Saturday, May 13, 2023

Missing mortgage contract innovation

From WSJ 

"Many Americans who want to move are trapped in their homes—locked in by low interest rates they can’t afford to give up. 

These “golden handcuffs” are keeping the supply of homes for sale unusually low and making the market more competitive and pricey than some forecasters expected.   

The reluctance of homeowners to sell differentiates the current housing market from past downturns and could keep home prices from falling significantly on a national basis, economists say."

What's going on? US 15 or 30 year fixed-rate mortgages have a catch -- you can't take it with you. If interest rates go up, and you want to move, you can't take the old mortgage with you. You have to refinance at the higher interest rate. It's curiously asymmetric, as if interest rates go down you have the right to refinance at a lower rate. 

As a result, yes, people stay in houses they would rather sell in order to keep the low interest rate on their fixed rate mortgage. They then don't free up houses that someone else would really rather buy. (In California, the right to keep paying low property taxes, which reset if you buy a new house also keeps some people where they are. And everywhere, transfer taxes add a small disincentive to move.) 

This is a curious contract structure. Why can't you take a mortgage with you, and use it to pay for a new house? Sure, mortgages with that right would cost more; the rate would be a bit higher initially. But fixed rate mortgages already cost more than variable rate mortgages, and people seem willing to pay for insurance against rising rates. I can imagine that plenty of people might want to buy that insurance to make sure they can live in a house of given cost, though not necessarily this house.  Conversely, fixed-rate mortgages that did not give the right to refinance, where you have to pay a penalty to get out of the contract if rates go down, would also be cheaper up front, yet people aren't screaming for those. 

Even the right to refinance at a lower rate is weird. A straightforward mortgage would have a 30 year fixed rate, but automatically lower that rate as other interest rates go down. Instead, you have to go through the formalities of refinancing, which adds a lot of fixed costs to the decision. I know a lot of very sophisticated finance people. Not one has ever reported that they've really solved the complex option pricing problem, when is it optimal to refinance a conventional mortgage?

The 15 and 30 year fixed rate mortgage, with right to refinance, is peculiar to the US. You can't make a psychological argument for it. Most of Europe has variable rate mortgages. And a lot less interest rate risk on bank balance sheets! 

So why are we here, and given that we are here why does this strange contract seem so resistant to innovation. I think the answer is simple: 15 and 30 year fixed rate mortgages were a creation of the federal government during the Great Depression. And today the vast majority of mortgages are securitized via Fannie Mae, Freddy Mac, VA etc, along with a generous government guarantee. Those have to conform to specific contract structures. You can't innovate better contracts and then pass the loan on via government agencies. (Commenters, correct me if I'm wrong. My recollection of the history is foggy here.) 

This all points to an interesting and usually unsung problem with extensive government intervention in the mortgage market: It freezes contract terms. Contracts that might be very popular -- such as the right to transfer the mortgage to a new house, or the right to settle up in both directions, marking the mortgage to market so you can pay a new higher rate -- don't get innovated. 


The is not, of course, a particularly original thought. Alexei Alexandrov, Laurie Goodman, and Ted Tozer at Urban Institute have a nice article advocating streamlined refinancing. They also point out the Fed should care, as it wants interest rates faced by borrowers to adjust more quickly. Ted Tozer points out that you can leave it behind -- a new buyer can assume an existing mortgage. However this feature doesn't often get used. 

I once was at the Swedish central bank talking about monetary policy. They were worried about raising interest rates. I presumed they were worried that too big to fail banks would have trouble. No, they said. In Sweden practically all mortgages are floating rate. And you can't just mail in the keys and default on mortgages. If you default, they take all your assets and garnish your wages. (So much for soft hearted socialist Scandinavia.  They are actually quite attuned to incentives.) The banks were going to be fine. They were worried that if they raised interest rates, people would do anything to pay their higher mortgage rates, and this would tank consumption. Talk about effective monetary policy! At the time however they were worried about house prices, and didn't want effective monetary policy. Long story short, mortgage contracts matter.  

(Martin Flodén, Matilda Kilström, Jósef Sigurdsson and Roine Vestman document this "cashflow channel" in the Economic Journal. It's on the back of my mind also in the search for better mechanisms to understand whether and how higher interest rates lower inflation.)  

