Real Estate Development Modeling

yayaa's picture
Rank: Gorilla | banana points 600

Hello Everyone,

I am curious to know how everyone underwrites and models a multi-family(mixed use with retail) development project. For example, if you or your firm is looking to acquire a current income producing project(future development site) and has plans on taking it down in the future, how would you value the property today and moving forward? When doing pro-formas for the development project you will be constructing in 2 years from when it is bought, there are many factors to look into such as construction costs, financing etc. I am aware of software such as Argus and Yardi which assist in the feasibility of the project. Is having software a must for development since it is the most specific and hands on project there is which involve many variables rather than an income producing property?

Bottom line, if you are acquiring a future development site, what steps do you think top development companies that are building large or even smaller scale multi-family projects take in achieving forecasting even though the future is not certain. What are the most important metrics in determining yield on the project. If anyone has any input or experience it would be highly appreciated. Also, any suggestions to what software or development model to use which is a good starting point to evaluate a development site. Thanks.

Private Equity Interview Course

  • 2,447 questions across 203 private equity funds. Crowdsourced from over 500,000 mem.
  • 9 Detailed LBO Modeling Tests and 15+ hours of video solutions.
  • Trusted by over 1,000 aspiring private equity professionals just like you.

Comments (45)

Best Response
Aug 15, 2017

Generally speaking, developers traditionally look at NOI yields (stabilized NOI/total project costs) to determine if there is a sufficient spread between that return and exit cap rates. If spreads are sufficient to justify the risk profile (maybe 150-200 bps) a lot of developers will pull the trigger.

That picture is getting increasingly complex with interest rate volatility in the market and rising construction costs. Now, some developers may accept a lower NOI yield or spread over exit caps if they just like the market fundamentals. For example, in many West Coast markets, construction costs are rising quickly and projected to continue rising given the skilled labor shortage. That puts downward pressure on development pipelines, creating a potential supply problem, which paints a rosier picture for market rent growth. Even if a project misses the mark and the yields are tight, rising construction costs and lengthier entitlement periods are making it more and more likely that you will not lose money if you build sooner, since replacement cost is ticking up with no clear reversal in sight. You may not capture the risk adjusted return you are seeking, but you are not going to be underwater like a 2009 mortgagee in the Florida panhandle.

    • 8
Aug 15, 2017

Going off of this. I built a dev model from scratch on excel and it is broken down into two "phases".

1) Development phase - this is a cash flow projection from the time you acquire site to the time you complete construction (delivery). During this phase, you are calculating your monthly costs, which then feeds into your monthly construction draws, which then feeds to your debt service (construction loan interest). Obviously there is no positive cash flow during this phase. Some people spread out there costs evenly during this phase. Some people believe hard costs occur in a "bell curve", there are a lot of soft costs to be drawn up front, etc. The distribution of your costs is very important because it dictates your construction interest.

2) Operation - This begins the day after delivery. You can begin receiving cash day one if you assume pre-leased units (I usually assume 15%), followed by your lease-up period which is derived from your leasing velocity (I usually plugged in 7.5% a month). Your expenses should kick in day one, and you are most likley still paying interest-only on your construction loan. It will be a while until you breakeven and start having positive cash flow. Depending on your mezz/equity, you - as the GP - will prob not receive a dime of operating cash distributions until you refinance or sell.

    • 4
Aug 17, 2017

Pretty much correct, but we don't see many 200 bps spreads anymore.

We pretty much look for a 9% yield w/ at least a 100 bps spread and an 85-90% P/E.

I come from down in the Valley, where Mr. when you're young, they bring you up to do like your daddy done.

Aug 18, 2017

We pretty much look for a 9% yield

What product type? We're generally content with a 7. Company I worked for last year was happy with a 6.5

    • 1
Aug 20, 2017

could you elaborate a bit more on the spread between return and exit cap rate? is Projection NOI/ actual total project costs(doesn't count in financing) VS Predicted CAP rate ???

Learn More

Side-by-side comparison of top modeling training courses + exclusive discount through WSO here.

