Sales forecasting: teaching with naive approaches.

Melancon, Melissa V. ; Babin, Laurie A.

INTRODUCTION

Projecting income statements and balance sheets is a key element in planning expenditures and arranging appropriate financing sources for a business. The usefulness of these projections is dependent upon the accuracy of sales' estimates. Most business courses, including introductory corporate finance, give only cursory attention to teaching students how to estimate sales' growth; many textbooks ignore sales' forecasting or treat expected sales as an automatically generated figure. Other sources suggest using multiple regressions, logit, probit, or curvilinear models to predict sales (West, 1994). Students are left in a quandary as to how to estimate future sales for business decisions.

Explosive growth in information technology and affordable, graphics driven computers exacerbate the confusion about sales forecasting. Many users embrace the erroneous belief that a sales forecast is a computer function and not a decision process. Lacking both a thorough understanding of the forecasting process and the skills to develop practical projections, students may use subjective measures or skip the forecasting part of planning. This knowledge gap leaves many underprepared to make critical business decisions affecting production, marketing, and financing. Easy to use computers and software programs may result in decreased accuracy as many forecasters misuse or misunderstand how to use the programs (McCarthy, 2006).

The purpose of this study is to present several simple (naive) methods that can be used as a foundation for teaching sales forecasting. The techniques lack statistical sophistication and are admittedly not as accurate as more complex methods; however, the projections provide students and with a foundation for learning to project sales to create sales driven financial statements.

LITERATURE REVIEW

Sales forecasting has been a topic of academic research for decades (Dalrymple, 1975, 1987; Mentzer & Cox, 1984; Mentzer & Kahn, 1995). The advent of computers has increased the number and complexity of sales forecasting models (Dalrymple, 1975). However, although researchers agree upon the importance of sales forecasting, there is little consensus upon which methods are best to use in teaching this valuable skill.

Myriad subjective approaches to projecting sales exist, such as sales force composition, customer expectations, and life cycle analysis (see Mentzer & Cox, 1984; Mentzer & Kahn, 1995: McCarthy, 2006). These judgment-based approaches rely on experienced salespeople and managers to predict the demand and to recognize trends in the markets. Students have not yet developed a knowledge base to draw upon in making inferences about demand, economic conditions, production capabilities, and marketing. Subjective approaches would be of limited value to this group. Supplementing these techniques with quantitative measures provides additional information.

Quantitative measures evolved from assuming constant sales to using an average of past sales data and then employing more complex statistical modeling techniques to account for seasonality, the cyclical nature of business, and other economic factors. The "bottom up" approach suggests interviewing customers and using past sales data to generate the next year's sales (Tichenor & Davis, 2009). Others suggest that sales forecasts include wide-ranging input sources, utilize multiple approaches, and include more objective techniques for projections. This was especially true when the forecasts were reviewed by top management (Ando, 2008; West, 1994). However, recent crises led many to conclude that historical sales and time series data are no longer fair proxies for future consumer behavior (Currie & Rowley, 2010).

Researchers and practitioners developed multiple approaches and adjustments to improve the accuracy of forecasts. Kumar and Patel (2010) cite National Retail Federation reports indicating forecast errors cost U.S. retailers about $200 billion annually. The authors use this economic justification to test clustering sales of multiple retail items into an aggregate sales forecast for the grouping as a means of reducing forecast error associated with individual items. Findings support use of this method to increase the efficiency of sales forecasting. Siriram and Snaddon (2010) test the accuracy of sales forecasts for information technology equipment using linear regression, logistics regression, and Bass curves. The authors conclude that logistics regression using the mean average percentage error was the best method to project new product sales. The results could be improved by including measures that account for market size and demographics, customer service, and technological development.

High technology industries have quick turnover and rapid innovation. Diffusion models and utility theory have been applied to develop a sales forecasting model for this market. A logistics approach is incorporated to measure the accuracy of sales projections for the CD and DVD market. One study suggests that the model improves the accuracy of the sales forecasts by incorporating factors that affect individual buying decisions without adding undue complexity. Further, the model does not require significant econometric expertise (Decker & GnibbaYukawa, 2010). Further adjustments seemed to improve accuracy. However, the increased complexity of these techniques may be overwhelming to students learning the process.