Roger Baris writes: 

I thought you might want to know a bit about the Danish mortgage market, which basically has the features both you and the Urban Institute mention.

1. A very old securitization market. Created in the early 19th century to help rebuild Copenhagen after the Brits shelled it when Denmark complied with Napoleon's "Continental System" which blockaded trade with England. I guess the Brits really hated being cut off from all that fine Danish butter and bacon.

2. At the initiation of the mortage, the borrower basically swaps his commitment to pay a 30-yr, fixed rate mortgage for bonds in a large, largely homogeneous  (with respect to maturity and interest rate) bond issuance. The bonds are then sold at the prevailing market price, with the proceeds providing the financing for the house buyer.

3. The loan is prepayable at any time. Like the UI proposal, the originating bank (which continues to service the loan) actually prompts borrowers with high interest rates to prepay (since the bank earns some fees in the process and therefore has an economic motivation). There is no re-underwriting of the loan at this point so long as the amount of the loan does not increase. This is a very quick and very cheap process. This is used if interest rates fall (in which case, the borrower's new loan is instantly securitized into a new bond issuance with uniform characteristics, with an equivalent amount of the old bond being prepaid) or if the borrower wants to prepay the loan for a house move. (If I remember correctly, the loan may be also be "portable" to a new house but in this case, the value of the new house has to be re-underwritten.)

4. Note that the servicer in a US mortage is contractually forbidden to prompt prepayments in this manner, although enforcement of this provision is sometimes difficult, as you can imagine.

5. If interest rates rise, conversely, the borrower has the option (if he/she wants to prepay the loan for any reason, especially a house move) of going into the bond market (which is highly liquid with large, uniform pools) and buying an equivalent face amount of bonds (at a discount because rates have risen) and then delivering these bonds to extinguish his/her mortgage liability.

6. In practice, all the bond market activity (both at issuance and extinguishing) is handled for the borrower by a mortgage bank.

 The net effect of all of this - the loan prepayment right, the "portability" and the option to deliver bonds to extinguish a loan - is to make the embedded interest rate option in a Danish fixed-rate mortgage more optimal than the same option in a US mortgage. This means that, as you would expect, the borrower pays a higher "spread" on the loan from the beginning in return for the greater optionality.

PS. The Danish mortgage system also has some interesting characteristics in terms of mutualizing risk; these characteristics also interact with the interest rate option. 


  1. I worked at a proptech startup and you nailed it re: conforming mortgages and the market power of Fannie and Freddy. It would be very expensive and difficult to stand up a parallel supply chain of unique mortgage products to market to customers, originate, then sell to institutional investors without the backing of F&F.

  2. What happens if you try to make a deal with your bank-- you allow me to transfer my low-interest mortgage to my new house, and I give you $10,000? Or even, you give me $8,000 and I refinance my mortgage at the new high interest rates (after which I pay it off and buy a new house). I suppose bank branch presidents are dullards not authorized to do anything so clever-- am I right on that?

    1. Eric, I had a similar thought. Both the mortgage holder and the bank can enter into a Coasian bargain. The mortgage holder keeps their mortgage by arranging a trade with bank. The bank has an incentive to reduce the its opportunity cost between the mortgage holder's low rate, say 2% and the current rate of 6%. Hypothetically both parties agree to a 3.5% rate . On the margin, both party and counter party are better off. The mortgage holder keeps their mortgage and moves while the lender gets a 100bps increase in loan interest.

    2. The problem is that mortgage lenders do not own the mortgages after they are funded. The notes are immediately sold to one of the Federal Government Agencies which in turn ties them up into packages that are conveyed to a trustee for the holders of the mortgage backed securities that fund the agency. The trustees have no authority to modify mortgages. If the mortgages are defaulted they can be put back to the agency at par, but if they are not defaulted payment is full is the only way out. This system created real problems during the Panic of 2008. But, there you have it.

    3. The ''mortgage holder'' is the mortgagee, i.e., the original lender or the government agency that purchases the mortgage contract from the original lender. The homeowner who mortgages her property in order to finance the property purchase is the mortgagor. She doesn't ''hold the mortgage'', nor does she have property rights in the mortgage contract. She can't enter into the bargain that you propose except by repaying the existing loan that is secured by the title deed to the property, and entering into a new mortgage contract on her next home. Whether the lender or originating bank will agree the non-market interest rate on the new loan or not depends on whether the mortgagor (borrower) is a Mark Zuckerberg and the lender (bank) is a First Republic Bank.

