WINNER, WINNER IT'S HOUSING FOR DINNER

Transcription

1 HARDLINES RETAIL Food Retail Market Underweight Broadlines Retail Market Underweight Drug Retail Market Overweight Hardlines Retail Market Overweight WINNER, WINNER IT'S HOUSING FOR DINNER Higher Wages Should Disproportionately Help Housing Investment Conclusion. For the first time in years there are some signs Labor s percentage of the National Income (i.e., non-corporate earnings) may move higher as wage growth has generally been accelerating and Median Income rose fairly quickly in In addition, the new administration has committed to bring back higher paying, middle-income jobs to the U.S. As such, we thought it would be interesting to see which industries could benefit the most from sustained, higher wage growth. Using the BLS Consumer Expenditure Survey income brackets and spending levels as a proxy for marginal propensity to consume (MPC), Housing comes out as the biggest beneficiary of rising incomes. We view this as potentially very good news for our housing related Hardliners like HD, LOW, and SHW. If rising labor costs do materialize and are sustained, this could put a premium on companies, in our opinion, that can generate strong revenue growth, something the competitive climate may restrain for Staple Retailers like Wal- Mart and Kroger. Housing could be an outsized beneficiary if Labor commands more of a faster growing National Income. Over the past 15 years, our analysis indicates that Housing has registered a higher Marginal Propensity to Consume (MPC), across all income brackets, than any other category the BLS tracks in its Consumer Expenditure Survey. The MPC for Housing reaches the high-teens among consumer units with average income before taxes between $30,000 and $99,999. This suggests that for every incremental dollar of income that a consumer unit makes in the $30,000 to $99,999 income range, it will spend approximately $0.16 on housing related purchases. This bodes very well for HD, LOW, and SHW. An extra 100bps of comp for LOW, HD, and SHW would yield a 2.4%, 1.6%, and 0.9%, increase in EPS, respectively, in FY18. Our analysis is predicated upon a 2 incremental margin for each company, which seems reasonable to us based on the fact that all three companies operate in a mature industry and each company would not have to change its cost structure much (if at all) if comps accelerated 100bps. An additional 100bps in comps, holding all else equal, could yield up to $0.11 in incremental EPS in FY18 for LOW; HD could generate up to $0.11 in FY18 EPS; and SHW could produce up to $0.12 in additional EPS in FY18. Company Price Rtg Amazon.com, Inc.(AMZN) $ OP Best Buy Co., Inc.(BBY) $43.97 PP Costco Wholesale Corporation(COST) Dollar General Corporation(DG) Lowe's Companies, Inc.(LOW) Sprouts Farmers Market, Inc.(SFM) Target Corporation(TGT) The Home Depot, Inc.(HD) $ PP $76.02 UP $73.90 OP $19.06 PP $64.81 PP $ OP The Kroger Co.(KR) $33.21 UP The Sherwin-Williams Co.(SHW) United Natural Foods, Inc.(UNFI) Wal-Mart Stores Inc.(WMT) $ OP $45.57 PP $67.77 UP Whole Foods Market, $30.03 UP Inc.(WFM) Source: FactSet/Wolfe Research OP=Outperform, PP=Peer Perform, UP=Underperform, NR=Not Rated Scott Mushkin (646) smushkin@wolferesearch.com Michael Otway (646) motway@wolferesearch.com Cody T. Ross (646) cross@wolferesearch.com Benedict Shim (646) bshim@wolferesearch.com DO NOT FORWARD DO NOT DISTRIBUTE DOCUMENT CAN ONLY BE PRINTED TWICE This report is limited solely for the use of clients of Wolfe Research. Please refer to the DISCLOSURE SECTION located at the end of this report for Analyst Certifications and Other Disclosures. For Important Disclosures, please go to or write to us at Wolfe Research, LLC, 420 Lexington Avenue, Suite 648, New York, NY WolfeResearch.com Page 1 of 12

2 Labor as a % of National Income Madeline Kimball - mkimball@wolferesearch.com - Do not forward Investment Conclusion If Labor gains so too will housing. For the first time in years there are some signs Labor s percentage of the National Income (i.e., non-corporate earnings) may move higher. Wage growth has generally been accelerating and Median Income rose sharply in In addition, the new administration has committed to bring back higher paying, middle-income jobs to the U.S. Given these factors, we thought it would be interesting to see which industries could potentially benefit the most from sustained, higher wage growth. Winner, winner, it s housing for dinner! Indeed, using