Sales forecasting's impact on functional areas of the firm is evidenced by the market in the aftermath of the recent financial crisis. Retailers and researchers are reeling; previous models of consumer behavior failed. Adjusted earnings forecasts are given more credence by markets when buttressed by sales forecasting. Use of multiple methods is indicated by both researchers and practitioners (Keung, 2011).

Using data derived from several methods can reduce forecast error. One approach is to measure the mean absolute percentage accuracy of sales force participation, consumer sales data from point of sales terminals, and forecasts for inventory orders instead of the error in the measurement. This measure is refined by assigning probability estimates for acceptance to the forecasts as well as more reliance on statistical modeling. Markedly improved forecasting was the result of this hybrid method (Gallucci, Fall 2007).

Examining the number and variety of forecasting methods used shows reluctance among actual preparers to use more complex models (Mentzer & Cox, 1984). Therefore, teaching students to use simple models may overcome this reluctance and build confidence in using more sophisticated techniques. Financial statements for Killing-M-Softly Exterminator Company found in Tables 1, 2, and 3 provide data used to illustrate each of the methods.

SALES FORECASTING METHODS

Constant Sales

The simplest way to project sales is to assume that sales will be unchanged from the previous year. This naive method is easy and may be an accurate process as long as sales remain consistent over time. However, large period-to-period variability (e.g. seasonal sales) or constant growth in sales requires application of more refined estimation techniques (Cornett, Adair, & Nofsinger, 2011). Expected sales for any future time are determined by:

E([Sales.sub.t+k]) = Sales [.sub.t]For all k > 0

Using the sales data provided in the Table 1 and assuming constant sales, the projection for 2011 sales would be $315,350.00 which is what sales were for 2010. The limitation of this method is that it fails to account for the impact of firm or systematic shocks from the economy. Additionally, there is no adjustment for inflation or firm specific changes.

Average Historical Sales

Using average historical sales as a predictor of future sales is a relatively easy forecasting approach. This technique improves upon the constant sales approach by including more data. Statistically, an average may more closely represent the expected value. Whether weekly, monthly, or annual sales are used, more data points are assumed to be better estimators of the population parameter. Therefore, the number of periods of historical sales should be as large as possible to minimize the forecast error. The time period may be adjusted if a shift in historic sales has occurred. For the uninitiated, a rule of thumb would be to use the number of periods in one business cycle. Business cycles are not deterministic values; expansions and recessions occur at irregular intervals and last for varying periods that are dated according to when business activity changes direction (Romer, 2008). Business cycles have been as short as one year and as long as eight years (Romer, 2008). This uncertainty may cause anxiety for neophyte forecasters. One suggestion or convention may be to use about ten years of data to create the estimate. This would reflect at least one business cycle for the firm.

Expected sales would be determined by:

([Salese.sub.t+k]) = [[SIGMA].sup.n.sub.k][sales.sub.t] for all k>0

Thus, the average of the ten years of sales figures in our example is $264,281. This technique's major flaw is that outlier years can severely distort the average. Additionally, the average may not reflect economic conditions or potential growth in the firm.

Average Percentage Change in Sales

Average percentage change in sales has an intuitive appeal to users who feel more confident using percentages of growth. In this approach using a sample of ten years of data capture economic changes that affect the firm individually and the market in aggregate. The percentage change from one year to the next is calculated as:

% Change in Sales = Ending Sales / Beginning Sales -1

Average Percentage Change = [[SIGMA].sup.n.sub.1] % Change in Sales/N

This technique provides users with an estimate of an average growth that can be expected in the next year. In the example:

Average Percentage Change in Sales

Year   Sales          % change

2002   $240,000.00
2003   $255,000.00    0.06250
2004   $245,400.00   -0.03765
2005   $253,450.00    0.03280
2006   $240,000.00   -0.05307
2007   $260,550.00    0.08563
2008   $271,600.00    0.04241
2009   $297,180.00    0.09418
2010   $315,350.00    0.06114

Average Percentage Change 0.03559

New sales would be expected to be 3.559% higher than in 2010. Thus, the forecast for new sales would be = 315,350*1.035579 = $334,630.94. Note that this method loses a data point.