      Is there a 'Coasean' bargain to be made? It seems to be unlikely for a majority of borrowers.

  3. Mortgages are a form of collateralized lending (same as auto loans). The value of the underlying collateral matters as much as the ability of the borrower to pay back the loan.

    If you want to buy a house using non-secured credit, try putting one on your credit card with a 15%-20% annual rate.

  4. The mortgage market in the US is more due to history than to logic. The standard 30 year fixed rate mortgage is a risk bomb to the lenders. The lender assumes both an interest rate risk and a duration risk. If rates go up the present value of future payments goes down. Furthermore, as noted home owners who have low rate loans are disincentiveized to sell, so the average duration of the loans goes up further decreasing their present value.

    Further with low down payments (10% is all that is required for standard loans, but anything less than 20% is too low in my estimation) it can take years before the loan has a sensible amount of collateral risk. about 7 to 10 depending on the down payment and the interest rate.

    This is why the only financing party stupid enough to want to extend credit on those terms is the Federal Government through Fannie Mae/Freddie Mac. Both of which have been zombies since the panic of 2008. And, why no one has been able to figure out how to restructure them.

    I can understand why the Swedish Central bank found that its ability to raise rates was severely impaired by the wide spread practice of issuing floating rate mortgages. If rates go from 3% to 6% as they have in the US over the last couple of years, mortgage payments double, and all but the most affluent will be financially choked.

    Assumable loans, if they exist are very rare in the wild. And for good reason, they magnify the duration risk. California conducted a natural experiment on the subject. In 1974, the California Supreme Court declared the standard due on sale clause to be invalid. This meant that all mortgages in California were assumable. This was bonanza for California homeowners and another shot below the waterline for financial institutions. In the soaring interest rate environment of the 1970s (I signed a mortgage commitment at 17% in spring of 1980), it was a disaster. And Congress preempted the law in 1982 (Garn-St. Germain 12 U.S.C. §1701j–3). I can't imagine that the financial industry will ever let the assumable mortgage be a popular product.

    BTW, assumable mortgages are not an unmitigated blessing for selling home owners. If the buyer defaults on the mortgage, the lender can demand that the seller pay it off. For most middle class Americans that would be a financial disaster of unimaginable proportions. You can just imagine the kind of chaos that would cause in a 2008 type event.

    The political problem is simple. The system has pumped an enormous amount of nominal wealth into the hands of middle class homeowners. But, it is a bit of a generational Ponzi scheme. The ratio of house prices to household incomes has soared to over 5 from under 3 where it was in the 50s and 60s. Like all bubbles, this one too will burst. But, do not expect Congress to try to deflate it slowly. Every Federal policy since the 1930s has been based on subsidizing demand and restricting supply and has functioned pumped wealth away from the young to the old. Only a catastrophe will change it.

    An aside, but not irrelevant. The Fed owns $2.576 trillion of Mortgage backed securities, which represents 30% of its total assets. See table 5 of the may 11 edition of "Factors Affecting Reserve Balances - H.4.1" . Footnote 4 says: " Guaranteed by Fannie Mae, Freddie Mac, and Ginnie Mae. The current face value shown is the remaining principal balance of the securities." This means that the holdings are at face value not marked to market. I wonder if the Fed is in better shape than SVB was? Has anyone looked into this more deeply?

    1. I wrote the above comment. I did not realize that it was being posted anonymously. For that I apologize.

      BTW, read what I wrote about assumable mortgages. They are a poison pill and I would not touch them with a ten foot pole.

    2. "BTW, assumable mortgages are not an unmitigated blessing for selling home owners. If the buyer defaults on the mortgage, the lender can demand that the seller pay it off. For most middle class Americans that would be a financial disaster of unimaginable proportions. You can just imagine the kind of chaos that would cause in a 2008 type event."

      That scenario depends on the terms of the assignment agreed by the lender. If the assignment is accompanied by an absolute discharge by the lender of the original mortgagee's obligations under the mortgage contract, then the scenario you postulate is null.

      The lender's risk is no greater than it would be under the original mortgagee's debt servicing capability, provided the assignee's credit rating is as good as or better than the original mortgagee's credit standing.