the BLS expenditure survey income brackets and spending levels as a proxy for marginal propensity to consume (MPC), Housing comes out as the biggest beneficiary of rising incomes with an MPC of 15.3% (on average across all income levels) compared to the next closest category of Transportation with an MPC of 10.1% (a complete description of our methodology starts on page 9 of this report). Our analysis suggests that consumer units report that they could spend an additional $0.15 (on average across all income levels) on Housing related purchases and $0.10 on Transportation purchases, for each incremental dollar of income. We view this as potentially very good news for our housing related Hardliners like Home Depot, Lowe s and Sherwin-Williams. We would further note, that if rising labor costs do materialize and are sustained, this could put a premium on companies, in our opinion, that can generate strong revenue growth, something the competitive climate may restrain for our Staple Retailers like Wal-Mart and Kroger. Can Labor See More of the Economic Fruits? Labor compensation has been squeezed for decades. Labor as a percentage of National Income (i.e., noncorporate earnings) peaked in 1970 at 65.9% and declined to a trough of 59.5% in The most recent results (2014) show the ratio has climbed slightly to 60.4%, clearly well below peak levels (Exhibit 1). In addition, according to the U.S. Census Bureau, the Real Median Household Income in the U.S. still remains below 1999 levels ($57,909), despite increasing approximately 5.2% to $56,516 in In Exhibit 2 we depict the trend in total Real Median Household Income in the U.S. since We do note, however, that lower labor costs have helped push corporate profit margins higher. The question looms whether the recent gains in wages can continue and reverse this decades-old decline for labor compensation, and direct more of the fruits of a growing economy to labor. Exhibit 1: Labor as a Percentage of National Income 68% Exhibit 2: Real Median Household Income in the U.S. $60,000 66% $58,000 64% 62% 6 58% 56% Labor $56,000 $54,000 $52,000 $50,000 $48,000 Source: University of Groningen and University of California Davis Source: U.S. Census Bureau WolfeResearch.com Page 2 of 12

3 More recently, there have been some signs Labor is doing better. As noted on the prior page, Median Income rose strongly in 2015 and the recent data show low-end wages are growing at a higher level. In Exhibit 3 below, we illustrate the year-over-year change in Weekly Earnings of Full-Time Employees of the bottom quartile (the bottom 25% of earners) on the left. Looking at the chart, we can see that the bottom quartile of earners have seen the largest increase in 2016 since Further, a large part of President Trump s campaign platform is focused on bringing back better paying, middle-income jobs to the U.S. While it remains unclear whether this will transpire, the middle two quartiles have seen fairly lackluster wage gains since the end of the Great Recession (Exhibit 3 right side). If Labor can see sustained gains, we believe that this will put a premium on companies that can drive strong sales growth. To understand better where those incremental earnings could be spent and which companies could benefit in our coverage universe, we analyzed marginal propensity to consume (MPC) across income demographics. To conduct the analysis we used the BLS consumer expenditure survey as a proxy for MPC. A complete description of our methodology starts on page 9 of this report. Exhibit 3: Y/Y Change in Weekly Earnings of Full-Time Employees Bottom Quartile and Y/Y Change in Weekly Earnings of Full- Time Employees for the Combined Second and Third Quartiles 6% 5% 4% 3% 2% 1% -1% -2% 6% 5% 4% 3% 2% 1% Y/Y Change in Weekly Earnings of Full-Time Employees (Bottom 25% of Earners) Y/Y Change in Weekly Earnings of Full-Time Employees (2nd and 3rd Quartile of Earners) Source: Bureau of Labor Statistics Consumer Expenditure Survey Results The Marginal Propensity to Consume (MPC) is high in middle-income households. On average, from 2001 through 2015, the MPC remains at high levels until income reaches $150,000 and More. It is the highest among the $50,000 to $69,999 income level (62.3%), followed by the $30,000 to $39,999 income level (60.7%). The lowest MPC from 2001 through 2015 is the $150,000 or More income level (35.8%). If we were to breakdown the average MPC from 2001 to 2015 into three distinct five year increments ( , , and ), the MPC, across all income levels and across the three time frames we measured, generally behaved consistently. MPC was the highest, not surprisingly, among the middle/lower income levels and declines sharply as income levels rises above $150,000. WolfeResearch.com Page 3 of 12