Present Value

Given a series of historical sales, an average compounded growth rate can be generated using the present value of a single sum. This method uses the first sales amount as the present value and the final sales amount is considered to be the future value. The number of periods is equal to (n-1) periods since the first sales figure represents time period zero. Solving for an interest rate using a financial calculator, a spreadsheet program, or using mathematics is a quick process. The interest rate calculated is a proxy for the periodic growth rate. The compounded

sales growth rate would be g = [(FV/PV).sup].1/n -- 1 for all n > 0. For the sample data the growth rate is

equal to[(315350/240000).sup.1/8] - 1 = g = 1.[31396.sup.125] -1 =.03472 =.03472= 3.742%. Therefore, new sales

would be predicted as $315,350 *(1.03742) = $326,298.84. The limitation of using this approach is that an endpoint that is positively or negatively skewed will provide a skewed growth rate that may not truly reflect the changes in sales.

ROE Method

The return on equity method estimates the amount of growth that a firm can maintain without changing its capital structure. Capital structure is the amount of debt and equity financing used by the firm (Higgins, 1977). See table 2 for the calculations of ROE and retention ratios. The ROE growth rate is calculated as

G = ROE * Retention Ratio

The growth rates are:

          ROE       RR        G         %

2002    0.5899    0.1152    0.0680     6.80%
2003    0.5601    0.1425    0.0798     7.98%
2004    0.4654   -0.0614   -0.0286    -2.86%
2005    0.4890   -0.0201   -0.0098    -0.98%
2006    0.3117    0.0204    0.0064     0.64%
2007   -0.5912    1.0000   -0.5912    -59.12%
2008   -3.8613    1.0000   -3.8613   -386.13%
2009    0.6118    1.0000    0.6118     61.18%
2010    0.4513    1.0000    0.4513     45.13%

The nine year average using the ROE method to project sales indicates that sales in 2011 would decline by 36.37%. Thus, sales for 2011 would be expected to be $315,350(1.0 .3637) = $200,657.21. As illustrated by this example, sales may decline over time. This method is preferred by practitioners for ease of use and relative accuracy. However, the use of this technique alone may not provide sufficient confidence for managerial decisions. There may be a tendency for managers to manipulate either the ROE or the dividend payout to inflate sales growth. The use of multiple methods may alleviate this concern.

Sustainable Growth Rate

The sustainable growth rate is the amount by which the assets of the firm can grow without changing its capital structure (see Higgins, 1977 for a full discussion of this issue). This is an alternative to the ROE method and is calculated as follows:

9 = ROE* (1- D)/l - [ROE* (l-D)]

The sustainable growth rates for Killing-M-Softly Exterminators are as follows:

       ROE        RR         ROE*RR

2002    0.5899     0.1152      0.0680
2003    0.5601     0.1425      0.0798
2004    0.4654    -0.0614     -0.0286
2005    0.4890    -0.0201     -0.0098
2006    0.3117     0.0204      0.0064
2007   -0.5912     1.0000     -0.5912
2008   -3.8613     1.0000     -3.8613
2009    0.6118     1.0000      0.6118
2010    0.4513     1.0000      0.4513

       1-(ROE*RR)SGR             SGR %

2002    0.9320     0.0729      7.2936%
2003    0.9202     0.0868      8.6764%
2004    1.0286    -0.0278     -2.7770%
2005    1.0098    -0.0098     -0.9753%
2006    0.9936     0.0064      0.6407%
2007    1.5912    -0.3715    -37.1540%
2008    4.8613    -0.7943    -79.4293%
2009    0.3882     1.5757    157.5661%
2010    0.5487     0.8224     82.2374%

The average of this nine year period would provide an estimated growth rate of 15.12%. Proforma sales for 2011, then, would be 315350 *(1.1512) = $ 363,030.42.