Aug 21, 2017

Sorry - a couple amateur questions - why are they looking at the spread between NOI yield and exit cap rates? Is that essentially just the PF NOI yield vs. the transaction environment.

Is the interest rate impact you're referring to just the cost of debt financing, and exit multiple impact? Or am I missing something else?

Aug 21, 2017

Because if you can't sell it for more than you paid to build it, you're doing something wrong. The spread between stabilized yield on cost and exit cap is just this--the "cushion" the developer has between each dollar he pays to generate NOI and the dollar amount the market will pay for that same NOI stream. Make sense?

Aug 15, 2017

The one thing I rarely hear about is Argus Developer. Does anyone use Argus Developer for their development modeling?

Aug 15, 2017

Exactly... no one specifically mentions it.

Aug 15, 2017

Haha, so basically, no?

Aug 18, 2017

Nope. MF modeling might be the easiest there is. No idea why you'd pay for a program like that for apartments.

Aug 15, 2017

Development for MF is tough though. This isn't a typical modeling of existing units. It's putting together all the variables and cash flows etc.

The program can simplify and streamline the process with simple inputs leaving less margin for error.

Aug 18, 2017

We tested it for our MF modeling and decided against it. There wasn't a way to play with JV structures or Opex assumptions very easily. We stuck with excel and it's flexibility.

Aug 21, 2017

Yes our land development group does

Sep 19, 2017

Nobody uses ARGUS Developer in the US , some use in UK / ASIA. It used to be called Circle Developer but then ARGUS bought it. No idea what ARGUS plans on doing with Developer as I don't think it has any relation to ARGUS Enterprise (could be wrong) which continues to vomit itself across the institutional marketplace.

Learn More

Side-by-side comparison of top modeling training courses + exclusive discount through WSO here.

Aug 15, 2017

[Double post]

Aug 15, 2017

The one development model I have worked with looked a lot like an lbo model. It broke down equity ownership and the debt they'd need - including bridge financing, had a sources and uses table, dcf, sensitivity analysis, a cash flow waterfall, and a final IRR calculation based on a predicted sell date. No Argus was involved.

    • 1
Aug 21, 2017

Argus for modeling in the operations component of the project. Feed your NOI and TI/LC numbers through to your model to produce cash flow. Highly suggest reviewing GetREFM's development modeling course as this isn't something one can just easily explain over a blog. Much easier to watch videos and practice until it's all clear.

Sep 20, 2017

This really depends on your investment strategy as a company, and for a specific project. I can tell you what I personally look for, which is probably different from what the more institutional investors on this thread focus on.

We always like to have multiple exit strategies that give us optionality under different future market conditions. For most projects we are both a significant equity contributor and also the sponsor, so we underwrite conservatively (which means we do relatively few deals) and try to structure deals so we have the choice to invest more or divest at multiple milestones in development. Generally speaking we like to keep equity in a deal forever, regardless of scale, so we aren't purely fixated on exit caps. Cash flow is very much king for us, and on top of that we don't like deals where we aren't likely to realize our promote. Where does all this lead? We like difficult, unloved, non-competitive, high-barrier infill sites with a lot of potential for growth. We don't clip coupons. Our competitive advantage is in the quality of execution, so we can structure deals with non-market terms that give us some upside that wouldn't be available to merchant builders.

This generally sounds pretty scary to my institutional buddies, because it's ambiguous and driven by narrative. The deal structures are one-off, and the projects require a ton of active management. On the flip side, I don't control the market, so relying on cap rate spreads sounds pretty risky to me. It all depends on what kind of risk you are good at managing.

    • 3
Aug 15, 2017

Very thoughtful. Thanks for your input.

Since you brought it up, how do long term investors who want to hold onto assets analyze deals. I'm familiar with the fact that looking at COC is more important than IRR but how would you value a building on a long term hold since Cash on Cash is one year in time?

What would be the most important factors in an excel spreadsheet to consider? Ex: As you mentioned, exit caps isn't something to look for when doing a 5 year pro forma

Sep 20, 2017

We always underwrite an exit scenario at stabilization, just in case the market tells us that's the best strategy. But generally, we hold. I'll look at a 10 or 15 year sale just to have a capital event against which I can calculate IRR.

