Homebuilder Destruction Wed, Apr 15, 2020 This morning, the NAHB updated their monthly reading on homebuilder sentiment, and like most other recent economic data, the results were not good. Homebuilder sentiment fell below 50 for the first time since June of 2014 (readings below 50 are considered contractionary). The index plummeted to its lowest reading since June of 2012 falling from 72 in March down to 30 in April. As shown in the table below, that 42 point decline in just one month was the largest ever recorded in the data going back to 1985. The same can also be said for every one of the sub-indices as well which are now at multi-year lows. That stands in stark contrast to where things were just one month ago. In March, the headline index, as well as several other sub-indices, were in the 95th percentile or better of all readings. In other words, before the COVID-19 pandemic homebuilders' outlook was historically optimistic. One interesting thing to note in this month's survey was the declines in homebuilder sentiment were broad across geographic regions but the areas of the country most affected by the virus saw larger declines and are now at lower levels. Namely, sentiment in the Northeast is the lowest of the four regions at 19 compared to the next lowest, 25, in the Midwest. That is as the bulk of US COVID-19 cases can be found in the Northeast, particularly in the tri-state area.
Retail Sales Chaos Wed, Apr 15, 2020 Expectations for this morning’s March Retail Sales report were already low but not low enough. While economists were expecting the headline reading to fall 8.0%, the actual result was even weaker at -8.7%. On the bright side (if you can even say that) readings ex Autos and Ex Autos and Gas were both better than expected. The overall economic impact of the COVID outbreak is obviously negative, but there have been some big winners (and even bigger losers), and this month's Retail Sales report illustrated some of these major shifts. Looking through the various sales trends in our monthly update there were some crazy looking charts, but two that especially stood out were Clothing and Food and Beverage Stores. In the case of Clothing, its 50% decline took the monthly sales rate to the lowest level since 1995. Meanwhile, sales at Food and Beverage Stores took off. With people forgoing clothes in favor of food, no wonder everyone is looking like slobs these days!
Keep Your Eye on Jobless Claims Everyone on Wall Street is debating and looking for the perfect dataset or indicator to tell if the bear market ended on March 23, 2020. It may very well have, but we won’t know for sure until we have a little more hindsight. But we must push on. With the USA on virtual lockdown and many businesses closed for weeks with the prospect of being closed for several more weeks if not months, a recession is quite likely. We know the market leads recession by several months. On average since 1948, bear markets start about 8-9 months ahead of recession though the last bear 2007 began just about 2 months before the recession. Bear markets end about 3-4 months before the end of the recession. The last recession ended in June 2009 about three months after the bear market low in March 2009. So if February 12 was the beginning of the bear market and March 23 was the bear market low then the recession could be rather short-lived. The lingering effects of the pandemic could mute economic activity for some time if there is not clear medical solution for COVID-19. But once the coast is clear there is likely to be a good deal of pent up demand. We applaud all the scientific efforts to develop a vaccine and therapies and all fiscal stimulus to help business and individuals. One metric we have always found useful in identifying bear market lows is the trend of Weekly Initial Jobless Claims. Initial Claims the past three weeks have been massive and historic, threatening to push unemployment near levels of the Great Depression. But if the trend continues, Initial Claims may have peaked already. In the chart we have omitted Initial Claims of 3.3 million, 6.9 million and 6.6 million the past three weeks as they would literally be off the chart. Also we have plotted DJIA on a logarithmic scale so that the highs and lows of the past would be discernible. All the bear market lows for DJIA associated with a recession line up quite well with the spike peaks of Weekly Initial Jobless Claims. Bear markets like 1987, which was a computer driven crash, were not accompanied by recession or an increase in Jobless Claims. But all the big bear market lows of 1970, 1974, 1982, 1990, 2001 and 2009 were marked by the peak of Jobless Claims. Tomorrow’s Weekly Jobless Claims is expected to be lower than last week. So keep your eye on Jobless Claims for continuing confirmation that the low is in.