Simple One Variable Regression (Linear Trends)

Least squares regression fits the sales figures to a line in which the squared distance from the line is minimized. Predicting sales by this technique is limited by the sample size. Predicting sales with past sales, we get:

Sales = [alpha] + [beta][Years.sub.i] + [epsilon]

The equation of the line predicting sales is Sales = [alpha] + [beta][Years.sub.i] + [epsilon]. Using this method, sales for 2011 would be =8124 (10) + 221,661= $304,901. See table 4 for full regression results including p-values.

DISCUSSION AND CONCLUSIONS

Estimates of sales for the next year based on these methods ranged from $200,643.25 to $363,030.42. The best estimate for 2011 would be the average sales from all of the methods. Research indicates that using multiple methods provides greater accuracy in predicting future sales (Gallucci, 2007; Keung, 2011). For this example, the average new sales would be $291,830.95.

Although sales forecasting accuracy is deemed a crucial driver for future business decisions, methods suggested in recent literature involve complex models or statistical approaches. Many students have neither the access to these complex modeling programs nor the ability to use such systems. While the methods demonstrated here have limited accuracy, they do offer simple approaches to predict future sales growth. These provide a foundation for teaching sales forecasting. In turn, students will learn to develop proforma financial statements that will enable better decisions about production, marketing, and financing in business.

LIMITATIONS OF THE STUDY

Several limitations exist in this study. Projecting sales using the assumption that sales are flat does not reflect changes in the economy nor does this method incorporate growth by the firm. The average of sales and the average historical percentage change in sales methods are easily skewed by an outlier, while the present value method may be skewed by an endpoint that is very large or very small. Using a single variable regression in which sales are assumed to be a function of time does not account for seasonality of sales, market shocks or firm specific variables such as advertising, in leverage or operational efficiency. Further, the limited number of data points used in most sales regression models may not provide sufficient data points for fitting the least squares line estimate. A greater error is possible using this technique.

The ROE method and the sustained growth rate technique incorporate the projected retained earnings if the firm maintains its current capital structure. The leverage in the firm (amount of debt) is incorporated into these methods. These techniques are subject to manipulation by management because the calculation for ROE can be altered by what is used as net income and how total equity is defined.

RECOMMENDATONS FOR FUTURE RESEARCH

Future research would include expanding the study to survey the actual sales projection methods used by businesses and practitioners. In this way, teaching methods would focus on the techniques used. This would ensure that the material covered in classes is relevant for the current business climate and enhances the business education students receive.

This work should be extended to incorporate the development of the proforma financial statements. The combination of the sales projections and the creation of the projected statements would provide a basis for teaching students to analyze the financial health of the firm. This would create opportunities for students to practice critical thinking and decision making to determine whether to undertake capital budgeting projects, what capital structure should be maintained for the firm, and what would be the best ways to obtain needed financing. All of these areas of corporate finance are critical in efficiently and effectively managing a firm to maximize its value.

REFERENCES

Ando, T. (2008, September 2). Measuring the baseline sales and the promotion effect for incense products: a Bayesian state-space modeling approach. Annual Institute of Statistical Mathematics, 60, 763-780.

Cornett, M. M., Adair, Jr., T. A., & Nofsinger, J. (2012). Finance Applications and Theory. New York: McGraw-Hill Irwin.

Currie, C. S., & Rowley, I. T. (August 2010). Consumer behaviour and sales forecast accuracy: What's going on and how should revenue managers respond? Journal of Revenue & Pricing Management, 9(4), 74-76.

Dalrympe, D. J. (1987). Sales forecasting practices: resultw from a United States survey. International Journal of Forecasting, 3, 379-91.

Dalrymple, D. J. (December 1975). Sales forecasting methods and accuracy. Business Horizons, 18, 69-73.

Decker, R., & Gnibba-Yukawa, K. (2010). Journal of Product Innovation Management, 27, 115129.

Gallucci, J. A. (Fall 2007). How to Improve Forecasts with Hybrid Forecast Inputs. Journal of Business Forecasting, 1417.