    3. What you are describing is what is known as a novation. I suppose the lender could contractually agree to do it in advance. I have never seen such a thing in the wild, and the standard Fannie Mae/Freddie Mac documents do not provide for it. They call for payment in full on sale and thus the discharge of the seller. Assumptions like California in the late 70s or in some commercial mortgages do not involve lender consent or novation. In those the buyer agrees with the seller to "assume" the mortgage and to pay it, but the seller is not discharged. Incidentally, even though the seller is personally liable for the note and is liable to reimburse the lender for a deficiency on foreclosure, the buyer has no personal liability to the lender only to the seller, if that is the contract of sale so provides.

    4. "How liability is assigned when a mortgage is transferred. When mortgaged property is transferred liability is assigned in one of two ways depending upon the state. 1.) Mortgagor-grantor retains liability -- in about half the states, the grantee (buyer) recognizes the existence of the mortgage but promises nothing to the mortgagee. Hence, if after transfer a deficiency results from a foreclosure sale, the mortgagor is liable to the mortgagee for this deficiency; the grantee loses, at most, his equity in the property. 2.) Grantee assumes liability -- in the other states, the grantee (the buyer) assumes liability for the mortgage (i.e., he "assumes" the mortgage). Hence, if after transfer a deficiency results from a foreclosure sale, the grantee is liable to the mortgagee for this deficiency; there is a separate agreement between the mortgagor and the grantee (for the benefit of the mortgagee) and they both sign the deed; the mortgagor is liable only for that part of the deficiency not paid by the grantee; because of their agreement, the mortgagor can seek the unpaid portion from the grantee. Sometimes, the mortgagee insists on his approval of the grantee before transfer."

      A Roman square clarifies the points made above.
      "1.) Mortgagor-Grantor retains liability: Mortgagor-Grantor is responsible for the deficiency; Grantee's liability is null.
      2.) Grantee assumes liability: Mortgagor-Grantor is responsible for deficiency not paid by grantee; Grantee is responsibility for the deficiency."

      Source: The Vest-Pocket Real Estate Advisor, Martin J. Miles, Englewood Cliffs, NJ, (c) 1989, Prentice Hall, Inc., pp. 212-13.

      "Novation" is not mentioned, but may be implied when the mortgagee insists on approving the assumption of the liability by the grantee (buyer).

      Fanny-Mae and Freddy-Mac, &c., are issuers of MBS and CMBS instruments It follows from that that they would stipulate a standard contract which requires the mortgagor to terminate (settle) the mortgage loan by accelerating payment of the outstanding balance of the loan when the mortgaged property is sold (title is transferred).

      The reference cited reflects practices in use before securitization became common-place.

    5. Is state law banking regulation that requires certain provisions, or caselaw that covers gaps in the contract? If it's just contract law, then presumably both parties could renegotiate the contract, so the Coasian bargaining would work. A problem might still arise if the lender had sold the mortgage and more than two parties now had legal rights. If it was sold subject to standardization regulation, that would ruin everything, or if it were sold to a government agency that didn't care about profit or was too big to be managed properly.

    6. For what it's worth... The mortgage is a contract between lender and borrower. Assignment, by the lender, of the mortgage to an investor does not alter the mortgage contract. Viz., "8.2 How to assign a mortgage. From the mortgagee's point of view, the mortgage represents an investment (an annuity) that can be sold. If he sells the mortgage, he assigns his rights in the mortgage to the buyer of the mortgage (assignee). The assignee (and new mortgagee) receives exactly the same rights as were held by the previous mortgagee; moreover, the lien priority is unaltered by the assignment. The assignment is usually made through a written instrument; the promissory note must be assigned with the mortgage." What follows this short introduction covers the technical legal requirements that make the assignment effective.

      There is no indication that state law (other than the state's law of contracts) bears on the assignment. The mortgage and the mortgagor's promissory note are unaltered. State law governs (i) the mortgage theory (i.e., Lean Theory, Title Theory, or Intermediate Theory), (ii) the evidence of pledge of security (i.e., Trust Deed, or Mortgage), and, (iii) the usual method of foreclosure (i.e., Strict Foreclosure, Judicial Sale, Power of Sale, or Entry and Possession).

