Friday, October 27, 2023

Prices vs. inflation and a mortgage puzzle

Mickey Levy's excellent WSJ oped leaves some thoughts. 

Inflation has fallen, though I still suspect it may get stuck around 3-4%. But prices "are 18.9% higher than its [their] pre-pandemic level." And some important prices have risen even more. "Rental costs continue to rise in lagged response to the 46.1% surge in home prices."  Those who are taking a victory lap about the end of inflation (the rate of change of prices) are befuddled by continuing consumer (and voter) anger. 

Well, prices are not the same thing as inflation.

Our current monetary policy has, for decades, forgotten past mistakes. If inflation surges to 10%, and prices rise 10%, it is considered a victory, and battle over when inflation gets back to 2%, though prices are still 12% higher than they were. It wasn't always this way. Under the gold standard, prices were stable for long periods, which means that bouts of inflation were quickly followed by bouts of deflation. I'm absolutely not advocating for the gold standard, but it's worth remembering that price stability rather than inflation stability is possible.  

Why the fuss? As Mickey points out, not all prices rose the same amount. In particular, "Increases in wage and salaries ... haven’t kept pace with the CPI and have resulted in a decline in real wages." No wonder "the public isn’t pleased." 

This offers an interesting moment to rethink the basic idea of bygones will be bygones in monetary policy. Wages are, by common agreement, a lot more sticky than prices. Eventually wages will catch up, but it will be a long and contentious process. Wouldn't it be better, after a bout of inflation, for prices to come down quickly,  to match the current level of nominal wages? "Real wage restoration" has a nice ring to it that even Fed doves and inequality worriers might appreciate. 


Mickey also points to a fascinating puzzle of our mortgage markets: 

Tens of millions of homeowners who locked in mortgages at rates below 3% between 2020 and 2021 are now unwilling to sell. The result is a shortage of homes on the market and higher prices. ..That in turn has impaired labor mobility, historically an important factor for production and labor-market efficiencies.

In the US, if you wish to protect yourself against rising mortgage interest rates by buying a fixed rate mortgage, you can only do it bundled with one particular house. You cannot easily say, "I don't want to get hit by interest rate rises, but I might want to move and still be able to afford a big house."  So we are stuck with this interesting puzzle, that higher interest rates to combat inflation lead to people staying parked in houses they really don't want, unwilling to move to take a better job somewhere else, to downsize, to cash out of a house-poor expensive area (Palo Alto), to upsize for more children or elderly parents, and so on. 

30 year fixed rate non-transferrable mortgages with a complex option to prepay and refinance when interest rates go down, and little consequence for default (whew)  are not a law of nature. It is not this way around the world, and I've been fascinated to talk to economists at other central banks about their very different worries. 

In many countries almost all mortgages have floating rates, that quickly catch up to any rise in short term interest rates. In many of these countries, it is not easy to default. If interest rates rise from 2% to 6% and you can't afford to triple your monthly payment, you can't just give the bank the keys as you typically can in the US. In Sweden, I was told,  if you default on a mortgage, the bank will grab all your other assets, and also garnish your wages for several years. You will live on their minimum social assistance income, about $15,000 at the time of this conversation, for several years. 

As a result, central banks in these countries have a completely different set of worries about raising interest rates. Rather than worry about defaults that will imperil banks, they worry that people will stop spending on everything else before defaulting on their mortgage. So monetary policy (raising interest rates), surprisingly ineffective right now in the US, can be dramatically more "effective" with that sort of mortgage market. (Yes, inducing a fall in consumption is the point of raising interest rates.) I don't know of mainstream models that include this distinction but it seems first order for the effect of interest rates on the economy. Central bankers also worry a lot more about public backlash when mortgage costs for the whole population can swiftly double or more. 

Back to the US. Why can you not keep your mortgage when you change houses? Why must protection against interest rate rises -- a form of insurance, really -- be tied to staying in one house? 

Put that way, you can come up with a dozen legal, regulatory, and perhaps even economic reasons. The 30 year fixed rate itself is an invention of 1930s federal housing policy. Banks hold very few mortgages. Pretty much the whole mortgage market gets securitized with a credit guarantee by federal housing agencies (Fannie, Freddy, etc.). So if their rules for acceptable mortgage says you can't change houses, well, you can't change houses, no matter what demand. Subsidies for a particular version of a product kill product innovation. One real estate economist I asked this of suggested that the mortgage originators like it this way, as it forces you to pay fees to move. And one can speculate that lenders don't want you to substitute a worse house as collateral. I don't think that holds, because acceptability of the house follows simple rules, but it's possible. 