4 Marginal Propensity to Consume Marginal Propensity to Consume Madeline Kimball - mkimball@wolferesearch.com - Do not forward Exhibit 4: The average marginal propensity to consume and by 5-year period: , , $15,000 to $29,999 $30,000 to $39,999 $40,000 to $49,999 $50,000 to $69,999 $70,000 to $99,999 $100,000 to $149,999 $150,000 and More $15,000 to $29,999 $30,000 to $39,999 $40,000 to $49,999 $50,000 to $69,999 $70,000 to $99,999 $100,000 to $149,999 $150,000 and More Note: The five year segments represent the average MPC during the five year period. The average MPC for the income brackets $70,000 and higher in do not include data from 2001 or 2002 because the BLS did not breakout the higher income brackets those years. To calculate the average MPC for the five year period we calculated the weighted average income and expenditures each year. We then calculated the difference between expenditures for each specific category in Income Bracket B and Income Bracket A (where Income Bracket B was one income bracket higher than Income Bracket A) and the difference in income from Income Bracket B and Income Bracket A. Finally we divided the change in expenditure for Category 1 by the change in Income. Source: Bureau of Labor Statistics, Wolfe Research Winner, Winner it s Housing for Dinner Housing could be an outsized beneficiary if Labor commands more of a faster growing National Income. It is no surprise that our Marginal Propensity to Consume (MPC) analysis indicates that lower/middle income consumer units have a higher marginal propensity to consume than higher income consumer units. What was surprising, however, was how much housing may stand to benefit. In fact, the MPC for housing reaches the high-teens among consumer units with average income before taxes between $30,000 and $99,999. Housing recorded the highest reading we saw across the 13 different expenditure categories measured by the BLS. This bodes very well for HD, LOW, and SHW, indeed. Housing has the highest Marginal Propensity Consume (MPC). Over the past 15 years, Housing has registered a higher MPC, across all income brackets, than any other category the BLS tracks in its Consumer Expenditure Survey. In Exhibit 5 on the next page we illustrate the outsized MPC in the Housing category compared to the next eight largest categories, on average from For reference, the average MPC on Housing across all income levels is approximately 15%, while the second largest bucket (in terms of MPC) is Transportation at roughly 1. Further, the range in average MPC from on Housing across all income levels is 610bps, peaking at 17.3% in the $15,000 to $29,999 income level and dipping to 11.2% in the $150,000 and More income level. Simply put, if you make more money, you spend a lot more on housing according to the consumer expenditure survey. WolfeResearch.com Page 4 of 12

5 Marginal Propensity to Consume Marginal Propensity to Spend on Housing Madeline Kimball - mkimball@wolferesearch.com - Do not forward Exhibit 5: Average MPC by Category Exhibit 6: MPC on Housing Over Three 5-Year Time Periods 2 15% 1 5% $15,000 to $30,000 to $40,000 to $50,000 to $70,000 to $100,000 to $150,000 $29,999 $39,999 $49,999 $69,999 $99,999 $149,999 and More -5% Food Housing Apparel and Services Transportation Healthcare Entertainment 2 18% 16% 14% 12% 1 8% 6% 4% 2% $15,000 to $29,999 $30,000 to $39,999 $40,000 to $49,999 $50,000 to $69,999 $70,000 to $99,999 $100,000 to $149,999 $150,000 and More Personal Care Products and Services Edcuation Personal Insurance and Pensions Note: The five year segments represent the average MPC during the five year period. The average MPC for the income brackets $70,000 and higher in do not include data from 2001 or 2002 because the BLS did not breakout the higher income brackets those years. To calculate the average MPC for the five year period we calculated the weighted average income and expenditures each year. We then calculated the difference between expenditures for each specific category in Income Bracket B and Income Bracket A (where Income Bracket B was one income bracket higher than Income Bracket A) and the difference in income from Income Bracket B and Income Bracket