Michael Feldman
Prof. Wyner
STAT 6130
12/17/24
Part 1: Executive Summary
My analysis identifies the key drivers of rental costs for a 50,000 sqft office space and
provides clear cost estimates to help guide your decision-making. I’ve evaluated the most
impactful factors—lease length, location, renovation age, occupancy, and amenities—and
identified significant savings opportunities while maintaining flexibility and quality.
Baseline Scenario:
To establish a baseline, I assumed a 3-year lease for 50,000 sqft of newly renovated,
modern-wired space in the new suburb closest to the airport with 100% occupancy. Under
these conditions, the estimated total annual rent is $3.4 million, or $68 per square foot. The
prediction range is estimated to be between $3.1 million and $3.7 million.
Major Cost Drivers: Premiums and Discounts:
From there, I examined how the major cost drivers impact rent, starting with lease
length. Longer leases provide measurable savings: extending the lease to 6 years reduces costs
by approximately $2 per square foot (about $100,000 annually), while a 10-year lease delivers
a discount of $4 per square foot (roughly $200,000 annually).
Location plays a critical role, with city center properties carrying a significant premium
of $6 per square foot (around $300,000 annually) compared to the baseline. On the other hand,
leasing in older suburban areas offers a consistent discount of $2 per square foot (or $100,000
annually), making it a cost-effective option without sacrificing space.
The age of renovations also factors into rent. While newly renovated spaces are included
in the baseline, properties renovated 4 or more years ago reduce costs by about $1 per square
foot, translating to $50,000 annually for 50,000 sqft. Similarly, occupancy rates affect pricing:
spaces at 80% occupancy can offer a discount of $2 per square foot, or roughly $100,000
annually, though lower occupancy may reflect less competitive properties. I also looked at the
impact of amenities. Modern wiring is already included in the baseline, as it’s essential for
operational needs. However, buildings with additional features like on-site fitness centers and
restaurants command a premium of up to $3 per square foot ($150,000 annually).
Adjusted Scenario:
To demonstrate how these factors play together, I adjusted the baseline scenario. For 50,000
sqft in the center of the new suburb, assuming:
• Renovations completed 4 years ago, 80% occupancy, Modern wiring included
• The estimated total annual rent drops to $3.1 million, or $62 per square foot. The
prediction range was estimated to be between $2.8 million and $3.4 million.
Bargains and Savings Opportunities + Premiums on New Properties + Full Amenities:
In terms of savings, my analysis highlights several clear opportunities. Leasing in older
suburbs consistently trims $2 per square foot off the cost, saving $100,000 annually.
Extending the lease to 6 or 10 years delivers meaningful long-term savings of up to $200,000
annually. On the other hand, new properties with full amenities come with a premium of $3 per
square foot, or about $150,000 annually (as mentioned), but they offer potential benefits such
as higher employee satisfaction and improved tenant experience.
By prioritizing lease terms, location, and property features, you can achieve significant
cost savings while securing a high-quality, flexible office space that meets your team’s needs.
Michael Feldman
Prof. Wyner
STAT 6130
12/17/24
Part 2: Technical Summary
My analysis aimed to estimate rental costs for a 50,000 sqft office space and identify the
key factors influencing rent. I developed a multiple linear regression model to produce clear,
actionable cost estimates while ensuring statistical rigor through careful data preparation, model
specification, and robust diagnostics. This summary explains the fitted model term by term, the
values used in the executive summary, and the reasoning behind excluding minor contributors
with negligible effects.
Data Cleaning and Outlier Removal
Prior to building the model, I carefully reviewed the dataset of 225 rental properties to ensure its
quality. I excluded three observations due to their extreme or invalid values, as they distorted the
analysis:
1. A lease with 197.16 sqft was excluded as it was an unrealistically small outlier for
commercial office space.
2. A lease with negative total rent was removed as it was invalid.
3. A lease with an unusually high RentTotal of $60.6 million was excluded as an extreme
outlier that significantly skewed model performance.