So, today's bright idea: Why don't banks also routinely sell retail fixed for floating swaps? These are standard financial contracts that have been around for decades. Here's how it works: You take out a floating rate mortgage, at say 2%. Fixed rate mortgages are, say, 3%. So along with your mortgage, you agree with the bank that you will pay the bank 3% a year, fixed for 30 years, and the bank will pay the floating mortgage rate. That's 2% now, but if interest rates rise, the bank has to pay 5% and you keep paying 3%. Now you can sell the house. When you get a new house, you use the floating rate (5%) to pay the new higher mortgage on the house, while you keep paying 3% out of pocket. We have synthesized a portable mortgage. 

Why doesn't this happen? I await your speculation in the comments. Banks trade fixed for floating swaps among themselves all the time, so laying off the risk is not hard. 

As usual, I am drawn to wonder what tax or regulation is in the way.  I can think of  a few. First, you will get the mortgage interest tax deduction only on the actual mortgage. The actual 3% fixed mortgage lets you deduct the whole 3%. You will pay income tax on fixed for floating payments. This is really an insurance payment, like fire insurance, which shouldn't be taxable, but the IRS will likely treat it as such. Perhaps if insurance companies sold the product they could lobby Washington for rules to extend the tax exemption,  but then you lose some of the efficiency of banks doing what is properly the business of banks. 

Heaven knows how bank regulators and consumer financial protection regulators will do to tangle up a perfectly sensible product. The Fed finally caved in to political pressure to put climate in financial regulation: 

 banks must manage their balance sheets for physical risks from climate change, such as flooding or drought, as well as the “stresses to institutions or sectors arising from the shifts in policy, consumer and business sentiment, or technologies associated with the changes that would be part of a transition to a lower carbon economy.”...

Shifts in policy? OK Citi, what will happen to your balance sheet if a Republican gets elected and cancels electric vehicle mandates? Oh, maybe that's not the "change in policy" banks are supposed to anticipate.  

Banks will also have to conduct a “climate-related scenario analysis”—don’t call them stress tests—that extend “beyond the financial institution’s typical strategic planning horizon” and account for potential losses in “extreme but plausible scenarios.” 

Well, if anvils fall from the sky... But I digress. With this sort of thing coming from regulators, who has the time to create and get approval for a new product that might actually serve homeowners and help to unlock housing supply in places where it's scarce? 

If you can think of other regulatory (or economic) barriers, comment away.  Or maybe, just maybe, we're waiting for a sharp fintech company to figure out that floating + swap is a product consumers would want. 

(Swaps also require counterparties to post collateral, i.e. put in enough cash that the other side can be sure the payment stream keeps going, and in the event of default come out even. On the household side, equity in the house should serve as it does for mortgages; it would be like a second mortgage claim on home equity. That's a reason to bundle retail swaps with the mortgage issuance. One could have banks post collateral too, though enough priority in bankruptcy should do the trick as these are retail contracts. )  

(The madness of refinancing fixed mortgages is also puzzling. I know a lot of very good financial economists, and not one can figure out the optimal time to refinance a fixed rate mortgage. Why not just sell mortgages that adjust down but not up? Sure, you'll pay a few basis points extra, but we save a lot of complexity and the costs of getting that option wrong. The same answers as above may apply.) 


I long ago proposed "health status insurance," which is insurance against health insurance premium increases. If you get sick and your health insurance increases, the premium insurance pays the higher health insurance, or a lump sum so you can do it. 

What I'm suggesting here is also  "mortgage rate rise insurance." Get a floating rate mortgage and separate mortgage rate insurance. If the mortgage rate rises, the insurance pays the higher rate, or a lump sum so you can do it. 

While we're at it, the major difference between buying and renting is that buying allows you to stay in the house no matter if rents go up, where renters bear the risk of higher rents. It's puzzling that you can get long-term commercial leases, but residential leases allow the rent to rise each year. Rent increase insurance could work the same way. Rent for a year, buy rent increase insurance, and if the rent goes up, rent increase insurance pays the difference or pays a lump sum so you can do it. (Renters synthesize this somewhat by renting and then voting in rent control!) 

Update 2:

As many commenters point out, another way to handle the problem is for borrowers to be able to buy their own mortgage back, as is done in Denmark.  


