Property Type as Regime
A hotel, an office building, and an apartment complex are not three settings of one model. They are three different games, with different rules for what a number means, where trouble shows up first, and how a loan gets out. Reading them the same way is the deepest version of the mislabeling error — not putting a loan in the wrong frame, but not believing the frames are different at all.
The previous chapter was about labels that lie. This one is about the harder mistake underneath it: treating property type as a tweak — a different intercept on the same line — when it is actually a different line. Type doesn’t just shift the level of a healthy DSCR; it changes the slope of everything. A hotel’s income is repriced every single night, so its coverage is volatile by construction and a one-quarter dip below 1.0x can be pure seasonality. An office’s income is locked into multi-year leases, so its coverage is smooth and reassuring right up until a lease rolls and doesn’t renew — its danger is concentrated at rollover, invisible in between. Multifamily reprices monthly but gently, a thousand small leases averaging out. Retail lives or dies on an anchor. Same metric, four different physics.
So the same DSCR trajectory carries opposite meanings across types, and reading it without the regime is how you get the loan exactly backwards. A hotel wobbling below coverage for a quarter is often noise; an office doing the identical thing may be a tenant that has already given notice and a number that is never coming back. A multifamily running sub-1.0 in its second year could be a lease-up going to plan; a stabilized office running sub-1.0 is a structural problem. You cannot carry one distress threshold across a mixed book — sub-1.0 is designed in one regime, seasonal in another, and catastrophic in a third. A screen that ranks a whole pool by raw DSCR is blending regimes and mis-ranking every loan against loans that play a different game.
The reading move is to read each loan inside its type’s regime, which means knowing the regime’s tells cold. For a hotel, watch the revenue itself (RevPAR) and read coverage as inherently noisy. For an office, watch the lease rollover schedule — the anchor’s expiry against the loan’s maturity — because that is where distress announces itself long before DSCR moves. For retail, watch the anchor tenant’s own health. For multifamily, watch occupancy and concessions, and give lease-up its designed sub-1.0 room. The platform’s screens carry this awareness — the Stress view and the Vintage lens don’t score a hotel and an office on one flat rule — and your reading has to carry it too, or the tools will be smarter than the reader driving them.
From outside a trust, this compounds, because a trust is a mix of types, and reading the pool means reading each sleeve in its own regime rather than trusting one blended number. It also tells you where the market’s largest mistakes cluster: they gather wherever participants apply one type’s instincts to another, or one era’s regime to a type that has changed underneath it. Reading office in 2024 with the pre-2020 frame — treating long leases as safety when the question is whether the tenant renews at all — is a regime error playing out across an entire asset class at once. The reader who knows there are different games, and which one each loan is actually playing, is reading a different book than the crowd scoring everything on one line.