I received a very astute question/concern from a user of our All-in-One(Ai1) model in our Ai1 Support Forum late last month. I initially set out to answer the question in writing, but the more I thought about my response, the more I concluded a video was necessary to fully explain the methodology here. Below find a brief explanation of the concern raised, the video I recorded explaining why I chose the methodology I did, and a link to download the Workbook used in the video.
Month 0 vs Month 1 as Time 0 in Development Analysis
In a typical real estate development model, cash flows are forecast on a monthly basis with return metrics likewise calculated on a monthly basis. This is mainly because the development cash flows (negative cash flows) that occur early in the analysis are spread out over a series of months (12+) rather than all occurring upfront; as is most often the case with acquisition analysis. Thus, most often the first period (i.e. time 0) of the analysis – where land costs, venture costs, and pre-development costs are often modeled – will usually be named month 0.
However, the naming of time 0 (whether month 0, month 1, or otherwise) is largely semantical in monthly analysis. This is because the Excel function most commonly used to calculate the monthly IRR – XIRR() – relies on a date being assigned to each cash flow. So, the first period might occur 12/31/2018, and regardless of whether that period is named month 0 or month 1, the result of the XIRR() calculation will be the same.
Treatment of Time 0 in the All-in-One
The All-in-One model forecasts all cash flows on a monthly basis – acquisition, value-add, and development – and then rolls those cash flows up to an annual string of cash flows to calculate annual metrics. When the construction module is turned off (i.e. acquisition analysis), the total acquisition cost assumption (M29 of the Summary tab) flows to month 0 (first period) of the DCF as a negative cash flow. Thus, when the construction module is turned off, the first period is named month 0.
However, when the construction module is turned on (M13 of the Summary tab is greater than 0), the first period of the analysis is actually month 1, not month 0. As a result, when the monthly cash flows roll up to annual cash flows, the first period of the analysis is year 1, not year 0. This leads the same IRR on a monthly basis – whether month 0 or month 1 are used for the first period – but a different IRR on an annual basis between the two methods.
As a result, the concern raised was that the annual IRR calculation was incorrect. In the following video, I’ll address this concern.
I first discuss why the annual IRR in development analysis is always going to be less accurate than the monthly IRR, regardless of whether you use month 0 or month 1 for time 0. In other words, if you’re analyzing a development deal, the annual IRR is not the metric for you. And then, I use a basic example (download Excel workbook below) to show my thought process around why I ultimately chose to use month 1 as the first period for development analysis in the All-in-One and the implications of that methodology.
Video Explanation – Treatment of Time 0 for Development Deals in the All-in-One
This was a useful exercise brought about by a great question raised by one of our Ai1 users. I think the moral of the month 0 vs month 1 conundrum, is first and foremost, you should rarely pay attention to annual IRR in development analysis. And second, there’s seldom one right way to model anything. As you’re building your own real estate models, you’ll come across methodology questions all the time that don’t have a clear answer. What’s most important is that you think through the options, and understand why you made the decision you did.
Huge thanks to Ai1 user Rob for bringing this topic to light!