Jobless Claims Still Dismal Thu, Apr 16, 2020 The pain has not stopped for the US labor market. Initial jobless claims remain at extremely elevated levels coming in at 5.245 million this week. Although this is still extremely high compared to other readings throughout history, this week's print was a small improvement from the past few weeks coming in below expectations (5.625 million) as well as the past couple of weeks' readings which have been well above 6.5 million. Last week, seasonally adjusted claims totaled 6.615 million which makes the 1.37 million drop to 5.245 million this week the largest single week decline on record. Obviously, that is little consolation given the fact that it was still the third-largest weekly reading on record. Put differently, although there were not as many added this week as the past few weeks, there are still a historically massive number of jobless claims being filed. Over the past four weeks, there have now been over 20 million jobless claims filed on a non-seasonally adjusted basis. In terms of total US population, that is roughly 6% or over 12% of the labor force. This week contributed another 4.972 million to that number compared to the past two weeks which saw prints of over 6 million. Again, that 1.24 million claims decline this week is the largest decline ever recorded even though it is not much of a positive. By now, every "normal" claims number has rolled off of the four-week moving average. As such, it has again pressed higher this week to a record of more than 5.5 million.
Philly Fed Collapses Thu, Apr 16, 2020 Yesterday, we highlighted the collapse of the first of the April regional Federal Reserve indices which was from the New York Fed. This morning, we got an update from the Philadelphia Fed, and the results were not much better. The headline index fell from -12.7 in March to -56.6 in April. That is the second-lowest level of the index on record behind the record low of -57.1 back in July of 1980. That deterioration has come very rapidly. As recently as February, before COVID-19's impact had been fully felt, the index was at 36.7 which is at the upper end of its historic range in the 97th percentile of all readings. In just the last two months, though, the headline index has experienced its two largest declines on record: -43.9 this month and -49.4 last month. The components of the index are similarly at or near record lows. As shown in the table below, indices for demand and employment all experienced their largest declines ever this month. Overall, demand is weaker while delivery times are longer as both prices paid and received fall. Demand is one of the key areas of weakness in the regional Fed reports. The indices for New Orders and Shipments have both fallen by their largest amounts ever to record lows. As a result of weak demand, prices are beginning to fall. The Prices Paid component is now down to -9.3 which is its lowest level since May 2015. Prices Received, on the other hand, is even lower at -10.6, which is the worst reading since July of 2009. The 17.4 point decline this month was also the fourth largest month over month decline for Prices Received on record. Last month's report had yet to capture the massive drop in employment as a result of COVID-19. That changed this month with the index for Number of Employees cratering to -46.7. That is the second-lowest reading ever behind March of 2009 when it fell to -51.5. Meanwhile, Average Workweek is down to its lowest level ever at -54.5.
Another Week, Another Massive Jobless Claims Number Thu, Apr 23, 2020 A massive number of people continue to file jobless claims with another 4.427 million initial jobless claims coming in this week. Similar to last week, the good news is that was less than forecasts of 4.5 million and the 4.427 million number is a drop from the 5.237 million last week. In fact, that 810K decline is the second largest on record behind last week's 1.378 million drop. Additionally, this week was the lowest print since the first claims number above 1 million for the week of March 20th. The bad news is that regardless of the somewhat slowed pace on initial jobless claims, the numbers remain at historically unprecedented levels. Before seasonal adjustment, this week saw 4.267 million claims filed compared to 4.965 million last week. Again, while that number is still extremely high, the past two weeks have marked a significant deceleration. While the level of claims are off of their records from a few weeks ago, the moving average put in a new record high today. The moving average has continued to grind higher reaching 5.786 million. Granted, that 280K increase from last week was the smallest since the surge in claims began in late March. As previously mentioned, the pace of initial jobless claims have slowed for the time being, but that does not take away from the massive number of people reporting as jobless. In total, since the beginning of March, just under 27 million Americans, or roughly 8% of the population, have filed initial jobless claims (seasonally adjusted). Although they are released with a one week lag to initial jobless claims, continuing claims this week printed at their highest level on record at 15.976 million.
