FORECASTING PROJECT for Almost Family, Inc. (AFAM) BUS621 – Business Statistics FP2 Team Eha Bus 621 Almost Family Inc. Forecasting Project Page 1 I. Independent Variables {This team mistakenly did not provide the descriptions and rationale of the variables under consideration. This is a huge omission. Don’t make this mistake!} II. Data and Forecasts The following table shows a compilation of independent economic variables that the group thinks will have a bearing on the Y-value (quarterly revenues) for the forecast model. Each of these variables will be tested on their correlation to each other and the strength by which they affect quarterly revenue values. Year Quarter 2006 2007 2008 2009 2010 2011 2012 2013 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Real Gross Medicare Quarterly Gasoline Regular Consumer Revenues Domestic Allowed Real Disposable Grade Retail Price Incl Unemployment Price Index Commercial U.S. Prime ($Million) Product Expenditures Personal Income Taxes, U.S. Average Rate (%) (Index Employment Lending Rate Number of ($Billion) ($Billion) ($Billion) ($/gal) Units) (Millions) (%) Branches $ 20.79 12,896 $18.60 9,535 2.34 4.7 1.99 89.3 7.43 115 $ 21.85 12,949 $20.10 9,619 2.85 4.6 2.01 89.6 7.9 115 $ 22.95 12,950 $19.40 9,665 2.84 4.6 2.03 90 8.25 115 $ 26.23 13,038 $19.70 9,792 2.26 4.4 2.02 90.4 8.25 115 $ 31.95 13,056 $19.30 9,836 2.37 4.5 2.04 91 8.25 115 $ 32.71 13,174 $20.80 9,851 3.02 4.5 2.07 91.3 8.25 115 $ 32.06 13,270 $20.10 9,890 2.85 4.7 2.08 91.4 8.18 115 $ 35.41 13,326 $20.40 9,944 2.97 4.8 2.1 91.7 7.52 115 $ 39.03 13,267 $20.20 10,087 3.11 5 2.13 91.8 6.21 115 $ 48.70 13,311 $21.70 10,288 3.76 5.3 2.16 91.5 5.08 115 $ 58.71 13,187 $21.00 10,054 3.85 6 2.19 91.1 5 123 $ 66.17 12,884 $21.30 10,048 2.30 6.9 2.14 90.1 4.06 124 $ 69.20 12,711 $21.40 9,927 1.89 8.3 2.12 88.8 3.25 124 $ 74.85 12,701 $23.10 9,916 2.32 9.3 2.13 87.8 3.25 124 $ 76.29 12,747 $22.30 9,760 2.57 9.6 2.15 87.5 3.25 126 $ 77.51 12,873 $22.70 9,746 2.60 9.9 2.17 87.3 3.25 126 $ 81.78 12,948 $23.40 9,882 2.71 9.8 2.17 87.3 3.25 126 $ 85.61 13,020 $25.20 10,034 2.81 9.6 2.17 87.6 3.25 126 $ 84.90 13,104 $24.30 10,063 2.72 9.5 2.18 87.9 3.25 126 $ 84.64 13,181 $24.70 10,087 2.88 9.6 2.2 88.3 3.25 127 $ 82.59 13,184 $24.50 10,196 3.30 9 2.22 88.7 3.25 128 $ 81.72 13,265 $26.40 10,158 3.80 9 2.25 89.2 3.25 128 $ 86.21 13,307 $25.50 10,126 3.63 9.1 2.26 89.5 3.25 166 $ 89.33 13,441 $26.00 10,122 3.37 8.7 2.27 90 3.25 166 $ 89.95 13,506 $25.90 10,214 3.61 8.3 2.28 90.5 3.25 166 $ 86.89 13,549 $27.90 10,292 3.72 8.2 2.29 90.8 3.25 166 13,598 $27.00 10,316 3.67 8.1 2.3 91.2 3.2 166 13,646 $27.40 10,342 3.56 7.8 2.32 91.6 3.33 166 13,706 $21.00 10,388 3.41 7.7 2.32 92 3.34 166 13,761 $22.60 10,456 3.54 7.8 2.32 92.4 3.31 166 With the exception of Number of Branches, the above data and forecasts were obtained from the following sources: Bus 621 Almost Family Inc. Forecasting Project Page 2 Column D Column E Column F Column H Column I Column J Column K Column L Real Gross Domestic Product ($Billion) Bureau of Economic Analysis/Global Insight - http://www.eia.gov/forecasts/steo/query/ http://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/SustainableGRatesConFact/Downloads/sgr2013p.pdf M. Kent Clemens, F.S.A. Centers for Medicare & Medicaid Services Office of the Actuary N3-26-04 Centers for Medicare & Medicaid Services Baltimore, MD 21244 Bureau of Economic Analysis/Global Insight - http://www.eia.gov/forecasts/steo/query/ EIA Weekly Petroleum Status Report - http://www.eia.gov/forecasts/steo/query/ http://data.bls.gov/timeseries/LNS14000000 Medicare Quarterly Allowed Expenditures ($Billion) Real Disposable Personal Income ($Billion) Gasoline Regular Grade Retail Price Incl Taxes, U.S. Average ($/gal) Unemployment Rate (%) Consumer Price Index (Index Units) Commercial Employment (Millions) U.S. Prime Lending Rate (%) Bureau of Labor Statistics/Global Insight - http://www.eia.gov/forecasts/steo/query/ Bureau of Labor Statistics/Global Insight - http://www.eia.gov/forecasts/steo/query/ Federal Reserve Board/Global Insight - http://www.eia.gov/forecasts/steo/query/ Number of Branches data was obtained form the Almost Family Inc Website http://www.patientcare.com/press_releases.php and the forecast was obtained by analyzing company earnings reports which indicated no plans for expansion in the next four quarters hence we expect the number of branches to remain constant at 166. III. Viability of Variables a. Correlation Matrix Multicollinearity is a statistical phenomenon in which two or more predictor variables in a multiple regression model are highly correlated. This was initially measured using the correlation matrix model and any pair of variables with correlation larger than ±0.9 indicates a strong relationship so one of the pair of variables will be deleted to resolve the issue of multicollinearity. Another measure of multicollinearity is the Variance Inflation Factor (VIF) which the group will utilize as well. Correlation results for AFAM variables are as follows: Correlation Matrix Revenues ($Million) Real Gross Domestic Product ($Billion) Medicare Quarterly Allowed Expenditures ($Billion) Real Disposable Personal Income ($Billion) Gasoline Regular Grade Retail Price Incl Taxes, U.S. Average ($/gal) Unemployment Rate (%) Consumer Price Index (Index Units) Commercial Employment (Millions) U.S. Prime Lending Rate (%) Number of Branches Revenues ($Million) 1.000 .168 .901 .631 .295 .944 .895 -.557 -.965 .690 Medicare Real Gasoline Real Gross Quarterly Disposable Regular Grade U.S. Domestic Allowed Personal Retail Price Incl Consumer Commercial Prime Product Expenditures Income Taxes, U.S. Unemployment Price Index Employment Lending Number of ($Billion) ($Billion) ($Billion) Average ($/gal) Rate (%) (Index Units) (Millions) Rate (%) Branches 1.000 .454 .679 1.000 .709 1.000 .810 -.097 .543 .589 -.018 .541 .524 .794 .917 -.353 -.815 .797 .669 .381 .829 .203 -.574 .554 26 sample size ± .388 critical value .05 (two-tail) ± .496 critical value .01 (two-tail) 1.000 .079 .648 .369 -.219 .496 1.000 .729 -.788 -.942 .523 1.000 -.174 -.831 .798 1.000 .616 -.064 1.000 -.590 1.000 Highly Correlated Data Medicare Quarterly Allowed Expenditures & Consumer Price Index Unemployment Rate & US Prime Lending Rate Highly correlated independent variables are highlighted in yellow. These are: Consumer Price Index with Medicare Quarterly Allowed Expenditures, and Unemployment Rate with U.S. Prime Lending Rate. To address the multicollinearity between these pairs of variables, we will initially remove Consumer Price Index (CPI) and Unemployment Rate (UR) from the data set. These were chosen because they have the Bus 621 Almost Family Inc. Forecasting Project Page 3 lowest correlation with revenue and somewhat high correlations with many other variables. This leaves us with five variables that have at least 95% confidence of non-zero correlation (because the sample correlation greater than the critical value of +/- .388) and not too high correlations with other variables. We note two of the remaining variables have sample correlation greater than .9 or less than -.9. All the variables are easily of confidently forecasted. 1. Bus 621 Almost Family Inc. Forecasting Project Page 4