Higgins, R. C. (1977). How much can a firm afford to grow? Financial Management, 6(3), 7-16.

Keung, E. C. (2011, November). Do suplementary sales forecasts increase the credibility of financial analysts' earnings forecasts? Accounting Review, 85(6), 2047-2074.

Kumar, M., & Patel, N. R. (February 2010). Using clustering to improve sales forecasts in retail merchandising. Annals of Operations Research, 174(1), 33-46.

McCarthy, T. D. (2006). The evolution of sales forecasting management: A 20-year longitudianl study of forecasting practices. Journal of Foreasting, 25, 303-324.

Mentzer, J. T., & Cox, J. E. (1984). Familiarity, application, and performance of sales forecasting technieuqes. Journal of Forecasting, 3(1), 27-36.

Mentzer, J. T., & Kahn, K. B. (1995). Forecasting technique, familiarity, satisfaction, usage, and applicaion. Journal of Forecasting, 14, 465-476.

Romer, C. D. (2008). Business Cycles. Retrieved October 17, 2011, from The Concise Encyclopedia of Economics: http://www.econlib.org/library/Enc/BusinessCycles.html

Siriram, R., & Snaddon, D. R. (May 2010). Forecasting new product sales. South African Journal of Industrial Engineering, 21(1), 123-135.

Tichenor, C., & Davis, B. (2009, November). Improving the Quality of Foreign Military Sales using Benford's Law. The Defense Institute of Security Assistance Management Journal, 184-189.

West, D. C. (1994). Number of Sales Forecast Methods and Marketing Management. Journal of Forecasting, 13, 395-407.

About the Authors:

Laurie A. Babin is a professor of marketing and holds the Abell Endowed Entrepreneurship Professorship in the College of Business Administration at the University of Louisiana at Monroe. She received her Ph.D. from Louisiana State University in 1992 and was at The University of Southern Mississippi for sixteen years before joining the faculty at the University of Louisiana at Monroe in 2007. Dr. Babin has published articles in several national and international journals and conference proceedings.

Melissa V. Melancon is an assistant professor of finance at ULM with a DBA in finance from Louisiana Tech University. She has CFM and CMA certifications and is currently pursuing a Charter Financial Analyst designation. Dr. Melancon holds the Chase Bank research fellowship. Current research interests include ethics, pedagogical research, teaching technologies, and corporate finance. She is developing new research in motivations for ethical decision making. Dr. Melancon has published articles in national journals and conference proceedings.

Melissa V. Melancon

Laurie A. Babin

University of Louisiana at Monroe

Table 1

Killing-M-Softly Exterminator Company Income Statements

Income
statement                 2002           2003           2004

Sales              $240,000.00     $255,000.0     $245,400.0
COGS               $162,000.00    $172,200.00    $174,240.00
Gross Profit        $78,000.00     $82,800.00     $71,160.00
  Depreciation
Expense             $10,000.00     $10,000.00     $10,000.00
S&A Expense         $13,440.00     $15,300.00     $14,969.40
EBIT                $54,560.00     $57,500.00     $46,190.60
Interest Expense     $1,815.33      $3,075.64      $2,222.27
EBT                 $52,744.68     $54,424.36     $43,968.33
Taxes               $21,097.87     $21,769.74     $17,587.33
Net Income          $31,646.81     $32,654.61     $26,381.00
Dividends Paid      $28,000.00     $28,000.00     $28,000.00
Add to R/E           $3,646.81      $4,654.61     -$1,619.00

Income
statement                 2005           2006           2007

Sales              $253,450.00    $260,550.00    $240,000.00
COGS               $180,115.00    $183,335.00    $267,000.00
Gross Profit        $73,335.00     $77,215.00    -$27,000.00
  Depreciation
Expense             $10,000.00     $10,000.00     $15,000.00
S&A Expense         $15,967.35     $17,717.40     $16,320.00
EBIT                $47,367.65     $49,497.60    -$58,320.00
Interest Expense     $1,622.35      $1,857.85     -$1,527.91
EBT                 $45,745.30     $47,639.75    -$56,792.09
Taxes               $18,298.12     $19,055.90    -$22,716.83
Net Income          $27,447.18     $28,583.85    -$34,075.25
Dividends Paid      $28,000.00     $28,000.00          $0.00
Add to R/E            -$552.82        $583.85    -$34,075.25