      The assignor assigns his rights to the mortgage (and promissory note) to the assignee, and as such does not retain any rights in the mortgage or the promissory note. It is the rights, and all of the rights, in the mortgage contract that are assigned, under the legal doctrine that an owner of property (e.g., rights in mortgage investment cash flows) has the sole unalienable right to dispose of that property at his absolute discretion. So, the issue of "two parties now had legal rights" in the mortgage does not arise following an assignment of the mortgagee's rights in the contract.

      Since the assignment of a mortgage (and promissory note) is an assignment of rights to property, namely, the investment and upon an event of default, the right to recover from the mortgagor any deficiency through the state courts, including foreclosure and/or sale of the underlying security (the real property), is subject to state law.

      "If it were sold subject to standardization regulation" likely doesn't apply because the assignment agreement is a contract between private parties and does not involve the public interest represented by the state (no alteration of the mortgage contract or the lien on real property arises through assignment).

      "If it were sold to a government agency that didn't care about profit...", is also not likely to be an issue -- government agencies that purchase mortgage contracts for the purpose of securitization of the same have an incentive to not lose money on a transaction because the government agency engaged in securitization is set up as a self-financing entity. If it loses money on a given transaction, it must make up the loss on another transaction, or cease to operate (i.e., management is set out on the street).

      Bottom-line: State law requires that mortgage contracts conform to specific forms (provisions). Case law governs judicial determinations in the event that there is a legal question to be answered (e.g., foreclosure, or perfection of the lien on real property). The two are complementary.

      The brief excerpts above are from "The Vest-Pocket Real Estate Advisor", M. J. Miles, Englewood-Cliffs, NJ, (c) 1989 Prentice-Hall, Inc., p. 199.

  5. My concern will be more on whether first time home buyers and non first time home buyers would get the same benefits if we allowed the old mortgage contract to be assumed. Mortgage assumability helps the non-first time home buyers to get out of their houses, increase housing supply in the market, but it’ll also increase demand. And with record low old mortgage rates, it’s much less costly for non first time home buyers to switch houses than for the first time home buyers to get their first home. So will the increase of supply really lower the house price and improve affordability? I don’t have the answer to this. I do think ARM is a better contract in terms of reducing risk and imposing more negative impact on house price when rates increase. But this will increase the likelihood of defaults and cause another “affordability” concern (somehow the government thinks affordability also means housing market stability)

    1. John — the key is to get Fannie and Freddie (and FHA/VA to a lesser extent) to agree to portable and automatic refinance mortgages. When I was at Fannie (many years ago), there was dust if portability and it’s really not a hard thing to underwrite the borrower rather than the property (as long as the new property meets certain minimum requirements). But can you get FHFA to go along? If they’re interested, it could happen — but if they’re not, nothing will change.

  6. Another approach is to let borrowers prepay their mortgage at the current market value. The Danish mortgage works this way. Effectively, borrowers can repurchase the bonds in the market, at the current market price, and hand them back to the mortgage lender to eliminate the loan.

    Hence, borrowers that got a good deal by taking out a fixed-rate mortgage at a low rate can capitalize the value of this deal, if they need to relocate, and they are not locked in.

  7. I suspect some of the interest rate derivative guys could bundle a low fixed that was transferable. Extra spread to make the hedges work?

  8. "I once was at the Swedish central bank talking about monetary policy. They were worried about raising interest rates. I presumed they were worried that too big to fail banks would have trouble. No, they said. In Sweden practically all mortgages are floating rate. And you can't just mail in the keys and default on mortgages. If you default, they take all your assets and garnish your wages. (So much for soft hearted socialist Scandinavia. They are actually quite attuned to incentives.) The banks were going to be fine. They were worried that if they raised interest rates, people would do anything to pay their higher mortgage rates, and this would tank consumption."

    Swedes and Canadians have that in common, if nothing else.

  9. In The Netherlands you can take your mortgage with you. So a lot of people buy bigger houses and finance it with their cheaper mortgage from a previous house.

  10. An interesting paper published in the Journal of Financial Research presents a technical model of duration of default-free mortgage-backed securities.