  1. I guess one question I have is what if something happens to the bank and they don’t make the payment for the float rate? It’s very clear in a mortgage that the house itself is the collateral right. Maybe I’m missing something or just not fully understanding

    1. That's why banks sell most of their mortgages to the government and many keep the servicing rights to collect the fees but eliminate the interest rate risk. Banks make most of their money nowadays from fees, mainly servicing fees on loans. Banks are supposed to practice asset/liability management and match their asset (income) rates/maturities with their liability (cost) funding/maturities. However, they keep most of their commercial loans which either float with rate changes or are short-term loans that reprice. A large bank recently stopped making consumer loans and mortgages I think.

  2. Canada has no 30 year fixed, it’s max 5 years or floating, their housing bubble dwarfs by miles what is happening in the US and their inventory is even lower.
    I bought a house with 2.6% fixed 30 years and I know many people who did the same. Why? Mainly as a hedge against inflation. What other options does the worker have to protect themselves from currency dilution? Buying a house is an obvious choice unless you are a Wall Street financier with first access to the newly printed currency. When money isn’t sound, a lot of weird stuff happens as people try to avoid being defrauded. Some will work less, just the minimum, others will gorge in debt, you name it, read Thomas Mann describing how inflation tore apart the fabric of society in pre nazi germany you will recognize today’s environment fully. At some point one had to step back and ask the fundamental questions, what do we need an economy for, is it to serve broader public needs or is it just a vehicle for a few to extract wealth and plunder public treasury. Why is the Fed still currently buying impaired bonds at face value from private interests? The ultimate endgame of inflation is total break down in trust. Your house is the last thing you want to sell in such scenario. For a clue of how high the trust is, witness how quickly costco run out of gold coins when they tried to sell them to the public.

    1. CMHC-insured 25-year term fixed rate mortgage loans are available in Canada. They comprise 0.5% of the total mortgage loans issued in H1-2023.

    2. The loans exist, but very few banks offer them. When I bought in 2020, most banks were reluctant to offer 10 year fixed rate mortgages, I think Scotia and RBC were the only ones that would at the time, and placed a massive rate premiun on that product.

    3. This comment has been removed by the author.

  3. "So, today's bright idea: Why don't banks also routinely sell retail fixed for floating swaps? These are standard financial contracts that have been around for decades."

    Mortgages have been around for longer. You don't see banks reselling mortgages back into the retail market do you?

    "Why doesn't this happen? I await your speculation in the comments. Banks trade fixed for floating swaps among themselves all the time, so laying off the risk is not hard."

    Laying off the risk to who:
    1. The retail investor?
    2. Other banks?
    3. The taxpayer?
    4. The bank shareholders?

    1. Banks may trade fixed for floating swaps among themselves (Company A floating rate for Company B fixed rate), but I highly doubt any liabilities of the bank itself are of the floating rate variety.

      "The Basel Committee has designed two liquidity ratios to ensure that financial institutions have sufficient liquidity to meet their short-term and long-term obligations: LCR and NSFR. These two requirements are intended to reduce risks in case of episodes of financial turbulence."

      "LCR - Liquidity coverage ratio"
      "NSFR - Net stable funding ratio"

      "The LCR is the percentage resulting from dividing the bank’s stock of high-quality assets by the estimated total net cash outflows over a 30 calendar day stress scenario."

      "The NSFR is defined as the ratio between the amount of stable funding available and the amount of stable funding required."

      A bank floating rate liability would NOT be considered stable funding.

    2. Bank deposits are floating rate.

    3. Bank deposits trade at par. $1 of deposit asset (for depositor) = $1 of liability (for bank).

  4. This comment has been removed by the author.

  5. The complexity here is that homeowners own options linked to fixed rate mortgages. We’re mortgages floating rates the call option would have no actual value. Because you’re repurchasing an asset that trades at par at par.

    In your example there is no explicit option. Obviously you could embed a receiver in the swap contract, but it’s an option on the derivative not the actual loan.

    To have the same thing occur under the current system you would have to be able to novate your mortgage to someone else. That would be a regulatory and legal mess. The infrastructure just doesn’t exist for it now.

  6. Something you’re not considering is that collateral for swaps must typically be composed of liquid assets. The issue with home equity is that it wouldn’t be able to be re-hypothecated trivially. In practice most swaps dealers rely on being able to re-hypothecate collateral to post margin on their hedges.

  7. Denmark has something like this with the embedded option resembling more of a straddle than a call option.

    But the main issue remains that government's trying to distort markets even with noble intentions will always have unintended consequences. That includes FNMA and FHLMC pre-GFC and when the Fed decided to buy MBS as part of QE (now proud owners of ~30% agency MBS outstanding)

    Its gross negligence that the Fed was buying MBS in 2021- after HPA was 20% y/y, inflation was greater than 5% and BTC was at 60k.