Income
statement                 2008           2009           2010

Sales              $271,600.00    $297,180.00    $314,350.00
COGS               $310,312.00    $223,045.20    $229,667.00
Gross Profit       -$38,712.00     $74,134.80     $84,683.00
  Depreciation
Expense             $15,000.00     $15,000.00     $15,000.00
S&A Expense         $18,468.80     $20,505.42     $22,004.50
EBIT               -$72,180.80     $38,629.38     $47,678.50
Interest Expense     $4,121.92      $7,492.85      $5,821.73
EBT                -$76,302.72     $31,136.53     $41,856.77
Taxes              -$30,521.09     $12,454.61     $16,742.71
Net Income         -$45,781.63     $18,681.92     $25,114.06
Dividends Paid           $0.00          $0.00          $0.00
Add to R/E         -$45,781.63     $18,681.92     $25,114.06

Table 2

Net Income, Return on Equity

                        2002      2003       2004

ROE = NI/TE           0.5899    0.5601     0.4654
Div. Payout Ratio =    0.885     0.857      1.061
  Div. /NI
Retention Ratio =     0.1152    0.1425    -0.0614
  (1-D)

                         2005      2006      2007

ROE = NI/TE            0.4890    0.3117   -0.5912
Div. Payout Ratio =     1.020     0.980    0.0000
  Div. /NI
Retention Ratio =     -0.0201    0.0204    1.0000
  (1-D)

                         2008      2009      2010

ROE = NI/TE           -3.8613    0.6118    0.4513
Div. Payout Ratio =    0.0000    0.0000    0.0000
  Div. /NI
Retention Ratio =      1.0000    1.0000    1.0000
  (1-D)

Table 3

Killing-M-Softly Exterminator Company Balance Statements

Balance Sheet                      2002           2003           2004

Cash                          $5,520.00      $6,120.00      $5,644.20
Acct Receivables             $21,041.10     $21,657.53     $20,842.19
Inventory                    $13,315.07     $13,681.64     $14,321.10
Total Current Assets         $39,876.16     $41,459.18     $40,807.49

Gross Fixed Assets          $100,000.00    $100,000.00    $100,000.00
Accumulated Depreciation     $10,000.00     $20,000.00     $30,000.00
Net Fixed Assets             $90,000.00     $80,000.00     $70,000.00
Total Assets                $129,876.16    $121,459.18    $110,807.49

Acct Payables                $15,090.41     $16,040.55     $16,230.58
Accruals                     $11,400.00     $11,985.00     $12,760.80
Notes Payable                $29,738.95     $17,132.21      $9,133.70
Total Current Liabilities    $56,229.36     $45,157.76     $38,125.07

Long Term Debt               $20,000.00     $18,000.00     $16,000.00
Total Liabilities            $76,229.36     $63,157.76     $54,125.07

Equity
Common Stock                 $50,000.00     $50,000.00     $50,000.00
Retained Earnings             $3,646.81      $8,301.42      $6,682.42
Total Equity                 $53,646.81     $58,301.42     $56,682.42

Total Liabilities           $129,876.16    $121,459.18    $110,807.49
  and Equity

Balance Sheet                      2005           2006           2007

Cash                          $5,829.35      $6,513.75      $4,080.00
Acct Receivables             $22,914.66     $26,411.92     $24,986.30
Inventory                    $16,284.37     $17,077.78     $27,065.75
Total Current Assets         $45,028.38     $50,003.45     $56,132.05

Gross Fixed Assets          $100,000.00    $100,000.00    $150,000.00
Accumulated Depreciation     $40,000.00     $50,000.00     $65,000.00
Net Fixed Assets             $60,000.00     $50,000.00     $85,000.00
Total Assets                $105,028.38    $100,003.45    $141,132.05