A. Finally we divided the change in expenditure for Category 1 by the change in Income. Source: Bureau of Labor Statistics, Wolfe Research Note: The five year segments represent the average MPC during the five year period. The average MPC for the income brackets $70,000 and higher in do not include data from 2001 or 2002 because the BLS did not breakout the higher income brackets those years. To calculate the average MPC for the five year period we calculated the weighted average income and expenditures each year. We then calculated the difference between expenditures for each specific category in Income Bracket B and Income Bracket A (where Income Bracket B was one income bracket higher than Income Bracket A) and the difference in income from Income Bracket B and Income Bracket A. Finally we divided the change in expenditure for Category 1 by the change in Income. Source: Bureau of Labor Statistics, Wolfe Research As income increases so too does the desire own a home. Consumer units across all income levels have a higher MPC for Owned Dwellings than Rented Dwellings. If we were to focus on the largest subcategory of Housing, which is Shelter, we can see that Shelter is broken down into three subcategories: (1) Owned Dwellings, (2) Rented Dwellings, and (3) Other Lodging (includes all expenses for vacation homes, school, college, hotels, motels, and other lodging while out of town). In Exhibit 7 on the next page we can see that the MPC of Owned Dwellings represents the overwhelming majority of the MPC for Shelter. In fact, the MPC on Owned Dwellings is 6.8% for the $15,000 to $29,999 income level, steadily rises to a peak of 10.8% at the $50,000 to $69,999 income level, and gradually falls off to 5.6% for the $150,000 and More income level. Meanwhile, the MPC on Rented Dwellings peaks at 1.5% in the $15,000 to $29,999 income level, gradually declines to a trough of -2.7% for the $50,000 to $69,999 income level, and slowly increases to -0.1% for the $150,000 and More income level. This data suggests that demand for housing and housing turnover could increase, both of which are historically positive indicators for sales at Home Depot, Lowe s and Sherwin- Williams. WolfeResearch.com Page 5 of 12

6 Marginal Propensity to Spend on Shelter Madeline Kimball - mkimball@wolferesearch.com - Do not forward Exhibit 7: Average Marginal Propensity to Spend on Shelter % 1 8% 6% 4% 2% -2% -4% $15,000 to $29,999 $30,000 to $39,999 $40,000 to $49,999 $50,000 to $69,999 $70,000 to $99,999 $100,000 to $149,999 $150,000 and More Shelter Owned Dwellings Rented Dwellings Other Lodging Note: The average MPC for the income brackets $70,000 and higher in does not include data from 2001 or 2002 because the BLS did not break out the higher income brackets those years. To calculate the average MPC for the five year period we calculated the weighted average income and expenditures each year. We then calculated the difference between expenditures for each specific category in Income Bracket B and Income Bracket A (where Income Bracket B was one income bracket higher than Income Bracket A) and the difference in income from Income Bracket B and Income Bracket A. Finally we divided the change in expenditure for Category 1 by the change in Income. Source: Bureau of Labor Statistics, Wolfe Research We estimate consumer units exhibited a 3% MPC for Home Improvement. While there is not a category designated home improvement we added together expenditures on Maintenance, Repairs, Insurance, and Other Expenses, Other Household Products, Floor Coverings, Major Appliances, Small Appliances and Miscellaneous Housewares, and Miscellaneous Household Equipment. At first blush a MPC of 3% seemed a little lower than we would have expected as HD, LOW, and SHW exhibit strong correlations to various housing metrics (discussed later). However, we would note that this survey reflects consumers belief about what they spend and not actual spending data. Therefore, we believe it might be as important to look at the slope of the MPC line rather than the actual MPC number. Under this lens, we point out that consumer units exhibited a 2.8% MPC for Home Improvement in the $30,000 to $39,999 income level and generally increased to 3.4% MPC for Home Improvement in the $100,000 to $149,999 income level. WolfeResearch.com Page 6 of 12