    1. I believe I've read that in Denmark one can buy back one's mortgage at its actual value. This would be a great solution: if I have a 3% mortgage and prevailing rates are 5%, the value of a 30-year $100 mortgage is much less than $100 (something like $79). As it is now, I have to pay $100 to buy back that mortgage (by selling my house). If I could buy it back for far less, there is no disincentive to refinancing at a higher rate to buy a different house.

    2. This comment has been removed by the author.

  8. The cost of moving from one house to another has become much larger than the past, Namely transfer taxes, real estate fees, inspections, surveys title insurance ,moving transport, etc- now about 10 % of the purchase price.This causes real immobility cost ,outweighing the mortgage cost.

  9. You can keep the mortgage when a way. You keep the ownership of the house you're leaving, rent it out and then rent where you're going. It's a bit convoluted, but apparently it is happening too some extent.

  10. C
    In the UK mortgages are much shorter. The rise in rates shifts demand curve in resulting in a price decline. The 30 year fixed rate reduces the supply of existing housing. The supply effect is strong enough to cause the price to rise.

  11. This seems wrong by at least an order of magnitude: "you'll pay a few basis points extra" for "mortgages that adjust down but not up". Maybe I should write down a model and show you? (I take few to mean 3-5.)

  12. This comment has been removed by the author.

    1. Kaplan post is still up,

  13. Buffered dynamic systems require a level target with a rate and direction of change control function. Scott Sumner supports price level/path targets…

  14. Another way to address the homeowner lock-in problem would be to create a mechanism for mortgage lenders (the “lenders” most frequently are FNMA or FHLMC MBS) and mortgage borrowers to negotiate paying off the loan early. Imagine a low coupon mortgage loan, say a 30-year 3% or 3.5%, that trades in the market at a price of 80 (give or take). The borrower would like to move, but can’t afford to because rates are up 400bps. The owner of the mortgage pool would be unambiguously better off receiving a prepayment of the loan at say 85% of par (the lender could always go out and buy more like-coupon MBS at a price of 80%, locking in an easy 5%). Paying down the existing mortgage at a 15% discount to par would better position the borrower to either “pay-down” the coupon on a new mortgage, or to borrow less at the current higher rate -either way facilitating a more affordable move to a new house despite higher rates. Both sides would be better off, so these transactions should be happening. But there is no mechanism for this to occur. So borrowers remain stuck in the wrong house, and lenders remain stuck in loans well below the market to borrowers who at least in some cases would be happy to pay a modest premium to pay off those loans.

  15. In an era of falling rates, the option to refinance at a lower rate wass probably more valuable in most peoples minds, and through the agency system, pretty good value (eg low cost to the consumer). I dont think you get that option with your proposal (you lock yourself in both ways)? But, you could do something with American swaptions to guarantee refinancing rates. It would be an interesting corollary to the prepayment modelling done in MBS prepayment, given retail don't exercise rationally. That said, retail would have to pay commercial banks rates, and the only folk that are going to make margin are the commercial banks (although they might claim its hedging/capital costs).

    I dont know how common the interest tax deduction is in other countries. To me the obvious question might be why not move/refinance if the IRS picks up the bill?

    In Europe there used to be an argument put forward that high unemployment in Southern Europe was due to an immobile work force. Is this a risk you would see emerging in the US?

    Lots of interesting question around the economic value associated with providing a public good. Is that why you introduced climate scenarios into the blog ;)

  16. The solution you want exists in the Danish mortgage system. Danish mortgages are full pass through, callable, pre-payable etc structures…essentially identical to US agency MBS. The one key difference is that the homeowner can buy back the loan at market price. This is done by purchasing the MBS bond that contains the loan in the secondary market and then handing it to the mortgage bank. If it helps intuition, think of it as a “step in“ whereby the bond now takes the place of the homeowner in providing the payments to the bank. Why is this helpful?

    Example: the 1% 2050 MBS trades at 65 (vs 100 par). You get a 35% reduction if you finance this so-called “buyback“ with a new loan. Or if you sell your home at a lower price you are not stuck with a valuation mismatch vs a home loan that remains payable at par to maturity.

    This is NOT an additional embedded option. It is a secondary market transaction that can be done by the homeowner. It is a great advantage to homeowners, but also to bond investors. Why?

