1 of 10

HOME SALES PRICING �AND �INVESTING IN�MORTGAGE BACKED SECURITIES

BY KAYDIA CASTANEDA

2 of 10

PROJECT GOAL

  • Are sales prices significantly impacted by either of the following:
    • If the home was remodeled
    • If the home was built between 2000 and 2010

3 of 10

HYPOTHESES

​

  • Null Hypothesis 2 (H0):

Homes that were built between 2000 and 2010 have no effect to sales price of a home.

  • Alternate Hypothesis 2 (Ha):

Homes that were built between 2000 and 2010 do impact sales price of a home.

​

  • Null Hypothesis 1 (H0):

Home remodeling has no effect on sales price of a home.

  • Alternate Hypothesis 1 (Ha):

Home remodeling does have an impact on sales price of a home.

4 of 10

DATA OVERVIEW

  • Sales prices of homes are normally distributed.

DESCRIPTIVE STAT OF SALES PRICES

​

​

Mean

$180k

Standard Error

2079.105324

Median

$163k

Mode

$140k

Standard Deviation

79442.50288

Sample Variance

6311111264

Kurtosis

6.53628186

Skewness

1.88287576

Range

$720k

Minimum

$35k

Maximum

$755k

Sum

$264,144,946

Count

1460

Data sourced from Kaggle.com

5 of 10

METHODS

  • Created a new column to categorize homes as remodeled or not remodeled using 1’s and 0’s so that the pivot table could process the data.
  • Analyzed housing prices by decades to find if any specific decade was an outlier. Based on initial analysis, it wa decided to compared 2000 - 2010 to all years before 2000.
  • Created histogram of sales prices to ensure data showed normal distribution.
  • Conduct statistical analyses using “two sample assuming unequal variable” - T-test.
    • In the test for homes remodeled, the control group was comprised of homes not remodeled
    • In the test for the year homes were built, the control group was comprised of homes built before 2000.
  • Results were graphed for visualization

6 of 10

ANALYSIS OF REMODELED HOMES

  • Null Hypothesis 1 (H0):

Home remodeling has no effect on sales price of a home.

​

  • The null hypothesis is not rejected

​

7 of 10

ANALYSIS OF YEAR BUILT

  • Null Hypothesis 2 (H0):

Homes that were built between 2000 and 2010 have no effect to sales price of a home

​

  • The null hypothesis is rejected

​

8 of 10

RESULTS

  • Hypothesis 1 Conclusion:
    • There is not evidence of a statistically significant difference in the average sales prices for houses that were remodeled and houses that were not remodeled.

​

  • Hypothesis 2 Conclusion: There is evidence of a statistically significant difference in the average sale prices for houses that were built between 2000 and 2010 and those homes that were built before the year 2000.

​

  • The statistical analyses showed that the sales price of having a home that was built between 2000 and 2010 will sell between 75k and 93k more, at a 95% confidence rate.

9 of 10

FINAL MESSAGE AND RECOMMENDATION

​

Recommendation:

Invest in mortgages of homes built before 2000 because there is evidence of a statistical difference in sales price from the homes that were built between 2000 and 2010.

​

    • Homes built before 2000 have lower sales prices and have a higher chance of having a higher portion of the mortgage paid off.
    • Allows investor to invest/purchase more mortgages because of lower costs and high chance of return if homeowner defaults – risk could be potentially lower.

​

​

​

For reference, analysis workbook here

10 of 10

THANK YOU FOR YOUR TIME

Questions?