7 Marginal Propensity to Consume Madeline Kimball - mkimball@wolferesearch.com - Do not forward Exhibit 8: Average Marginal Propensity to Consume Home Improvement % 12% 8% 4% $15,000 to $29,999 $30,000 to $39,999 $40,000 to $49,999 $50,000 to $69,999 $70,000 to $99,999 $100,000 to $149,999 $150,000 and More Housing Home Improvement Note: The average MPC for the income brackets $70,000 and higher in do not include data from 2001 or 2002 because the BLS did not breakout the higher income brackets those years. To calculate the average MPC for the five year period we calculated the weighted average income and expenditures each year. We then calculated the difference between expenditures for each specific category in Income Bracket B and Income Bracket A (where Income Bracket B was one income bracket higher than Income Bracket A) and the difference in income from Income Bracket B and Income Bracket A. Finally we divided the change in expenditure for Category 1 by the change in Income. Source: Bureau of Labor Statistics, Wolfe Research In addition to favorable MPC trends in the Housing category, we believe housing fundamentals remain solid and well below prior peak levels. In Exhibit 9 we depict several key economic indicators over time and track correlations (from 2001-present) against the same store sales of HD, LOW, and SHW. Looking at the current level for the six economic indicators and comparing them to peak levels (from 2002 to Present), we can see that they are all below peak levels, with Private Fixed Residential Investment -19% below peak; Housing Starts -43% below peak; and New Home Sales -56% below peak. Exhibit 9: Select Housing Indicators vs Retailer Same Store Sales Correlation ( Present) HD SSS LOW SSS SHW SSS Current Source: BEA, U.S. Census Bureau, S&P, Federal Reserve, BLS, Company Reports, Wolfe Research Peak ('02 - '11) Peak versus trough analysis Trough ('02 - '11) Current y/y Δ Current vs Peak Housing metrics PFRI % (19%) Housing Starts % (43%) New Home sales % (56%) Existing Home sales (2 quarter lag) % (23%) Pricing Case -Schiller 20 City (7%) Consumer health Unemployment Rate (0.57) (0.44) (0.82) 4.7% 4.4% 9.9% -30bps (6%) WolfeResearch.com Page 7 of 12

8 An extra 100bps of comp sales for each of LOW, HD, and SHW would yield a 2.4%, 1.6%, and 0.9%, increase in EPS, respectively, in FY18. Below in Exhibit 10 we depict the potential impact that an additional 100bps of comparable store sales would have on LOW s, HD s, and SHW s FY17 and FY18 EPS. Our analysis is predicated upon a 2 incremental margin for each company, which seems reasonable to us based on the fact that all three companies operate in a mature industry and each company would not have to change its cost structure much (if at all) if comps accelerated 100bps. According to our analysis, we estimate than an additional 100bps of comp sales for each of LOW, HD, and SHW equates to up to $0.11, $0.11, and $0.12 in EPS in FY18, respectively. Exhibit 10: 100bps Incremental Comp Impact on LOW s, HD s, and SHW s EPS LOW EPS Impact Under Increased Comps FY17E FY18E Net Sales $68,187 $71,666 HD EPS Impact Under Increased Comps FY17E FY18E Net Sales $99,994 $107,270 Sales Impact from Extra 100bps of Comp $290 $679 Sales Impact from Extra 100bps of Comp $941 $1,010 Profit Impact from Extra 100bps of Comp $58 $136 Profit Impact from Extra 100bps of Comp $188 $202 Taxes $21 $50 Taxes $70 $75 Net Income $36 $86 Net Income $119 $127 EPS Impact from Extra 100bps of Comp $0.04 $0.11 EPS Impact from Extra 100bps of Comp $0.10 $0.11 Weighted Avg. Shares Outstanding Weighted Avg. Shares Outstanding 1,163 1,108 Tax Rate Stock Price Impact Analysis Estimated EPS Impact from Extra 100 bps of Comp $0.04 $0.11 Times: Current NTM P/E Multiple 16.2x 16.2x Estimated Impact to LOW Share Price $0.70 $1.74 Tax Rate Stock Price Impact Analysis Estimated EPS Impact from Extra 100 bps of Comp $0.10 $0.11 Times: Current NTM P/E Multiple 19.4x 19.4x Estimated Impact to HD Share Price $1.98 $2.23 Estimated Impact to LOW Share Price $0.70 $1.74 Divide: LOW Current Share Price $73.90 $73.90 % Impact to LOW Shares 0.9% 2.4% Estimated Impact to HD Share Price $1.98 $2.23 Divide: HD Current Share Price $ $ % Impact to HD Shares 1.4% 1.6% SHW EPS Impact Under Increased Comps FY17E FY18E Net Sales $12,502 $13,103 Sales Impact from Extra 100bps of Comp $78 $83 Profit Impact from Extra 100bps of Comp $16 $17 Taxes $5 $5 Net Income $11 $11 EPS Impact from Extra 100bps of Comp $0.11 $0.12 Weighted Avg. Shares Outstanding Tax Rate 31.5% 31.5% Stock Price Impact Analysis Estimated EPS Impact from Extra 100 bps of Comp $0.11 $0.12 Times: Current NTM P/E Multiple 22.0x 22.0x Estimated Impact to SHW Share Price $2.48 $2.65 Estimated Impact to SHW Share Price $2.48 $2.65 Divide: SHW Current Share Price $ $ % Impact to SHW Shares 0.8% 0.9% Note: Figures represent potential impact to Wolfe Research s FY17 and FY18 EPS estimate Source: Wolfe Research estimates, company filings, FactSet WolfeResearch.com Page 8 of 12

