Yes. I’ve attached two financial models that I created this past semester.
A ground-up mixed-use multifamily deal modeled from construction through stabilization and exit. I sized the construction loan with a C-PACE tranche included, then took it out with permanent debt at the refi. NOI builds off the unit mix, parking, and ground-floor retail. The model returns levered and unlevered IRR, equity multiple, and a three-tier GP/LP promote that flexes through each hurdle. Rents and vacancy benchmarked against CoStar comps, and the levered return is stress-tested against construction cost and the permanent loan rate.
A two-asset project-finance model for a renewable portfolio in the USVI. I sized the senior debt to DSCR with a back-levered structure and monetized the ITC through a Section 6418 transferability sale. The model also includes a downside case on ITC pricing.
One week from the offer.
Intermediate and climbing quickly. I’m comfortable in Excel, but I reach for Python (pandas) when the data work gets heavier. Python lets me build API integrations, work with larger datasets, and iterate on complex models faster than a single workbook allows. Excel becomes the wrong tool for cleaning and normalizing large macro data for prediction modeling.
Yes. The high intensity, long days, and chance to grow are the main reasons I am applying. That was not meant to be sycophantic, but I believe I thrive in environments with a lot of entropy, making decisions with incomplete information and embracing spontaneity. In demanding settings I have personally experienced asymmetrical growth, and the impact of my work has grown right along with it. That is exactly the draft I want to sign up for when I apply to a boutique firm with a lean team doing big things.
Debt Yield · Leverage / Recovery Test
NOI ÷ Loan Amount
Debt yield matters most in a risky, volatile market, because it answers one big question: if this deal goes bad, how fast do I recover my money? It is the lender’s cash-on-cash return on day one, and because it ignores financing terms entirely, it’s the cleanest read on real leverage at origination.
DSCR · Serviceability Test
NOI ÷ Annual Debt Service
This answers the primary question lenders have: can the borrower make the payments, and in case of a shortfall or a slow quarter, how much cushion do I have before coverage breaks?
The key distinction is that DSCR moves with the loan terms and debt yield does not. The same property and loan can show a different DSCR just by cutting the rate, stretching amortization, or going interest-only. Additionally, LTV is just as soft, since it leans on an appraised value driven by cap rates.
Debt yield strips all of that out and moves only with NOI and loan size, so it cannot be gamed by cheap financing or aggressive valuations (subscription lines or forward commitments). When rates are high, debt service rises and DSCR becomes the binding constraint: it sizes the loan down, which in turn lifts debt yield so it clears comfortably. When rates are low, DSCR passes easily and debt yield becomes the constraint that limits proceeds.
Internal Rate of Return (IRR) is the discount rate that sets a deal’s NPV to zero. It’s the time value of money over the holding period, signaling how efficiently capital is compounding. The levers that move it are refinances and a quicker return of capital, but it also assumes that you can reinvest distributions at that same rate (which rarely holds). I have always thought of IRR as a metric of how expensive this future stream of income is. On the other hand, Equity Multiple (EM) or TVPI is the magnitude of wealth created — just cash distributions over equity invested, irrespective of the time it took to generate that wealth. IRR is the speed; EM is the size.
A Personal View
I have an aversion to markets / companies / funds hyper-obsessed with IRR. When all the value lies at exit, the GP/LP playbook for close-end RE funds quietly rewards optimizing for the IRR number. On long-term ground-up bets I’d argue great developers should optimize for EM rather than IRR. I guess I am arguing for long-term wealth generation over short holds. I also think that’s why people are moving to DPI or MIRR as a metric over IRR. Open to arguments on opportunity costs with an EM-focused approach!
Mostly shortcuts. I’ve been progressively weaning myself off the mouse. Not fully there yet, but close, and it’s a deliberate habit I keep working at.
I’ve built PPMs and OMs academically (in mock case studies and semester projects) but not sell-side documents in a commercial setting. It is near the top of what I want to learn here, and given the modeling and writing I already do, I’d expect to learn fast.
I am based in Berkeley now, but I’d relocate to Miami for the internship. Being physically with the team is worth a lot early in a career; you absorb an enormous amount by osmosis — in overhearing conversations, and chats by the water cooler. However, if I work remotely or hybrid, I am pretty comfortable with a setup like that as well. I have always loved being the “headphones on, crunching Excel late into the night” analyst. For the first few years of my career I really want to be quantitatively bent, while I build relationships in the industry to really thrive.
One of my mentors liked Emile’s LinkedIn post, and it caught my eye.
The sports I compete in are tennis and triathlon (since I am in the club team at school). I have been an avid skier for most of my life, and being near Tahoe, I have been able to ski into late spring. Recreationally, I also play golf and climb (bouldering and top-rope).
~ $27 / hour.