The Elephant In The Land Acq Room: Underwriting For Inflation
Underwriting new projects has never been harder (at least outside of downturns). Yes, it was never easy when done well, but a confluence of factors has raised the degree of difficulty:
- Shortages of entitled ground have made the competition for it fierce, and as previously discussed, improvements in the public builders have allowed them to be aggressive yet profitable bidders for it.
- Long approval time frames and a frequently random and capricious process to get there raise the distance into the future we are estimating values. How much emphasis does one place on comparable analysis today if the delivery date of a home could be five years or more from now?
- Larger deposits for land and much higher entitlement expenditures mean the cost of speculating on approvals has increased substantially. There is no consistent source of capital for processing other than your own. So, how do you decide which projects are worth the risk?
Moreover, we no longer supply the bottom of the market, so affordability calculations are skewed. If you take the bottom third of the market and eliminate it, the median goes up, and the income ratios get worse. But, as supply falls and we no longer supply entry-level to any degree, the question is not the overall affordability of a market (sad and unnecessary as that is) but more of a matching exercise – are we under- or over-supplied in the price points we are serving? If there are more buyers than supply at each price point, affordability can be awful, and we still have price up-lift pressures.
Most of us learned to analyze a project from the top down: Look at comparable projects or triangulate amongst different types of projects if there are no direct comps. Sometimes – for unique product or locations – we would construct "alternatives analysis" looking at what buyers could get for our projected price elsewhere, or where they would have to go get what we are offering. But start with what you think the home would be worth if delivered today. Then, subtract estimated costs and required rates of return and arrive at land values. An imperfect science perhaps, but at least grounded in current reality.
Being honest, the industry has worked based on using this approach and then making several aggressive assumptions for schedules, costs, and price premiums to get to a hurdle number required for approval. It's not ideal, and in the past, I have argued for more realistic hurdles and more rigorous numbers, but if the approach worked to turn what was probably a 6% margin into a 9% to get done, it was not a tragedy (usually). But many people today find it more like being aggressive to turn a 0% into a 6% margin. That is scary.
So, what to do? Wait it out – and for how long? – hoping land values allow more room for traditional underwriting? Dive in and hope for the best? For most, the first option is not feasible as they look at delivery pipelines. The second option seems like going to Vegas and putting it on black.
I think the question is properly framed as, how does one underwrite for deliveries that often will occur five years in the future (+3 yr. approvals + 1 yr. LD + 1 yr. models and units………….)?
The elephant in the room is how much inflation you need to factor into your proforma. [As a side note, it has always amused me to hear people who tend to push their assumptions hard and unrealistically flinch at explicit inflation. We are a funny industry.]
The reality is there is no great answer to this. But I think it is possible to make more intelligent bets on the future than simply “to go with your gut” or, randomly, to project 3% annual revenue and cost inflation, or about in that neighborhood. You could argue for using past inflation rates in your market. At least there would be a basis in history, but I worry that in most markets that would include a period when supply and demand were not nearly so out of balance as they have been during the pandemic period. Which period is valid? Either? Average of both?
It probably makes more sense to examine the future supply/demand balance and compare it to the previous balance for clues.
The problem with the current comp analysis is that, in most cases, it is based on non-replicable costs. Land, horizontal development, and tap and impact fees have increased dramatically since many of today’s open projects were approved/developed. This raises a question: Will future home prices in these areas be higher, or will we not build anything there? You could argue that land prices will fall, but I would not bet on it. Most landowners today are reasonably well-capitalized. They can and will just wait for their number.
Since my underlying assumption is that we are not going to stop being able to build and sell homes (although they are likely to be higher priced and lower in total volume), how might one approach this logically? I think the answer is that we are going to have to get better at our macro analysis of markets.
When doing large-scale land deals, we routinely did a “pipeline analysis.” In other words, we scheduled out the potential competitors over time, the likely time frame those projects would deliver, and price points, and assumed all would get their “fair share” of the market. We would then determine – after those filters – whether that parcel's market seemed over- or under-supplied. Often, the analysis would cross over into other sub-markets that might be alternatives for our end users. Is an adjacent sub-market running out of lots? This might suggest that ours might get more than historical sales. Or, the opposite: A new area may open up that will syphon off some demand?
While many would say they already do this on modest-sized deals, it is usually a feeling they have based on struggling to find deals to buy rather than a detailed parcel-by-parcel analysis.
Intuition is probably fine for “it’s getting tighter,” but it is not up to the task of answering how much tighter: For example, 20% fewer deliveries five years from now? How about 50% less?
I am not arguing this is an exact science. There are too many variables for that to be true. Will that project – burdened by too much infrastructure to open – find a creative solution and get developed? Will entitlements get harder or easier – thereby changing what is delivered (I would bet on the former)?
Still, we are trying to improve our odds, not find guarantees.
What are the paths of growth to focus on? Once upon a time, this discussion often focused on employment growth areas and future transportation infrastructure. These still matter greatly, but there is little infrastructure building in many markets, and the path-of-growth is defined by sewer and water capacity/availability. So, sad to say, in many ways, in some locations, the determinants will amount to what will be possible, not what should logically be.
One area where our data has improved dramatically is demographic analysis. Projecting income growth in areas is critical to answering whether buyers can pay more if supply and demand analysis says they will need to. Projecting the future is tricky and can be a bit circular. If an area will/can produce a lot of new housing, it will probably attract buyers from elsewhere, raising the income profile. However, we should consider this exercise more like a probability exercise than a precision activity: No numbers to the right of the decimal.
Then, the multi-million-dollar question.
What does this all add up to for projected home and lot prices? It is not like there is a formula to plug all this info into and get a definitive answer.
Here, I will go out on a limb with this suggestion: Determine what number would “work,” and then take a sanity check. It will rarely be super clear. Still, it seems a big stretch and a bad idea if you forecast sub-market volumes to be down by 10%, but prices will be up 25%. If, on the other hand, you think the sub-market will have 40% less volume, will not likely be “replaced” with another growing submarket, and you need prices to be up 15% five years from now, it seems plausible (let us not to forget to inflate costs.)
Working through these issues results in a replacement cost analysis married to a future supply and demand balance.
These ideas are best used where there is an extended period before the delivery of a product. Banking on price movements supported by macro analysis over five years is much different from thinking you can predict price increases two years from now.
You might wind up right, but remember, "early and right can be the same things as wrong.” We are talking here about having the time for anticipated changes to happen. And making these kinds of high-risk assumptions to justify a deal that would produce mediocre margins is way too risky. We still need to underwrite where we would operate today on present conditions, which hopefully is a floor under our speculative ceiling.
At this point in the column, I suspect that I have some past co-workers and partners who are double-checking who the author is. But it would be putting my head in the sand to suggest that if you merely try harder, you will find plenty of deals that pencil using today’s numbers. So, if we are taking risks and betting on the future, let us at least improve the odds of our bets.
Food for thought.