    See: "Prepayment Risk and the Duration of Default-free Mortgage-backed Securities", G. A. Anderson, et al., J. Fin. Res., Vol. XVI, No. 1, Spring 1993. File name: jfr_mbs_dur_93.pdf

    Abstract: "The conventional duration measure for mortgage-backed pass-through securities assumes that the prepayment rate is invariant to changes in market interest rates. In this paper the conventional duration is modified to take into account the interest-rate sensitivity of mortgage prepayments. Including interest-rate sensitivity is shown to reduce substantially the duration of a mortgage-backed pass-through security when the current mortgage rate is less than the contract rate."

  11. "What's going on?" The would-be social planner identifies a problem -- housing growing families during a period when the central bank is increasing the cost of money in order to dampen the rate of a general price inflation. The social planner would like to allow growing families that have an existing mortgage contract obligation at a lower interest rate to be able to transfer that obligation to a new replacement housing property more suitable to the needs of the growing family. The impediment is the mortgagor's loan contract with the mortgagee is non-transferrable -- it runs with the land until repaid in full.

    In the WSJ article, the example families have a perceived "space constraint" that they would like to relax by purchasing a larger house and property to accommodate a larger family. The problem they face is that the householders are unwilling to compromise on life-style in order to obtain the larger home they feel they need. Furthermore, they are not willing to buy bare land and arrange the construction of a new house themselves, unlike earlier generations. The would-be social planner's preferred policy is to make the mortgage loan 'transferrable' so as to make the move up in housing 'affordable' for the growing family. The constraint on the would-be social planner is that her preferred policy imposes a negative externality on the mortgagee during a period of rising interest rates. Since a mortgage contract is property to the mortgagee, the would-be social planner's preferred policy would be construed to be in essence an uncompenstated expropriation of property (the mortgagee's interest in the mortgage contract). Per Coase, the would-be social planner would pay the difference in current value between the existing mortgage contract and the mortgage contract that the growing family would have to enter into at the higher current mortgage rate, in order to compensate the mortgagee's loss incurred under the would-be social planner's preferred policy.

    The question raised in the blog article ought to be reframed -- is it socially optimal for growing households to be subsidized by the general taxpayer in order to permit the would-be social planner to implement her preferred policy option described above?

  12. Re: Danish mortgage market. Two papers suffice to describe the Danish mortgage market characteristics:

    First is an IMF paper describing the Danish mortgage market.
    The chief characteristics are (i) highly regulated, (ii) dominated by narrow banks, (iii) mortgage contracts have embedded call and delivery options.

    The mortgage contract runs with the land for the term of the loan and is not 'portable', i.e., the mortgagor cannot take a low interest rate loan with her when she sells her existing house and buys a replacement house. She can exercise the call option and realize a gain on the redemption cost if the current mortgage interest rate is higher than the contract interest rate. She then pays the higher market rate for the new mortgage contract to fund the new house purchase.

    Second is the BIS Quarterly Review, March 2004
    This paper is more technical and provides a comparison between the Danish mortgage market and the U.S. mortgage market. Table 1 in the paper provides the following comparative statistics.
    "Summary statistics for the Danish and US mortgage markets. Data for 2003:(1)"
    "Total volume of mortgage bonds in circulation:(2)
    Denmark 232
    USA 5,129
    "Daily turnover in mortgage bonds(2)
    Denmark 2
    USA 219
    "Total volume of mortgage loans as % of GDP:
    Denmark 101
    USA 81
    "Ratio of households’ debt to disposable income
    Denmark 192
    USA 112
    "Number of residential loan originators:
    Denmark 4
    USA 7,771
    "Share of owner-occupied dwellings:
    Denmark 59
    USA 68

    "(1) For the United States, third quarter.
    (2) In billions of US dollars (exchange rate used: DKK 6 = USD 1)." \end Table 1.

    "A key regulatory difference between the two markets is that Danish
    mortgage banks, unlike their US counterparts, cannot retain prepayment risk. All market risk, including prepayment risk, is passed on to investors in Danish mortgage bonds such as pension funds and commercial banks."

    The BIS report includes charts illustrating the comparative similarities and differences between the two countries.

    The Danish market is characterised by oligopoly (the top 5 mortgage banks share of the mortgage market is 95% according to the IMF report).

    The Danish alternative is effectively a highly concentrated set of specialized or narrow banks issuing MBS and CMBS pass-through bonds to investors to fund mortgage loans to households and firms holding real estate. In the U.S. context, it would be similar to removing all of the commercial banks of asset size below the systemically important U.S. banks (e.g., JPM-Chase, Bank of America, Wells Fargo Bank, etc.) and limiting the origination of mortgage loans to those large banks alone under strict federal supervision, in conjunction with the government-sponsored agencies Fanny Mae, Freddy-Mac, etc. originating the MBS and CMBS bonds to investors.