After these removals, the dataset contained 222 observations, providing a clean and reliable
foundation for the analysis.
Model Specification
The regression model was specified as: RentTotal=β0+β1(SqftLease)+β2(LeaseLen)+β3
(Age)+β4(Renovation)+β5(Location_SUBNEW)+β6(Location_SUBOLD)+β7
(Wiring_YES)+β8(Restaraunt_YES)+β9(Exercise_YES)+ ϵ.
The model was trained on 80% of the data and validated on the remaining 20%. Below are the
key coefficients and their interpretations, rounded for clarity:
Fitted Model Results
The following are the key coefficients, rounded for clarity in the executive summary:
1. Intercept (β0): $400,000
2. Square Footage (SqftLease): 66.7 per sqft
o As expected, square footage is the primary driver of rent. For a 50,000 sqft space,
this translates to $3.4 million annually at baseline.
3. Lease Length (LeaseLen):
o Longer leases offer meaningful cost savings:
§ A 6-year lease reduces rent by approximately $2 per sqft.
§ A 10-year lease offers a discount of about $4 per sqft.
4. Location:
o Compared to the city center:
§ New Suburb (Location_SUBNEW) reduces rent by $172,857 annually
(~$3 per sqft).
§ Old Suburb (Location_SUBOLD) offers a similar discount of $178,911
annually.
Michael Feldman
Prof. Wyner
STAT 6130
12/17/24
5. Renovation:
o Newly renovated properties command a slight premium, while properties
renovated 4+ years ago provide a discount of approximately $1 per sqft.
6. Occupancy:
o Lower occupancy (e.g., 80% occupancy) results in a discount of about $2 per
sqft, which translates to $100,000 annually for 50,000 sqft.
7. Amenities:
o Modern wiring (minor impact) adds a premium of $2,834 annually
o On-site fitness centers (Exercise_YES) add a premium of approximately $2 per
sqft (~$100,000 annually).
o Buildings with restaurants (Restaraunt_YES) add a more modest premium of $1
per sqft (~$63,000 annually).
Neglected Factors
While some predictors were statistically significant, their contributions to rent were too small to
be meaningful for a 50,000 sqft property. These factors were excluded from the executive
summary for simplicity:
o Building Age (Age): Coefficient values were close to zero, with a marginal impact of
less than $0.03 per sqft.
o Renovation beyond 4 years: While included for reference, its impact was small relative
to other factors.
Model Diagnostics
To ensure the robustness of the model, I performed a series of diagnostics:
1. R² and Model Fit:
o The model achieved an R2R^2 value of 0.89, meaning it explains 89% of the
variability in RentTotal. This reflects a strong model fit.
o The Actual vs. Predicted RentTotal plot confirms strong predictive performance
across almost all of the data range. Most points align closely with the perfect
prediction line
2. Residual Analysis:
o The Residuals vs. Predicted plot showed no clear patterns, indicating
homoscedasticity and linearity assumptions were met.
o A Histogram of residuals and a normal Q-Q plot confirmed that residuals were
approximately normally distributed
NOTE: See all graphs and charts at bottom of page 3
Values Used in the Executive Summary
1. Baseline Scenario: A 3-year lease of 50,000 sqft of newly renovated, modern-wired
space in the new suburb with 100% occupancy produced an estimated rent of $3.4
million annually, or $68 per sqft.
2. Adjusted Scenario: For the same space with 4 years since renovation and 80%
occupancy, the estimated rent is $3.1 million annually, or $62 per sqft.
3. All premiums and discounts were expressed per square foot and rounded to the nearest
dollar for simplicity.
Michael Feldman
Prof. Wyner
STAT 6130
12/17/24
Conclusion
This analysis provides a reliable and interpretable model for estimating rent costs for a
50,000 sqft office space. By focusing on major drivers such as square footage, lease length,
location, and amenities, I identified key premiums, discounts, and opportunities for cost savings.
Minor factors like building age were excluded due to their negligible impact. The model
diagnostics confirm its validity, and rounding ensures the results are both clear and practical for
decision-making.