing Director and Global Head of Index Investment Strategy at S&P Dow Jones Indices. “The National Composite Index gained 5.7% relative to its level a year ago, well ahead of July’s 4.8% increase. The 10- and 20-City Composites (up 4.7% and 5.2%, respectively) also rose at an accelerating pace in August. The strength of the housing market was consistent nationally – all 19 cities for which we have August data rose, and all 19 gained more in the 12 months ended in August than they had done in the 12 months ended in July.

“A trend of accelerating increases in the National Composite Index began in August 2019 but was interrupted in May and June, as COVID-related restrictions produced modestly-decelerating price gains. We speculated last month that the accelerating trend might have resumed, and August’s results easily bear that interpretation. The last time that the National Composite matched August’s 5.7% growth rate was 25 months ago, in July 2018. If future reports continue in this vein, we may soon be able to conclude that the COVID-related deceleration is behind us.
emphasis added

Click on graph for larger image.

The first graph shows the nominal seasonally adjusted Composite 10, Composite 20 and National indices (the Composite 20 was started in January 2000).

The Composite 10 index is up 5.2% from the bubble peak, and up 0.5% in August (SA) from July.

The Composite 20 index is 9.5% above the bubble peak, and up 0.5% (SA) in August.

The National index is 20.1% above the bubble peak (SA), and up 1.0% (SA) in August.  The National index is up 62% from the post-bubble low set in December 2011 (SA).

The second graph shows the Year over year change in all three indices.

The Composite 10 SA is up 4.7% compared to August 2019.  The Composite 20 SA is up 5.2% year-over-year.

The National index SA is up 5.7% year-over-year.

Note: According to the data, prices increased in 18 cities month-over-month seasonally adjusted.

Price increases were above expectations.  I'll have more later.

Tuesday:
• At 8:30 AM ET, Durable Goods Orders for September from the Census Bureau. The consensus is for a 0.5% increase in durable goods orders.

• At 9:00 AM, S&P/Case-Shiller House Price Index for August.  The consensus is for the Composite 20 index to be up 4.2% year-over-year.

• Also at 9:00 AM, FHFA House Price Index for August. This was originally a GSE only repeat sales, however there is also an expanded index.

• At 10:00 AM, Richmond Fed Survey of Manufacturing Activity for October. This is the last of the regional surveys for October.

The US is now mostly reporting 700 thousand to 1 million tests per day. Based on the experience of other countries, the percent positive needs to be well under 5% to really push down new infections (probably close to 1%), so the US still needs to increase the number of tests per day significantly (or take actions to push down the number of new infections).

There were 1,043,423 test results reported over the last 24 hours.

There were 62,315 positive tests.

Almost 18,500 Americans deaths from COVID have been reported in October. See the graph on US Daily Deaths here.

Click on graph for larger image.

This data is from the COVID Tracking Project.

The percent positive over the last 24 hours was 6.0% (red line is 7 day average).

For the status of contact tracing by state, check out testandtrace.com.

And check out COVID Exit Strategy to see how each state is doing.

The second graph shows the 7 day average of positive tests reported.

The dashed line is the July high.

Note that there were very few tests available in March and April, and many cases were missed (the percent positive was very high - see first graph). By June, the percent positive had dropped below 5%.

This is a new record 7-day average for the USA.

Note: This is as of October 18th.

From the MBA: Share of Mortgage Loans in Forbearance Decreases Slightly to 5.90%

The Mortgage Bankers Association’s (MBA) latest Forbearance and Call Volume Survey revealed that the total number of loans now in forbearance decreased by 2 basis points from 5.92% of servicers’ portfolio volume in the prior week to 5.90% as of October 18, 2020. According to MBA’s estimate, 3.0 million homeowners are in forbearance plans.
...
“The share of loans in forbearance declined only slightly in the prior week, after two weeks of a flurry of borrowers exiting as they reached the six-month mark,” said Mike Fratantoni, MBA’s Senior Vice President and Chief Economist. “There continues to be a steady improvement for Fannie Mae and Freddie Mac loans, but the forbearance share for Ginnie Mae, portfolio, and PLS loans all increased. This is further evidence of the unevenness in the current economic recovery. The housing market is booming, as shown by the extremely strong pace of home sales last week. However, many homeowners continue to struggle, as the pace of the job market’s improvement has waned.”
...
By stage, 25.02% of total loans in forbearance are in the initial forbearance plan stage, while 73.14% are in a forbearance extension. The remaining 1.84% are forbearance re-entries.
emphasis added

Click on graph for larger image.

This graph shows the percent of portfolio in forbearance by investor type over time.  Most of the increase was in late March and early April, and has been trending down for the last few months.

The MBA notes: "Total weekly forbearance requests as a percent of servicing portfolio volume (#) increased relative to the prior week: from 0.10% to 0.11%."

There hasn't been a pickup in forbearance activity related to the end of the extra unemployment benefits.