Election-year Mays: DJIA’s Second Worst Month May officially marks the beginning of the “Worst Six Months” for the DJIA and S&P. To wit: “Sell in May and go away.” Our “Best Six Months Switching Strategy,” created in 1986, proves that there is merit to this old trader’s tale. A hypothetical $10,000 investment in the DJIA compounded to a gain of $1,068,826 for November-April in 69 years compared to just $1,461 for May-October. The same hypothetical $10,000 investment in the S&P 500 compounded to $823,326 for November-April in 69 years compared to a gain of just $9,537 for May-October. May has been a tricky month over the years, a well-deserved reputation following the May 6, 2010 “flash crash”. It used to be part of what we called the “May/June disaster area.” From 1965 to 1984 the S&P 500 was down during May fifteen out of twenty times. Then from 1985 through 1997 May was the best month, gaining ground every single year (13 straight gains) on the S&P, up 3.3% on average with the DJIA falling once and two NASDAQ losses. In the years since 1997, May’s performance has been erratic; DJIA up eleven times in the past twenty-two years (three of the years had gains in excess of 4%). NASDAQ suffered five May losses in a row from 1998-2001, down – 11.9% in 2000, followed by twelve sizable gains in excess of 2.5% and five losses, the worst of which was 8.3% in 2010. Election Year Mays rank at or near the bottom, registering net losses on DJIA and S&P 500 (since 1952), NASDAQ (since 1972) and Russell 1000 and 2000 (since 1980).
Another Regional Fed Record Low Mon, Apr 27, 2020 What an encore. Following the largest decline ever and a record low reading in the Dallas Fed's Manufacturing index for March, April's fared even worse as the headline index fell even further, dropping from -70 to -73. Although that is a new record low and far from a sign of good conditions in that region, it wasn't as bad as the consensus reading of -75. That thin silver lining aside, this month's report was broadly weak. In addition to the headline number, the indices for current conditions on Production, Capacity Utilization, New Orders and New Order growth, and Shipments were all at their lowest levels ever (since 2004), and the six-month outlook indices for Company Outlook and Wages and Benefits were also both at record lows. As for the rest of the categories, if they were not at a record low they were close with most coming in at the 5th percentile or lower of all readings. As we have frequently highlighted over the past month for other regional Fed indices, demand has been absolutely crushed. Each of the categories for demand like those for new and unfilled orders and shipments have completely collapsed. On the bright side, although they still remain low and the increases were small, there was a slight pickup in the outlook indices across these indicators in April. The same can also be said for the categories covering employment. Outlook six months ahead for Employment and Hours Worked both picked up in April. Albeit, these are still around some of their lowest levels since the global financial crisis, and the index for current hours worked continued to fall in April as shutdowns remain in place.
Historically Bad GDP For Q1 As Services Suffer Wed, Apr 29, 2020 This morning the BEA released one of the worst GDP numbers in the modern history of the US economy. Total output fell 4.79% at annual rates, worse than the 4.0% decline (again, at annual rates) that economists had forecasted. That's not as bad as the worst numbers from the financial crisis, but the details under the hood were qualitatively different. Instead of being led down by falling capital expenditures and gradually decelerating consumption, Q1 saw an outright collapse in services and durable goods spending with large but historically less remarkable declines in business spending. Services spending has been the steady Eddie of US growth for the better part of a century now: it's never contributed more than an 82 bps headwind to total QoQ SAAR growth since World War 2. But in Q1 it was a 5% headwind, almost an order of magnitude more than its worst quarter in the global financial crisis. We outline the sobering numbers in more detail below.