9 Navigating the Consumer Expenditure Survey We aggregated the Consumer Expenditure Surveys from 2001 through 2015 to analyze trends in buying habits. According to the BLS, the Consumer Expenditure Survey collects information from the Nation's households and families on their buying habits (expenditures), income, and household characteristics. The strength of the survey is that it allows data users to relate the expenditures and income of consumers to the characteristics of those consumers. The survey consists of two components, a quarterly Interview Survey and a weekly Diary Survey, each with its own questionnaire and sample. Beginning in 2003 through 2014, the CES broke out consumer expenditures across 13 different income levels. These included: (1) less than $5,000, (2) $5,000 to $9,999, (3) $10,000 to $14,999, (4) $15,000 to $19,999, (5) 20,000 to $29,999, (6) 30,000 to $39,999, (7) 40,000 to $49,999, (8) 50,000 to $69,999, (9) $70,000 to $79,999, (10) 80,000 to $99,999, (11) 100,000 to $119,999, (12) 120,000 to $149,999, and (13) $150,000 and more. In 2001 and 2002, the BLS did not break out the $70,000 and More income level further. We explain later how we adjusted for this. The CES breaks out each income bracket by the number of Consumer Units. The BLS defines a Consumer Unit as consisting of any of the following: (1) All members of a particular household who are related by blood, marriage, adoption, or other legal arrangements; (2) a person living alone or sharing a household with others or living as a roomer in a private home or lodging house or in permanent living quarters in a hotel or motel, but who is financially independent; or (3) two or more persons living together who use their incomes to make joint expenditure decisions. Financial independence is determined by spending behavior with regard to the three major expense categories: Housing, food, and other living expenses. To be considered financially independent, the respondent must provide at least two of the three major expenditure categories, either entirely or in part. Importantly, the BLS also notes how many people are in each consumer unit by income level. For a frame of reference, the 2015 CES (the most recent one published) stated that there were 128,437,000 consumer units consisting of 2.5 people in the U.S. If we were to multiply the two together that would yield million people, which is very close to the million people in 2015 that the US Census Bureau reported. Given the two databases are less than 1% off in terms of the U.S. population, we have confidence in the figures provided in the CES. In order to analyze buying habits by income level we examined the Average Annual Expenditures. The CES breaks out Average Annual Expenditures into 13 categories, which include: (1) Food, (2) Housing, (3) Apparel and Services, (4) Transportation, (5) Healthcare, (6) Entertainment, (7) Personal Care Products and Services, (8) Reading, (9) Education, (10) Tobacco Products and Smoking Supplies, (11) Miscellaneous, (12) Cash Contributions, and (13) Personal Insurance and Pensions. We note that Food, Housing, Apparel and Services, Transportation, Healthcare, Entertainment, and Personal Insurance and Pensions are broken out into further subcategories, but we are not going to list them all out for the sake of brevity. Calculating Marginal Propensity to Consume. Now that we have our 13 categories of income level and their purchasing habits separated into 13 categories, we can analyze how the change of income levels affect purchasing habits. This concept is referred to as Marginal Propensity to Consume, or MPC. MPC equals the change in consumption divided by the change in income level. To put it in laymen s terms: if you make another dollar of income, what percentage of your dollar are you likely to spend on which categories. For example, let s say a person used to make $100 and then received a raise and now makes $200; let s also say that when the person made $100, he/she spent $5 on food and now that he/she makes $200, the person spends $8 on food. For this example, the MPC food is equal to ($8 - $5) divided by ($200 - $100), or 3%. WolfeResearch.com Page 9 of 12