    This is not what the would-be social planner had in mind ...

  13. The crucial thing might be to let some individual at the relevant agency know. If he can push it through, then he will be the expert, and can leave for private consulting and make a million dollars. Someone has to give him the idea, though. So if any readers know any individuals there, go ahead. It isn't any good just emailing the agency, etc.-- that will be ignored.

  14. I'm a neighbor of yours in Palo Alto. My home is locked up by capital gains taxes. In the late 1990's I paid $700,000 for this four-bedroom single family home which would now sell for $3.6 million, according to Zillow - incurring almost $1 million in capital gains taxes. If I die owning it, my children can sell it tax free. I am being paid $1 million to keep it off the market. I would much rather move to a comfortable condo and let a young family take over the home, but that isn't going to happen in my lifetime. Inflation will only exacerbate this problem.

    Also, the inflation effect on capital gains taxes is particularly cruel to people whose homes haven't appreciated in real terms - for them, a capital gains tax is confiscatory.

  15. While I agree with the general theme of the comments around the rigidity of the mortgage market - blaming it entirely on the federal government is a bit too easy and very wrong. Look at the jumbo market which can't be held by the GSE's? Do you see portable mortgages there or any other great innovations? Not really. Ok - so why not more innovation? Here is an idea - standardization beget economies of scale that are hugely valuable. Most products are streamlined and simplified because that makes them cheaper to produce and easier to sell to mass markets. Imagine even trying to explain to an average home buyer what it means to "take your mortgage with you". Nobody uses ARMs despite the fact that they are always a great deal because they are just too complicated for most people (they are NOT very risky as has been documented endlessly - they are the opposite). Innovation in the space hasn't worked for ARMs and its not clear that it would work for countless other nifty ideas that make sense in principlebut are complex to execute in practice.

  16. I've learned a lot from the comments. Banks and borrowers refinance mortgages. Why can't the borrower and lender modify the existing mortgage contract? the borrower keeps his mortgage paying a slightly higher interest and the bank uses the new dwelling as collateral. Slightly different than a refi but same principle. I believe it's a Coasian arrangement as the transaction costs to both parties are reduced. The borrowing party's cost of carry is reduced and the lender's opportunity cost is reduced. Of course, I could be wrong.

  17. The entire concept of a "30 year fixed loan" is nonsense. It would not exist except for government subsidy. But it lets people by in "cheap" (not really) and they aren't getting rid of it.

    The rest of it is all just trying to dance around that issue.

  18. Note that FHA, VA, and FarmersHome mortgages can be assumed. This gets around the lock in problem that motivates the initial discussion. GSEs do not offer this option for some reason. ARMs are available as an option in the US market but not popular for some reason.

    1. ARMs are not popular for a simple reason: risk avoidance. You might be able to afford a 1500/mo payment, but you will absolutely default if it bumps to 3500/mo.

  19. I'd note that we could mitigate a lot of this if property and homebuilding loans were available in the general space. Currently, you can only get a property/new construct loan if you are working with an approved builder in a community. There's tons of open land for sale, but no one will loan you money for it with less than 50% down. This has decimated the builder space, which now is little more than cookie-cutter multi unit/mcmansion pushers.

  20. Mortgages have two risk problems associated with unpredictable inflation.
    1. Real risk. The real value to bank and borrower keeps changing, creating risk for both.
    2. Liquidity risk. The borrower is liquidity-constrained and has wages whose real value lags inflation, and periods of unemployment. For this, a nominal mortgage might be best.
    So what fancy product might help with this but be better than a nominal mortgage? In practice, bank patience helps with (2). Some kind of analog to preferred stock might be good.

  21. Here's to another year of innovative discussions and uncovering the untapped potential in the world of mortgages!


Comments are welcome. Keep it short, polite, and on topic.

Thanks to a few abusers I am now moderating comments. I welcome thoughtful disagreement. I will block comments with insulting or abusive language. I'm also blocking totally inane comments. Try to make some sense. I am much more likely to allow critical comments if you have the honesty and courage to use your real name.