Typical May Trading: A Month of Transition with Negative Predisposition Once again it is that time of the year where we increasingly see and hear “Sell in May.” But when exactly in May is the “best” time to sell? Based upon the last 21 years of data, the best time could be early in May. The month has opened well, on average, recently with strength on the first trading day and on the second trading day for the most part. Afterwards, DJIA, S&P 500, NASDAQ, Russell 1000 and Russell 2000 all tend to drift sideways to lower until around May’s sixteenth trading day or so. It is on this day the NASDAQ and Russell 2000 begin to rally to finish the month. Beware though, this late-May rally typically fizzles before it can exceed the highs reached earlier in the month.
Over 30 Million Initial Jobless Claims Filed Thu, Apr 30, 2020 Initial jobless claims fell this week for the fourth week in a row since the record high print of 6.867 million back in the final week of March. Claims came in at 3.839 million which was 588K less than last week but also above estimates of 3.5 million. That brings the total number of initial jobless claims filed over the past six weeks (when there was the first print over 1 million) to over 30 million which is over 9% of the entire US population. On a non-seasonally adjusted basis, this week was only the third consecutive decline off of the peak of 6.211 million which came one week later than the seasonally adjusted data's peak. This week's decline brought jobless claims down to 3.489 million. Since the huge spike in jobless claims began in late March, the 4-week moving average has risen every single week, except for today. This week, the four week average fell 753.25K down to 5.033 million. As with the other readings, while that is still lower, the moving average remains extremely high compared to the rest of history.
Was Warren Right? Tue, May 5, 2020 Warren Buffett made headlines over the weekend when he disclosed that he had completely sold out of his airline holdings. While some questioned whether he was selling out at the lows, for now at least, he appears to be right. As shown in the chart below, the S&P 500 Airlines Group is currently right near its COVID lows, and a close at current levels would actually be a new closing low. Given the national stay at home orders still in place, it's not much of a surprise that airlines haven't bounced, but the same could be said for a number of the other 60+ S&P 500 industries. The table below lists the S&P 500 industries that are currently at their lowest levels relative to their 52-week range. At a level of just 3%, the Airlines group is trading by far the lowest level relative to its 52-week range. The next closest sector, Energy Equipment & Services, is currently trading at much higher levels versus its 52-week range at 20%. In total, there are just 13 industries that are trading in the lower third of their 52-week range, and most of them are clustered in the Financials and Industrials sectors. On the upside, the fact that there are more groups in the upper third of their 52-week range than the lower third suggests positive breadth. Sectors represented on this part of the list include Consumer Discretionary, Technology, and Health Care. While Airlines is the only industry on the verge of making new lows, there is also just one industry on the verge of making new highs too. The Metals and Mining industry definitely wasn't spared during the February and March declines, but the magnitude of the bounce has been even stronger with a gain of over 66% off the lows.
ISM Collapse All Around Tue, May 5, 2020 While the results for ISM's Non-Manufacturing index for the month of April were better than expected (the index was forecasted to come in at 38), the index fell 10.7 points from March down to 41.8. That is the fifth lowest level of the index since its inception in 1997 and the weakest reading since March of 2009. April also marked the first time that the headline number showed a contractionary reading (those below 50) since December of 2009. As shown in the table below, not only was the headline number at one of its lowest levels ever, but the 10.7 point decline was the largest month over month decline ever recorded. The same can also be said for the month over month declines for multiple sub-indices including those for Business Activity, New Orders, Employment, and the Manufacturing and Non-Manufacturing composite. On the other hand, the indices for Supplier Deliveries and Inventory Sentiment experienced their largest one month gains ever, but those are not necessarily positives as detailed later on. Breadth across each of the sub-indices was overall very weak in April as several readings are now at or near some of their weakest levels on record with most now sitting