Acct Payables                $16,777.84     $15,570.92     $24,871.23
Accruals                     $13,432.85      $7,165.13      $6,600.00
Notes Payable                 $4,688.10    -$34,446.04     $32,022.63
Total Current Liabilities    $34,898.78    -$11,710.00     $63,493.86

Long Term Debt               $14,000.00     $20,000.00     $20,000.00
Total Liabilities            $48,898.78      $8,290.00     $83,493.86

Equity
Common Stock                 $50,000.00     $85,000.00     $85,000.00
Retained Earnings             $6,129.59      $6,713.45    -$27,361.80
Total Equity                 $56,129.59     $91,713.45     $57,638.20

Total Liabilities           $105,028.38    $100,003.45    $141,132.05
  and Equity

Balance Sheet                      2008           2009           2010

Cash                          $6,246.80     $12,778.74     $13,517.05
Acct Receivables             $30,508.49     $37,452.82     $37,894.25
Inventory                    $33,156.62     $26,276.56     $30,202.78
Total Current Assets         $69,911.92     $76,508.12     $81,614.08

Gross Fixed Assets          $150,000.00    $150,000.00    $150,000.00
Accumulated Depreciation     $80,000.00     $95,000.00    $110,000.00
Net Fixed Assets             $70,000.00     $55,000.00     $40,000.00
Total Assets                $139,911.92    $131,508.12    $121,614.08

Acct Payables                $28,905.78     $20,776.81     $21,393.64
Accruals                      $7,469.00      $8,172.45      $8,644.63
Notes Payable                $71,680.58     $52,020.38     $15,923.28
Total Current Liabilities   $108,055.36     $80,969.64     $45,961.54

Long Term Debt               $20,000.00     $20,000.00     $20,000.00
Total Liabilities           $128,055.36    $100,969.64     $65,961.54

Equity
Common Stock                 $85,000.00     $85,000.00     $85,000.00
Retained Earnings           -$73,143.44    -$54,461.52    -$29,347.46
Total Equity                 $11,856.56     $30,538.48     $55,652.54

Total Liabilities           $139,911.92    $131,508.12    $121,614.08
  and Equity

Table 4

Killing-M-Softly Regression Results

SUMMARY OUTPUT

Regression Statistics

Multiple R          0.848759598
R Square            0.720392855
Adjusted R Square   0.680448977
Standard Error      14817.87593
Observations        9
ANOVA
                    df              SS                MS
Regression          1               3959962560        3959962560
Residual            7               1536986129        219569447
Total               8               5496948689

                    Coefficients    Standard Error    t Stat
Intercept           223661.1111     10764.93729       20.77681506
Year                8124            1912.979556       4.246778264

Sales Forecast      $ 304,901.11
  for 2011

SUMMARY OUTPUT

Regression Statistics

Multiple R          0.848759598
R Square            0.720392855
Adjusted R Square   0.680448977
Standard Error      14817.87593
Observations        9
ANOVA
                    df              F                 Significance F
Regression          1               18.03513          0.003808
Residual            7
Total               8

                    Coefficients    P-value           Lower 95%
Intercept           223661.1111     1.5E-07           198206.1
Year                8124            0.003808          3600.522

Sales Forecast      $ 304,901.11
  for 2011

SUMMARY OUTPUT

Regression Statistics

Multiple R          0.848759598
R Square            0.720392855
Adjusted R Square   0.680448977
Standard Error      14817.87593
Observations        9
ANOVA
                    df
Regression          1
Residual            7
Total               8

                    Coefficients    Upper 95%         Lower 95.0%
Intercept           223661.1111     249116.143        198206.1
Year                8124            12647.4779        3600.522

Sales Forecast      $ 304,901.11
  for 2011

SUMMARY OUTPUT

Regression Statistics

Multiple R          0.848759598
R Square            0.720392855
Adjusted R Square   0.680448977
Standard Error      14817.87593
Observations        9
ANOVA
                    df
Regression          1
Residual            7
Total               8

                    Coefficients    Upper 95.0%
Intercept           223661.1111     249116.1
Year                8124            12647.48

Sales Forecast      $ 304,901.11
  for 2011