This was a joy to read and I learned a lot. Thanks! Two high level comments.
First, you argue that upzoning won't be all that helpful because the real problem comes from the capital markets side, and the demand from investors for high IRRs. But don't the high IRR requirements for real estate development result partly from local land use regulations that add whole new dimensions of delay and risk?
As I see it, the right kind of land use reform (see my upcoming essay) would help unlock the financing because there wouldn't be so much red tape, waiting for government permission, and running risks that you'll never get it.
Second, you recommend new federally chartered agencies, expansion of the GSEs' scope of practice, and/or municipal real estate finance interventions similar to what the GSEs do, as a way to lower the financing costs of housing. Maybe-- but the GSEs or similar entities are rather odd and under-theorized, and open to the critique of being inefficient central planning that socializes risk.
I'd be interested in a Substack post by Mike Fellner entitled "The Case for the GSEs," or something like that. :) The operations of Fannie Mae and Freddie Mac don't map cleanly onto the core arguments economists make for why government should have a role in the economy. They're not performing a basic judicial/property rights enforcement function, even if there's a bit of adjacency. They're not doing monetary policy, even though there's some adjacency to financial regulation. They're related to meeting a basic human need (housing), yet they're not a welfare program. They're not exactly/explicitly a subsidy program-- and yet the implicit government guarantee that became explicit in 2008 argued had a big macroeconomic cost. Should the GSEs exist at all? Why? What's the theoretical case?
Two arguments come to mind for me:
1. They're a mechanism for subsidizing homeownership because homeownership (arguably) has positive externalities.
2. Some kind of economies of scale in housing finance are being secure through government leadership... but does that even make sense? I can't quite wrap my head around the model.
My concern is that interventions on the housing finance side, to reduce the IRR required for multifamily housing to clear investment hurdles, would just socialize more risk.
Thanks for the great feedback. I completely agree that delays crush IRRs and the potential for delays drives up requires returns. I wrote about it in another post. I guess I put that in a different bucket than pure land use. Maybe just slow permitting? But I agree, it's a huge problem.
Broadly, I view the GSEs are providing stability in the market place. Imagine the recovery from the GFC with no GSEs. There would have been no mortgage lending for 5 years.
So the "too slow recovery of mortgage lending after GFC with no GSEs" scenario seems like a good starting point for a "case for the GSEs" argument.
But I still want to understand the downsides of the non-existence of GSEs better. No mortgage lending ≠ no home sales. What if housing prices fell more, lots of people bought mobile homes and saved, and then housing markets reemerged at lower price points, with far less financing and far more cash? Is that good or bad, and why? What's the model?
I found this essay enlightening. Where I focus on the demand side of housing finance, this sheds light on the supply side, and I think the two pieces complement each other well.
It’s interesting that it really does come down to a supply-demand issue, just not of housing units themselves, but of finance.
In that light, my proposed transfer surcharge, which limits credit available for mortgage principal and redirects it toward returns on new construction, ends up addressing both sides of that imbalance.
This is interesting, but doesn’t really account for the crux of the abundance claim. Reality is, crediting everything you note here, in places like New York and San Francisco and LA, you can grow your stream of income by building up. You can build a 100-unit high rise on the same footprint as a 50-unit mid-rise. It will not cost twice as much (or close to twice as much) to build the 100-unit high rise, and your rents will more than double, both because there are twice as many units and because the top floors of a 40-story building will have better views than the top floors of a 20-story building, which will translate to significantly higher rents.
Then people who move into that new build will move out of other places, and others will move into their vacated units at marginally lower price points. Rinse and repeat.
I could very well cost you more than twice as much. And even if you do get economies of scale, the capital will be more expensive because the more complex the project, the higher the risk.
It wouldn't cost you twice as much. There are economies of scale to building straight up; extending plumbing, for instance, is much easier than building it from scratch. Same with electric, HVAC, etc. There's significantly less financial risk to just building a taller building in a high-demand expensive metro. That's the crux of the abundance argument.
I am sorry but nobody in the industry would agree with you because it's not reality. Complex projects are riskier and frequently run into diseconomies of scale. It sucks but it's reality.
Take a step back. Yeah, if you, say, need to add an elevator or whatever, you’re adding cost. But when talking about abundance, we’re talking about places with sky high land costs and demand. Building a skyscraper in Tulsa isn’t economical because land is cheap and you’re not gonna get a lot of rent out of a penthouse there.