in contractionary territory. One of the weakest areas of last month's report was Business Activity. The index only registered a reading of 26 in April compared to 48 in March. That 22 point decline is nearly double the previous largest MoM decline of 11.8 points from December of 2007 to January of 2008. This index is now sitting at its lowest level ever, surpassing the previous low of 34.4 from November of 2008. The same can also be said for New Orders. Demand in the services sector has been crushed by the COVID pandemic and the New Orders index fell to 32.9. That is also a record low, and the 20 point drop was the largest on record which follows last month's 10.2 point decline which is now the third biggest MoM drop. Only 2 industries surveyed, Public Administration and Finance & Insurance, reported growth in New Orders with another 16 reporting contraction indicating the weakened demand is occurring throughout the service sector. In addition to weak demand, the supply side has also taken a hit. Supply chains appear to have become severely disrupted from the pandemic. The index for Supplier Deliveries is the only one that is inverse (readings above 50 indicate slower deliveries rather than faster) in the ISM report. In other words, the index rises when it takes longer for suppliers to make deliveries which as ISM details "is typical as the economy improves and customer demand increases". But in the current scenario, the index has surged by the most of any month in its history to a record high of 78.3 on a function of supply weakness rather than demand strength (as indicative in New Orders' collapse). Supply chains have faced serious headwinds as all 18 industries reported slowed deliveries in April. This record high in Supplier Deliveries, which again in more normal circumstances could be taken as a positive sign for the economy, helped to bolster the weak headline number. With weak demand and long lead times, industries are reporting inventories as too high. The index for Inventory Sentiment saw a massive surge rising 14.8 points to 62.6. While that is only the highest level since July 2017, this furthers the point that supply chains are facing trouble and demand remains weak. That large increase in the number of industries reporting inventories as too high comes as inventories are appearing to be drawn down as that index is still at a contractionary reading of 46.9. Employment has also collapsed with a second consecutive reading below 50. All 18 industries reported that there was a decline in employment which sent the index to a record low of 30. As with the other categories mentioned, this was also the largest monthly decline ever. In the commentary section, one responder made an interesting note stating business has been impaired "due to employees afraid to work side-by-side with other employees." After today's release and last Friday's release of the manufacturing counterpart, the composite index is now at its weakest level (41.8) since March of 2009 (39.8) with the 10.3 point decline from last month the largest one month decline ever recorded for the composite. All in all, the two reports for the month of April gave a very bleak outlook for the US economy. ©2
10% of Population Filing Jobless Claims Thu, May 7, 2020 Initial jobless claims fell for a fifth consecutive week down from 3.839 million last week to 3.169 million this week. While that is still extremely high by historical standards, that is the lowest print since jobless claims starting printing more than a million in late March. In other words, a massive number of people are continuing to file for unemployment insurance, but the pace of increase has slowed down with the peak in the rearview for the time being. On a non-seasonally adjusted basis, this week was only the fourth consecutive decline with claims now down to 2.849 million. As with the seasonally adjusted number, this week's print was the smallest of the past several weeks since claims first began to print in the millions but is still very high compared to the rest of history. The four-week moving average has also seemed to have peaked for the time being as it fell for a second straight week. The moving average is now down to 4.174 million compared to over 5 million last week. While it could be expected given the massive volatility of the indicator over the past couple of months, that 859.75K decline from last week is the largest single week decline on record (second chart below). Since the March 20th print (the first one in the millions) to today, a total of 33.483 million initial jobless claims have been filed. That is more than 10% of the US population and over 20% of the labor force!. As such, continuing jobless claims (lagged by a week) came in at another record of 22.647 million this week.
