Countrywide Business Proposal

Running Head: COUNTRYWIDE BUSINESS PROPOSAL

Countrywide Financial Corporation Business Proposal

Countrywide Financial Corporation Business Proposal

News reports state that rising mortgage defaults and delinquencies have shut down the subprime mortgage markets (Lubby, 2008). Companies are facing huge losses because of the housing boom that has gone sour. One financial company that has been in the news more often than others is Countrywide Financial Corporation (CFC). CNNMoney.com reported on March 13, 2008 that while delinquency rates fell from 7.47% in January 2008 to 7.44% in February, foreclosures continue to rise ( This paper will focus on the financial problems facing Countrywide and recommendations for actions the lender should take to improve its financial status.

Countrywide’s foreclosure rates have increased due to loan defaults by consumers. “The nation’s largest residential mortgage lender, with a market share of 17%, said its foreclosure rate almost doubled in January from the same month a year ago” (Muolo, 2008). The dependent variable in this situation is Countrywide’s profits and market share. The company’s profit is declining along with their market share with the increase in defaults. This has also affected their workforce, forcing the company to develop department restructuring programs that will eventually lead to layoffs. The independent variable is the number of foreclosures occurring as homeowners face financial difficulty in paying their monthly mortgages.

Some analysts blame the subprime mortgages for the rise in foreclosures. Lenders offer subprime mortgages to high-risk borrowers. Generally, subprime borrowers pay 200 to 300 basis points above the prevailing prime rate given to borrowers who meet a minimum FICO score (credit score) and a sufficient down payment (Agarwal & Ho, 2007). FICO is a registered trademark of Fair Isaac Corporation. Countrywide is one of the largest mortgage lenders in the United States. As such, the rate of subprime mortgages is higher, which means their present financial status is dependent on the performance of the outstanding loans. Research reveals that the rate of foreclosures for Countrywide is higher than other lenders (Countrywide’s, 2008). This statement can be proven or disproved by hypothesis testing by taking a sample of mortgages from Countrywide and other mortgages combined. There are five steps to take in hypothesis testing. The first is stating the null and alternative hypotheses.

The null hypothesis states the value of a population parameter. For this test, the null hypothesis (H0) would be that the population mean of Countrywide foreclosures is greater than the mean of all other foreclosures. The alternative would be that the mean of Countrywide’s foreclosures is less than the mean of other lenders’ foreclosures and expressed in statistical and mathematical terms as follows:

H0: Ој1 > Ој2

H0: Ој1 < Ој2

The second step is to determine the level of significance.

The level of significance, which is the probability of rejecting the null hypothesis, would be .05, the normal level for consumer research projects (Lind, et al, 2004). The next step would then be to select the test statistic, a value used to determine whether to reject the null hypothesis. The sample information determines the test statistic. Large samples (greater than 30) use the z test statistic. The sample distribution of the sample mean determines the z value. The fourth step is to formulate the decision rule. The rule states the conditions under which one rejects the null hypothesis or not. Figure 1 demonstrates the region of rejection (Lind, et al, 2004). The critical value is the point that divides the point rejection and non-rejection of the null hypothesis (Lind, et al, 2004).

Figure 1

Source: Statistical Techniques in Business and Economics, 12th edition.

The final step of the hypothesis test is to compute the value of z and compare to the critical value. The comparison will be the basis of rejecting or not rejecting the null hypothesis. The formula for computing the test statistic (value of z) is:

Countrywide can use the results of the hypothesis testing as a means to develop strategies to combat the foreclosure rate, considering programs other lenders have implemented.

Countrywide is the nation’s number one home mortgage lender, approving three out of every four loans. This success rate has now turned into a downfall for the company. When times were good in the housing and mortgage sector, Countrywide was making significant profits. With sub-prime lending virtually exploding before our eyes, we are seeing the housing market tumble. In the fourth quarter of 2007, Countrywide lost 422 million dollars and announced that one third of their subprime mortgages were delinquent. On March 11, 2008, Countrywide announced a 16 billion dollar plan to restructure or refinance mortgages. The company expects the plan to affect over 82,000 loans (Lubby, 2008). Other lenders use programs such as forbearance mortgages to help homeowners avoid foreclosure. The hypothesis test will use a sample size greater than 30, so a z test will be preferred over a t test. A t test would not provide a large enough sample because the number of loans is so high. It is extremely hard to calculate the best sample size because of the different variables involved, coupled with the fact that the population standard deviation is also unknown. Based on the team’s research, there are more than 100,000 subprime loans, using a 95% confidence level, and +/- 3% error, the test will need to sample no less than 1,058 sub-prime loans based on charts used to calculate sample sizes found on Zoomerang website. This number may appear to be a small number, but when the number of customers/loans is over 10,000 the additional number of respondents grows rather slowly.

To select the sample we used a 95% confidence interval and 5% significance level of data used, that was collected from three years of previous foreclosure filings in all states. After the data was compiled Countrywide selected a sample size that would fairly represent