But we don’t have a housing cost issue in Tulsa. It’s hugely more expensive to build a 10 story mid rise than a mansion in Manhattan. But a 10 story mid rise in Manhattan is hugely more profitable than a mansion in the same footprint, under any assumptions. Same with a 100 unit building vs a 40 unit one. Hell, a penthouse alone could cost more than a townhouse on the same lot.
And that’s the point— when rents are sky high, you can generate lots of rent by building a ton of units in a space. And, more importantly, you then drive down costs elsewhere because you don’t need to do anything around the remaining housing stock besides updating it; you’re just eating into the premium paid for that stock because it’s less scarce. Construction costs for that stock aren’t relevant; they’ve already been paid.
Height only works for you when it means you can overcome disecononomies of scale that come with denser construction modes by spreading land costs over more units. This is correct. But even in a place like Manhattan the math does not always work. And frequently, height itself must be an amenity enabling higher rents. . (Views, perceived security, ect)
It generally works far better in the exact places where housing costs are an issue-- NYC, San Francisco, Seattle, etc. The logic is somewhat self-reinforcing-- the more scarce housing is, the more rents a landowner can collect from a parcel. At that point, the economics of building become better and better, all things being equal. Unless you break your city altogether and make it unlivable, which can also be done by pricing out anyone that performs lost of essential services (try living in SF on two teachers' salaries).
But generally what stops more dense housing from being built in those places isn't the economics of building-- it's zoning, NIMBY community boards, rent controls and "affordable" set asides, etc.
It’s not. It’s the economics no matter how much you don’t want to believe it. The economics for multifamily have been horrendous in SF and LA for the past decade or so. They’ve been just OK in NYC. You just don’t know what you don’t know.
I really enjoyed this article. Very clear, very concise-- thank you for sharing.
I do have a question-- shouldn't changes in land value (or more specifically, parcel desirability) be a factor considered with regards to the economics of a multifamily acquisition? Rent growth seems to capture only a piece of this consideration.
I'd also be curious to learn how mixed-use developments are evaluated for feasibility. Pretty common to see new ~4-story developments where the first floor is retail. Does the retail space become a major consideration to the economics of the venture, or is it just a footnote?
Mixed use development is harder to make work, since it requires investors who want exposure to both residential and commercial real estate. This is a smaller investor pool.
YIMBY here, enjoyed this essay. Tackling financing increasingly looks like the next big issue for federal YIMBYism; there’s already some movement at the state and local level to defer impact fees (see CA’s SB 937 from 2024). All important work, thanks for some thought-provoking recommendations
One part of this essay that I found off putting even as I agreed with the basic premise: As important as financing is, I’m always surprised by the either/or framing of financing vs cost reduction. Reducing the cost of development is the other side of making IRRs higher that can allow development to occur in a lower-rent environment. YIMBYs have been dealing with a lot more than zoning for years; we’ve been trying to tackle all the sources of cost and government-imposed risk. Discretionary permit processes, legal veto points, local fees, costly labor mandates, barriers to new construction techniques—we’ve been working on reforming all these and more. For example, when people poo-poo Austin’s building boom, it seems a bit one-sided. Sure, the city isn’t as much a pure success as YIMBYs say, but it’s at least partly a success! A lot of places in the U.S. saw similarly large COVID rent jumps without the following building boom and rent drops *bc Austin took action to lower costs that made such a boom possible* in a way most other places hadn’t
All this is to say, yes, let’s tackle financing. Yes, let’s explore vehicles to create lower cost debt and equity. AND let’s keep working to lower costs
I don't disagree directionally with any of that. I think where we might disagree is the magnitude. How much can you really reduce costs with deregulation? I think the upper bound of this is really about 15 percent max. More likely, it's about 5 percent. Nothing to sneeze at, but it's a one off.
On the other hand, if you could lower required returns such that new housing could go up at 1 percent rent growth instead of 3, you could massively lower housing costs in the long run.
This was a joy to read and I learned a lot. Thanks! Two high level comments.
First, you argue that upzoning won't be all that helpful because the real problem comes from the capital markets side, and the demand from investors for high IRRs. But don't the high IRR requirements for real estate development result partly from local land use regulations that add whole new dimensions of delay and risk?