10 How to adjust for the different income levels that the CES captures over time. We analyzed the MPC of consumer units across 13 income levels. However, we grouped these 13 income levels into eight buckets: (1) Less than $15,000, (2) $15,000 to $29,999, (3) $30,000 to $39,000, (4) $40,000 to $49,999, (5) $50,000 to $69,999, (6) $70,000 to $99,999, (7) $100,000 to $149,999, and (8) $150,000 and more. The reason we boiled down the 13 income levels to eight is because in 2015 the BLS changed the income buckets that it reports. For example, in 2015 the BLS consolidated the Less than $5,000, $5,000 to $9,999, and 10,000 to $14,999 income brackets into one bucket, Less than $15,000. Additionally, in 2015, the CES broke out the $150,000 and more income level into two buckets, $150,000 to $199,999 and $200,000 and More. In order to account for these slight differences in the way that the BLS reports the CES data, we applied a weighted average to sub-categories that equaled one category. For example, in 2014 the BLS reported that there were 6,115 consumer units in the Less than $5,000 income level that had an average Income Before Taxes of $2,392; 4,980 consumer units with an average Income Before Taxes of $7,987 in the $5,000 to $9,999 income level; and 8,243 consumer units that had an average Income Before Taxes of $12,876 in the $10,000 to $14,999 income level. Meanwhile in 2015, the BLS reported that the Less than $15,000 income level had an average Income Before Taxes of $8,169. In order to make 2014 comparable to 2015, we: 1. Added the number of consumer units in each income level in 2014 (6, , ,243 = 19,338) 2. Divided the number of consumer units for each income level by the total (6,115 / 19,338 = 31.6%; 4,980 / 19,338 = 25.8%; 8,243 / 19,338 = 42.6%). This gave us the proportion that each of the three income levels represented of the total Less than $15,000 income level. 3. Calculated the weighted average Income Before Taxes (and all expenditures) for the Less than $15,000 income level ((31.6% x $2,392) + (25.8% x $7,987) + (42.6% x $12,876) = $8,302). Below we note our four key assumptions underlying our analysis in which we based our conclusions: 1. Consumer units can only move up income brackets. 2. Consumer units can only move up one income bracket. 3. Because data was not provided for incomes greater than $70,000 in 2001 and 2002, all averages that should include these years were excluded. For example, the average annual expenditures for 2001 through 2005 for $100,000 to $149,999 income level is an average of 2003 through To calculate the average annual expenditures for 2001 through 2015 for the $100,000 to $149,999 income level we calculated an average from 2003 through There are only 7 income buckets for MPC, as the lowest income level would be left out because there is no lower income level of consumer units. WolfeResearch.com Page 10 of 12

11 DISCLOSURE SECTION Analyst Certification: The analyst of Wolfe Research, LLC primarily responsible for this research report whose name appears first on the front page of this research report hereby certifies that (i) the recommendations and opinions expressed in this research report accurately reflect the research analysts personal views about the subject securities or issuers and (ii) no part of the research analysts compensation was, is or will be directly or indirectly related to the specific recommendations or views contained in this report. Other Disclosures: Wolfe Research, LLC Fundamental Stock Ratings Key: Outperform (OP): Peer Perform (PP): Underperform (UP): The security is projected to outperform analyst's industry coverage universe over the next 12 months. The security is projected to perform approximately in line with analyst's industry coverage universe over the next 12 months. The security is projected to underperform analyst's industry coverage universe over the next 12 months. Wolfe Research, LLC uses a relative rating system using terms such as Outperform, Peer Perform and Underperform (see definitions above). Please carefully read the definitions of all ratings used in Wolfe Research, LLC research. In addition, since Wolfe Research, LLC research contains more complete information concerning the analyst s views, please carefully read Wolfe Research, LLC research in its entirety and not infer the contents from the ratings alone. In all cases, ratings (or research) should