US Economic Recessions Mon, May 11, 2020 With numbers like we saw in last Friday's employment report, the current downturn no doubt ranks as one of, if not, the steepest downturns in US history. What started in March still hasn't been officially designated a recession, but unlike prior downturns when investors and economists usually couldn't even initially agree on whether the economy was in a recession or not, the only question this time around is how long and deep it will be. Because the drop-off in economic activity was so sharp and sudden, by some measures we may have already seen the depths of the contraction. If that ends up being the case, though, one could even argue that whatever it is we've been going through the last several weeks doesn't even meet the criteria for a recession. Huh? According to the NBER, which is the organization in charge of classifying recessions, they are characterized as "a significant decline in economic activity spread across the economy, lasting more than a few months, normally visible in real GDP, real income, employment, industrial production, and wholesale-retail sales." Economic activity essentially peaked in February to early March before the lockdowns started, and from there, the economy came to an absolute standstill. With parts of the economy already starting to open up, though, if that trend continues, the trough of the contraction would have occurred in April. At that rate, the entire contraction would have spanned a period of two months at most. While the definition of 'few' isn't a specific number, if a couple is two, then a few would be three or more. Whether the length of the economic contraction ultimately meets the criteria for an official recession is irrelevant. The fact is that never before have we seen such a large number of Americans lose their jobs in such a short period of time, and if businesses aren't able to get back up and running soon, the dislocations it causes will take years to repair. So, how long does a pandemic induced recession typically last? Unfortunately, there aren't a lot of examples on that front. The chart below shows the length of US economic recessions going back to 1900, and while the recession brought on by the Spanish Flu back in 1918 wasn't the shortest recession of the last 120 years, at just seven months, it was the second shortest and wasn't even long enough to result in a bear market for the Dow. As crazy as the market's rebound off the lows has been in the last six weeks, can you imagine if Twitter was around in 1918 and in the midst of that pandemic the DJIA never even saw a peak to trough decline of 11%?
May Option Expiration Week: Challenging since 2009 Trading around May option expiration is mostly a mixed bag. DJIA has been down twenty of the last thirty-eight May expiration days with an average loss of 0.13%. The full week has a bearish bias for DJIA and S&P 500 with records of 21 declines and 17 advances over the past 38 years. More recently, DJIA has suffered declines in nine of the past eleven expiration weeks. S&P 500 recent record has eight losses in the past eleven expiration weeks.
Reversal to Neutral Wed, May 13, 2020 Currently, the only two sectors that are trading in overbought territory (1+ standard deviations above its 50-DMA) in our Trend Analyzer are Technology and Communication Services. Meanwhile, Financials, Industrials, Utilities and Real Estate are all closer to oversold territory than overbought without any of those sectors yet to have even moved above their 50-DMAs. As for short-term breadth levels as measured by the 10-day advance/decline line, as shown in the charts from our Daily Sector Snapshot below, those same sectors in addition to Consumer Discretionary and Materials have actually begun to touch oversold levels after the late day reversal lower yesterday and a lack of a push higher with lackluster breadth over the prior few days. Even the sectors that have been market leaders like Health Care, Communication Services, and Tech have more neutral 10-day A/D line readings at the moment even with prices at or near overbought levels.
When the rug gets pulled, the circuit breakers will trip. What's the max % market decline before the weekly breaker hits?
Sentiment Still Sits Bearish Thu, May 14, 2020 In last Thursday's Chart of the Day, we noted that for the first time in a decade less than a quarter of respondents were bullish or neutral while over half were bearish. Although that isn't exactly the case again this week, sentiment still leans heavily bearish. Bullish sentiment dropped down further to 23.31% from 23.67% last week. That is the lowest level of bullish sentiment since the COVID-19 pandemic began and the lowest number since last October when the percentage of investors reporting as bullish bottomed out at 20.31%. The move out of the bullish camp didn't exactly flow into the bearish camp, though. Bearish sentiment likewise fell slightly this week from 52.66% to 50.61%. Although that is slightly lower, the majority of those surveyed are still bearish. Simply put, bearish sentiment has been high for a long time. Of the previously mentioned past ten weeks, even if bearish sentiment was not above 50% it was at least one standard deviation above its historical average. That now ranks as the longest streak since 2008 (14 weeks) and the fourth-longest in the history of the survey daying back to 1987. With such a large divergence between bulls and bears, the bull-bear spread now stands at -27.3 which is slightly better than the -28.99 reading last week but still clearly at the low end of the past decade's readings.