As I see it, the right kind of land use reform (see my upcoming essay) would help unlock the financing because there wouldn't be so much red tape, waiting for government permission, and running risks that you'll never get it.
Second, you recommend new federally chartered agencies, expansion of the GSEs' scope of practice, and/or municipal real estate finance interventions similar to what the GSEs do, as a way to lower the financing costs of housing. Maybe-- but the GSEs or similar entities are rather odd and under-theorized, and open to the critique of being inefficient central planning that socializes risk.
I'd be interested in a Substack post by Mike Fellner entitled "The Case for the GSEs," or something like that. :) The operations of Fannie Mae and Freddie Mac don't map cleanly onto the core arguments economists make for why government should have a role in the economy. They're not performing a basic judicial/property rights enforcement function, even if there's a bit of adjacency. They're not doing monetary policy, even though there's some adjacency to financial regulation. They're related to meeting a basic human need (housing), yet they're not a welfare program. They're not exactly/explicitly a subsidy program-- and yet the implicit government guarantee that became explicit in 2008 argued had a big macroeconomic cost. Should the GSEs exist at all? Why? What's the theoretical case?
Two arguments come to mind for me:
1. They're a mechanism for subsidizing homeownership because homeownership (arguably) has positive externalities.
2. Some kind of economies of scale in housing finance are being secure through government leadership... but does that even make sense? I can't quite wrap my head around the model.
My concern is that interventions on the housing finance side, to reduce the IRR required for multifamily housing to clear investment hurdles, would just socialize more risk.
Thanks for the great feedback. I completely agree that delays crush IRRs and the potential for delays drives up requires returns. I wrote about it in another post. I guess I put that in a different bucket than pure land use. Maybe just slow permitting? But I agree, it's a huge problem.
Broadly, I view the GSEs are providing stability in the market place. Imagine the recovery from the GFC with no GSEs. There would have been no mortgage lending for 5 years.
See here for how delays kill returns.
https://open.substack.com/pub/precon/p/evaluating-the-impact-of-fees-property?utm_source=share&utm_medium=android&r=4uiw0
So the "too slow recovery of mortgage lending after GFC with no GSEs" scenario seems like a good starting point for a "case for the GSEs" argument.
But I still want to understand the downsides of the non-existence of GSEs better. No mortgage lending ≠ no home sales. What if housing prices fell more, lots of people bought mobile homes and saved, and then housing markets reemerged at lower price points, with far less financing and far more cash? Is that good or bad, and why? What's the model?
I found this essay enlightening. Where I focus on the demand side of housing finance, this sheds light on the supply side, and I think the two pieces complement each other well.
It’s interesting that it really does come down to a supply-demand issue, just not of housing units themselves, but of finance.
In that light, my proposed transfer surcharge, which limits credit available for mortgage principal and redirects it toward returns on new construction, ends up addressing both sides of that imbalance.
If you aren’t already subscribed to Mike you should!
This is interesting, but doesn’t really account for the crux of the abundance claim. Reality is, crediting everything you note here, in places like New York and San Francisco and LA, you can grow your stream of income by building up. You can build a 100-unit high rise on the same footprint as a 50-unit mid-rise. It will not cost twice as much (or close to twice as much) to build the 100-unit high rise, and your rents will more than double, both because there are twice as many units and because the top floors of a 40-story building will have better views than the top floors of a 20-story building, which will translate to significantly higher rents.
Then people who move into that new build will move out of other places, and others will move into their vacated units at marginally lower price points. Rinse and repeat.
I could very well cost you more than twice as much. And even if you do get economies of scale, the capital will be more expensive because the more complex the project, the higher the risk.
It wouldn't cost you twice as much. There are economies of scale to building straight up; extending plumbing, for instance, is much easier than building it from scratch. Same with electric, HVAC, etc. There's significantly less financial risk to just building a taller building in a high-demand expensive metro. That's the crux of the abundance argument.
I am sorry but nobody in the industry would agree with you because it's not reality. Complex projects are riskier and frequently run into diseconomies of scale. It sucks but it's reality.
Take a step back. Yeah, if you, say, need to add an elevator or whatever, you’re adding cost. But when talking about abundance, we’re talking about places with sky high land costs and demand. Building a skyscraper in Tulsa isn’t economical because land is cheap and you’re not gonna get a lot of rent out of a penthouse there.
