Confidence Vs Room for Error in Underwriting

When putting together an underwriting model for a multifamily investment, there exists a delicate balance between aggressiveness and margin for error. Being too extreme in either direction will yield unsatisfactory results. If you’re too aggressive in your assumptions, you’ll overpay and possibly get into trouble down the road. If you’re too pessimistic or build in too much room for error, you’ll never buy anything, especially in this fully-priced market! Today, I’ll discuss how we implement the right balance of aggressiveness and room for error as well as the difference between being conservative from a business plan and a projection perspective. 

In case you’re interested, here is an article about our take on conservative underwriting: Thoughts on Underwriting Fundamentals. Accurate underwriting doesn’t necessarily need to build in much room for error as the goal is to accurately reflect/project reality. Additionally, it isn’t necessary to project failure or incompetence because you wouldn’t do any deal if that were the case. Instead, modeling an investment based on realistic yet desired assumptions will output an IRR, which can then be used to determine whether the investment has merit on a risk-adjusted basis (does the hoped-for return compensate the investors well enough for taking on the risks associated with the market, asset, and business plan?). Furthermore, room for error can/should be modeled in via the following; ample reserves, a healthy spread between bridge loan and take-out proceeds (to mitigate refinance risk), and realistic tax increase (property taxes in Texas are a bit of a wildcard and typically go up significantly for a value-add deal since the assessor’s office knows you’re putting money into the property and raising the NOI). 

You don’t want to incorporate too much room for error and get too far away from reality (renovated rent comps show $800 rents but you input $700) as this will dramatically reduce the likelihood of being the winning bidder. Additionally, as you increase your purchase price during the bidding process, you will need to push on your numbers and become less confident with your underwriting as you continually change your assumptions.

Today, investors are competing with buyers who are not factoring in any room for error, meaning they are projecting to be able to manage the property perfectly, with high occupancy for 5+ years as well as consistent, above historical average rent growth. To be fair, we too are sometimes forced to make these potentially unrealistic assumptions in our own underwriting as this is the price of admission in order to get a deal under contract and funded. I feel the best way to walk this tightrope of valuation is by trying to remain realistic while embedding some optimism into your underwriting by incorporating contingencies, multiple exit strategies, flexible or conservative capital structures (debt and equity), and by avoiding binary risks. 

In less familiar markets, we tend to embed more conservatism and room for error in our underwriting. Conversely, in markets in which we feel we have a lot of knowledge and expertise, we can be more confident in our assumptions which in turn allows to bid more aggressively on deals. In order to gain market insight, we have narrowed our focus to mainly one market (see our white-paper “Why Houston Multifamily?”), are continually underwriting many assets in various submarkets, perform comprehensive comp work and consult boots on the ground expertise. Nonetheless, we try to get more aggressive in our underwriting without taking on higher default risk and other binary risks. For example, a lower exit cap isn’t materially detrimental to a deal because it doesn’t change any of the income and expense assumptions. However, even casual investors know this trick and will base an investment’s merit on its cash flows. On the flip side, one of the riskiest assumptions is to project a great increase in income from trailing financials to year one pro forma. Growth assumptions are overlaid and compound on top of unrealistic year one numbers which will significantly affect the trajectory of the whole underwriting. This is particularly impactful to a ten-year cash flow model as the effects of compounding are given many more years to stray even further from reality. Furthermore, it is helpful to discern where NOI increases are coming from. Expense reductions happen much faster than income increases so it behooves you to identify where the projected NOI increases are coming from in year one.  

Lastly, there is a difference between conservative underwriting from a business plan and forecasting perspective and conservative model construction. We previously discussed conservative business plans regarding rent growth, stabilization timelines, and projected sale values. However, model construction is almost never discussed. There are a million ways to model cash flows and calculate assumptions and they are certainly not all equal. For example, an exit cap rate is not all there is to a terminal valuation. Two models using the same exit cap assumption with the same exit year NOI could still come up with different valuations. This is because the NOI being divided by the exit cap could be based on T12 or T1 or somewhere in the middle (we base our exit NOI on T3 income over T12 expenses). This NOI could also be tax-adjusted by the future buyer’s likely projected increase in taxes. Does this NOI include capital expenditure reserves at $300 per unit? Another example is also related to reserves. Does the model assume that all capex reserves, which are reserved with the lender in an escrow account, are spent or are they partially refunded upon sale? Along these lines, are the tax and insurance escrows made upon acquisition refunded back to the owner upon sale or is this not accounted for? Does the rent growth assumption begin after rents have been stabilized at the new pro forma values or are they being organically grown while they are also being raised to renovated value (our model assumes no rent growth during our stabilization period, which is typically the first 12 to 24 months)? 

When evaluating any opportunity, you should always underwrite a deal yourself using your own model as this is the model that will yield you a higher level of confidence than a model you are not intimately familiar with. At the end of the day, underwriting is all relative due to the myriad formulas, assumptions and the varying methods of valuation mentioned above. This means that the returns of an investment should not be based entirely on the projected numbers (18% IRR) and instead be a factor of relative valuation. This means that if you underwrite 100 deals using your own process, you should pursue the deals which underwrite in the top 5 percentile. For us, most of the deals we see underwrite to anywhere from an 8% to negative IRR. This means we get excited when a deal pencils to a 15%+ IRR with an average risk coefficient.