Financial modeling for seniors housing and care is not as easy as it used to be, when revenues could be assumed to rise faster than expenses.
Have you noticed that financial modeling for seniors housing and care is not as easy as it used to be? In the past, forecasters (and buyers) would model 2% increases in annual costs and 3% increases in revenues. This would build in a growing profit amount, which always looked good to lenders and investors.
One of the problems with this was that annual capital expenditures were always, and I mean always, low-balled. Often it was a plug number at $300 per unit or bed, when it should have been three, four or five times that amount.
The problem today is deeper than that. How do you forecast revenues to rise faster than expenses when you often need to discount rents to get residents in the door? Add to that costs, especially labor, which are often rising faster than inflation.
How do you factor in legislated increases in minimum wage rates to your financial forecasts? For those buyers coming into the market and thinking they can manipulate a few numbers and make the deal pop, that just doesn’t work as well anymore. The era of financial manipulation may be coming to an end, as operations are just getting too difficult.
Learn some Excel bud – it really is not that hard.
That is my point exactly. Just using Excel is not going to cut it anymore, because garbage in, garbage out. You need a lot more knowledge and assumptions. How valid are your assumptions in Excel? And those assumptions will change more frequently than in the past.
Interesting and timely topic, Steve. I recently spoke with several investor groups on that subject last week. The overall consensus was that you can’t rely on global/fallback assumptions. You need to drill down in each market and develop assumptions based on the property and the competitive market.
I can run scenario analysis on any set of census and rate assumptions, by payor type, that I want to run through my model. I take into account the activity/seasonality trends over the past three years of historical and I project out five years into the future. I use a sliding scale, ratio-driven staffing tool that pivots off census, by position, that allows me to make sure my nursing labor hour PPD is in line where it should be. I then pivot projected staffing levels, given the census assumptions loaded, against current wage rates and include annual raises in future years (specified by month and %-increase of each position). In terms of non-wage expenses, I can select, by GL-account, if I want that expense to be a flat $ amount with annual growth kickers, or based on skilled mix PPD, just Med A PPD, just Med B PPD, or Managed Care PPD. All of this creates a 5 years pro forma that is loaded into our cash flow returns model, which we can run different financing structures through. Back in the day you may have done forecasting with department-level PPD’s and annual growth rates, but that’s not what we do….
Sounds pretty impressive. Just curious, how have your five-year forecasts stacked up to actual results?
I apologize for the delay in responding – I just saw that you had replied. They stack up quite well thus far. Obviously when there are unforeseen market forces that impact census temporarily (for example, if the market’s referral sources/Hospitals have an unexpected dip in volume – thus discharges to SNF), it will throw off our census forecasts for the year. However that is why we run scenario analysis reflecting best-case, most-likely-case, and worst-case, so that we understand where our range of returns may track. Another benefit of how our models are built is that they allow for a rolling updating of actual performance levels each month thereby allowing us to continually re-forecast, if needed. However for steadily performing assets the model has been more accurate than any operator budget model we’ve seen thus far. We marry our underwriting to an in-depth market analysis as well. No longer do we just rely on market demand studies for our assumptions concerning the local markets, but we source discharge data, market share, demographic growth rates, etc. directly from the same sources used by the firms who do market demand studies to ensure they’re capturing an accurate picture. Mixing our market analysis with the on-the-ground interviews we do during due diligence has given us a pretty thorough and accurate underwriting process. We even forecast managed care penetration rates in specific markets so that we can factor them into the census assumptions of the underwriting model. Obviously there are a lot more factors that go into the process, but my point is that it is quite detailed and leverages Excel’s functionality to accurately produce an output that we can be confident in basing our investment decisions upon.
J,
Just saw your second reply. Sounds pretty detailed analysis, which is good. But just curious how the analysis has stacked up when looking a few years out with actual data?
Steve, I tweaked and expanded my prior comment. See below:
Interesting and timely topic, Steve. I spoke with several private equity groups on that subject last week. The overall consensus was that you can’t rely on global/fallback assumptions in your underwriting. They drill down in each market and develop assumptions based on the property and the competitive market.
I completely agree. It is just not as easy as it was, and it won’t get easier despite demographics.