But we don’t have a housing cost issue in Tulsa. It’s hugely more expensive to build a 10 story mid rise than a mansion in Manhattan. But a 10 story mid rise in Manhattan is hugely more profitable than a mansion in the same footprint, under any assumptions. Same with a 100 unit building vs a 40 unit one. Hell, a penthouse alone could cost more than a townhouse on the same lot.
And that’s the point— when rents are sky high, you can generate lots of rent by building a ton of units in a space. And, more importantly, you then drive down costs elsewhere because you don’t need to do anything around the remaining housing stock besides updating it; you’re just eating into the premium paid for that stock because it’s less scarce. Construction costs for that stock aren’t relevant; they’ve already been paid.
Height only works for you when it means you can overcome disecononomies of scale that come with denser construction modes by spreading land costs over more units. This is correct. But even in a place like Manhattan the math does not always work. And frequently, height itself must be an amenity enabling higher rents. . (Views, perceived security, ect)
It generally works far better in the exact places where housing costs are an issue-- NYC, San Francisco, Seattle, etc. The logic is somewhat self-reinforcing-- the more scarce housing is, the more rents a landowner can collect from a parcel. At that point, the economics of building become better and better, all things being equal. Unless you break your city altogether and make it unlivable, which can also be done by pricing out anyone that performs lost of essential services (try living in SF on two teachers' salaries).
But generally what stops more dense housing from being built in those places isn't the economics of building-- it's zoning, NIMBY community boards, rent controls and "affordable" set asides, etc.
It’s not. It’s the economics no matter how much you don’t want to believe it. The economics for multifamily have been horrendous in SF and LA for the past decade or so. They’ve been just OK in NYC. You just don’t know what you don’t know.
I really enjoyed this article. Very clear, very concise-- thank you for sharing.
I do have a question-- shouldn't changes in land value (or more specifically, parcel desirability) be a factor considered with regards to the economics of a multifamily acquisition? Rent growth seems to capture only a piece of this consideration.
I'd also be curious to learn how mixed-use developments are evaluated for feasibility. Pretty common to see new ~4-story developments where the first floor is retail. Does the retail space become a major consideration to the economics of the venture, or is it just a footnote?
Thanks! To answer your questions;
1.Land prices are largely a reflection of rents.
Mixed use development is harder to make work, since it requires investors who want exposure to both residential and commercial real estate. This is a smaller investor pool.
YIMBY here, enjoyed this essay. Tackling financing increasingly looks like the next big issue for federal YIMBYism; there’s already some movement at the state and local level to defer impact fees (see CA’s SB 937 from 2024). All important work, thanks for some thought-provoking recommendations
One part of this essay that I found off putting even as I agreed with the basic premise: As important as financing is, I’m always surprised by the either/or framing of financing vs cost reduction. Reducing the cost of development is the other side of making IRRs higher that can allow development to occur in a lower-rent environment. YIMBYs have been dealing with a lot more than zoning for years; we’ve been trying to tackle all the sources of cost and government-imposed risk. Discretionary permit processes, legal veto points, local fees, costly labor mandates, barriers to new construction techniques—we’ve been working on reforming all these and more. For example, when people poo-poo Austin’s building boom, it seems a bit one-sided. Sure, the city isn’t as much a pure success as YIMBYs say, but it’s at least partly a success! A lot of places in the U.S. saw similarly large COVID rent jumps without the following building boom and rent drops *bc Austin took action to lower costs that made such a boom possible* in a way most other places hadn’t
All this is to say, yes, let’s tackle financing. Yes, let’s explore vehicles to create lower cost debt and equity. AND let’s keep working to lower costs
I don't disagree directionally with any of that. I think where we might disagree is the magnitude. How much can you really reduce costs with deregulation? I think the upper bound of this is really about 15 percent max. More likely, it's about 5 percent. Nothing to sneeze at, but it's a one off.
On the other hand, if you could lower required returns such that new housing could go up at 1 percent rent growth instead of 3, you could massively lower housing costs in the long run.
Hey guys! 👋My blog is all about helping rental property owners with property management and maintenance tips.
Check out one of my posts here: https://substack.com/@damian168479/note/p-182100015?utm_source=notes-share-action&r=6x9xlc
Feel free to share with other landlords!