Once I've looked at the nearer term exit scenarios, I just model out 20 years of operating income and look at cash flow. We, along with our long term partners, hope to wake up in 15 or 20 years with a bank account full of distributions and a high-value asset. As an owner it's not about driving IRR, it's about building wealth.

    • 2
Dec 23, 2017

Most RE deals I've worked on were with high-net worth equity investors and typically they have been focused on yield almost exclusively unless there was a juicy exit opportunity. And the models were pretty basic.

Got much more complicated with the institutional level projects where financing and the waterfall became more of an input.

Dec 23, 2017

Nice old thread about making it easy.

Cap on cost. If the market is a 4.5 cap, if you can build and stabilize to a 6.5 cap on cost, you're going to make money. I don't do ground up as a broker, but down and dirty I look at cap on cost.

Apr 12, 2018

Why would you need REFM is you can already do a multi-asset dev model?

Apr 12, 2018

I will exchange REFM, BIWS and other solid models I have against your platform model.
PM me if you have a platform model, and we can discuss the details specifically.
Everything will be obviously kept confidential.
To your question, - Why would you need REFM is you can already do a multi-asset dev model - Answer is simple, for your library, like books.

    • 1
Apr 12, 2018

I'm in brokerage so with several different companies I have seen single asset modeling done different ways. Perhaps you are only asking for acquisitions so I'm interested in hearing the answer as well, but I thought I'd give my experience until someone can answer you.

Generally for smaller dollar single assets i have seen underwriting from sponsors "clients" done with rent rolls and some proforma expenses(always too low).

Generally when the asset is over $40 million you tend to(and I do) multi cash flow statement models. I don't think the debt guys are really doing this and I don't think small developers are either. I think the mid-tier and top tier developers and REITS and REPE's do this, along with Equity Funds.

I was just looking today actually at some acquisition model for a property that did have a projected 10-year out proforma but they didn't do a DCF or MOIC or IRR. Also nobody is really looking at the assets on a granular or even monthly level. As far as I'm concerned there is a lot of dumb money in the asset class as you are playing with OPM so you don't need to be really smart.

Also as far as inflation and rent/growth expense rates NOBODY is researching what the actual numbers could be. I have never once seen a conservative model that didn't have growth rates at 5% for the rents(not Office) and expenses rising at 3% a year. You KNOW taxes are going up, why aren't you underwriting to that? I believe lenders and equity sources have so much capital that they are really just trying to lend on any assets possible.

Apr 12, 2018

Hold up. I work for a PE group.

We have never u/w a deal without giving some serious thought and read some serious research on projected market growth. If it's a hotel, you take into account PKF projected rates/occupancies. If it's a MF property/student housing, you take into account axiometrics projections. If it's an office building, run off of CoStar projections. These all have real data behind their forecasts which takes into account employment numbers and historical growth, etc. That being said, if any of these projections seemed rosy, we gladly beat up the growth rates to a point where we feel comfortable.

I can't speak for other funds, but we spend a hell of a lot of time thinking about untrended yields. If you need to rely on market growth to bring in returns near your cost of capital, it's probably a deal you should be passing on unless you have some great data backing your thesis.

As per taxes, Brokers/Operating Partners are almost always aware of rising RE taxes triggered by periodic reassessments or a sale. If you are at any reputable group, you are u/w this into your proforma.

Apr 12, 2018

We definitely don't have an exact science. Maybe it's possible to use a regressions on historical/projected data quantify growth.. but we definitely don't do that. We use an intuitive approach and take into account data provider's growth projections, growth rate historical averages, net absorption historical avgs vs. upcoming supply, growth at comparable times in prior market cycles (not always applicable), current rents vs. prior peak rents etc... If occupancy is at sky-high levels, that's usually an indicator the market is poised for above-avg rent growth to normalize occupancy at lower levels.

My point was that we don't just use a flat 3% and run with that hoping for the best. Didn't realize your post was only speaking on behalf of brokerages.