not be used or relied upon as investment advice and any investment decisions should be based upon individual circumstances and other considerations. Wolfe Research, LLC Sector Weighting System: Market Overweight (MO): Market Weight (MW): Market Underweight (MU): Expect the industry to outperform the primary market index for the region (S&P 500 in the U.S.) by at least 1 over the next 12 months. Expect the industry to perform approximately in line with the primary market index for the region (S&P 500 in the U.S.) over the next 12 months. Expect the industry to underperform the primary market index for the region (S&P 500 in the U.S.) by at least 1 over the next 12 months. Wolfe Research, LLC Distribution of Fundamental Stock Ratings (As of December 31, 2016): Outperform: 39% 1% Investment Banking Clients Peer Perform: 49% 1% Investment Banking Clients Underperform: 12% Investment Banking Clients Wolfe Research, LLC does not assign ratings of Buy, Hold or Sell to the stocks it covers. Outperform, Peer Perform and Underperform are not the respective equivalents of Buy, Hold and Sell but represent relative weightings as defined above. To satisfy regulatory requirements, Outperform has been designated to correspond with Buy, Peer Perform has been designated to correspond with Hold and Underperform has been designated to correspond with Sell. Wolfe Research Securities and Wolfe Research, LLC have adopted the use of Wolfe Research as brand names. Wolfe Research Securities, a member of FINRA ( is the broker-dealer affiliate of Wolfe Research, LLC and is responsible for the contents of this material. Any analysts publishing these reports are dually employed by Wolfe Research, LLC and Wolfe Research Securities. The content of this report is to be used solely for informational purposes and should not be regarded as an offer, or a solicitation of an offer, to buy or sell a security, financial instrument or service discussed herein. Opinions in this communication constitute the current judgment of the author as of the date and time of this report and are subject to change without notice. Information herein is believed to be reliable but Wolfe Research and its affiliates, including but not limited to Wolfe Research Securities, makes no representation that it is complete or accurate. The information provided in this communication is not designed to replace a recipient's own decision-making processes for assessing a proposed transaction or investment involving a financial instrument discussed herein. Recipients are encouraged to seek financial advice from their financial advisor regarding the appropriateness of investing in a security or financial instrument referred to in this report and should understand that statements regarding the future performance of the financial instruments or the securities referenced herein may not be realized. Past performance is not indicative of future results. This report is not intended for distribution to, or use by, any person or entity in any location where such distribution or use would be contrary to applicable law, or which would subject Wolfe Research, LLC or any affiliate to any registration requirement within such location. For additional important disclosures, please see WolfeResearch.com Page 11 of 12

12 The views expressed in Wolfe Research, LLC research reports with regards to sectors and/or specific companies may from time to time be inconsistent with the views implied by inclusion of those sectors and companies in other Wolfe Research, LLC analysts research reports and modeling screens. Wolfe Research communicates with clients across a variety of mediums of the clients choosing including s, voice blasts and electronic publication to our proprietary website. Copyright Wolfe Research, LLC All rights reserved. All material presented in this document, unless specifically indicated otherwise, is under copyright to Wolfe Research, LLC. None of the material, nor its content, nor any copy of it, may be altered in any way, or transmitted to or distributed to any other party, without the prior express written permission of Wolfe Research, LLC. This report is limited for the sole use of clients of Wolfe Research. Authorized users have received an encryption decoder which legislates and monitors the access to Wolfe Research, LLC content. Any distribution of the content produced by Wolfe Research, LLC will violate the understanding of the terms of our relationship. WolfeResearch.com Page 12 of 12