    Because when homeowners do this in size, as has been the case recently in Denmark, it provides support to the bond market. It acts like QE, just that it is not the central bank that is buying. Net issuance in MBS has been negative in Denmark because of these bond buybacks and Danish MBS have tightened in OAS unlike US agencies, which widened substantially in the selloff.

    I am not sure how much tweaking but be required to introduce this feature to the US mortgage market, but can imagine it would have provided quite a benefit in the current environment.

    1. Great point, and it rather opens up/begs the question why cant corporates buy their own bonds back at market price? I am guessing its because it would force the holders to mark to market. The closest most corporates could get would be to buy somebody else's bond and hold as an offset.

    2. This comment has been removed by the author.

    3. "Great point, and it rather opens up/begs the question why can't corporates buy their own bonds back at market price?"

      A few reasons:

      1. Some (not all) debt contracts include a prepayment penalty.

      2. Corporates deduct the interest expense on debt payments from their tax liability. Repurchasing the debt early eliminates that tax benefit.

      3. Companies under Chapter 11 reorganization may be forced to convert existing debt to common equity or re-purchase the debt by selling common equity to the public.

      Corporate decisions are by and large made by the equity ownership of the company. And so those equity holders for obvious reasons like to limit the amount of new equity shares that are sold.

    4. This comment has been removed by the author.

  17. Great post as usual. Since you mentioned "decades" and "forgotten past mistakes", I have a question. If we superimpose a trend of 2.3 percent p.a. to the 2000 price level, there were "undershooting" mistakes and recently we clearly overshot. But we are right where the 2.3 percent trend is. Are you saying MP was "too tight" before the inflation surge of 2021? The possibility of "bouts of inflation were quickly followed by bouts of deflation", is now "bouts of below average inflation were followed by a bout of above average inflation". Is this not good enough to anchor inflation expectations, as opposed to price level expectations? Is deflation desirable in some sense? Are there periods of strong growth in modern economies under recent (post 1980s) monetary regimes accompanied by bouts of deflation? I agree that in the short term, if there are real wage losses, the "let the bygones…" approach is problematic, but that doesn't follow that deflation is desirable, or that periods of deflation are accompanied by real wage growth and/or robust economic growth (are they?). Of course, I am not making any model-based assessment here, I am just asking for examples of strong growth (and real wage growth) and deflation in modern economies in the post-1980 period. Without a model, I still think that the best inflation rate is the one you don’t have to think much (or at all) about it, say 2 percent. And yes, if over a 5–10-year period the central bank is delivering a 2 percent drift on average, then it’s all good. Common sense, one of the least common things these days, beats any model in my book. I have never seen deflation consistently associated with good economic outcomes. Thanks

  18. The whole problem will be solved if a "secondary market" for mortgages develops.

    For instance, banks could create a mortgage that includes the right for the mortgage holder to "buy" the mortgage from the bank at the actual value of the debt. The formula to work out this "actual value" (as a function of the mortgage term and prevailing interest rates) could be included in the mortgage contract.

    That would be equivalent to a "portable" mortgage.

  19. "Why don't banks also routinely sell retail fixed for floating swaps?"

    The problem with that is that the rate for the fixed leg depends on the term of the swap. And the amount owed in the mortgage changes with time (assuming a classical French amortization), so what the mortgage holders really need is a "swap ladder" with different amounts for different terms. The more "ladders" in the "swap ladder" the better the adjustment between the mortgage and the swap payments but the more complex the product.

    It would be easier to include an "obligation for the bank to sell the mortgage to the mortgage holder at fair value" if the property is sold.

  20. On the subject of inflation it's intriguing to see the soft reaction (relatively speaking) of fixed income earners, such as the life insured and bond holders, to the permanent loss imposed by the 2022/3 inflationary shock. As deflation is out of the horizon, this huge Net Present Loss will never be recouped--a subdued response to this shock.

    Coming from Argentina (with inflation running today at 10% per month...), I'm no stranger to witnessing how individuals try to safeguard their assets from inflation (often unsuccessfully, particularly the poor).

    A soft reaction is not the best to control politicians and spenders...

    Alfredo Irigoin, PhD

    1. The reaction was larger than you think if you know where to look:

      Federal interest payments to the public have jumped from $508.5 Billion in the 3rd quarter of 2020 to $981 Billion in 3rd quarter of 2023.

      This is driven by three things.
      1. Rollover of existing federal debt at higher interest rates
      2. End or remittances from the Fed back to Treasury
      3. Unwinding of central bank quantitative easing

      And all three things are still ongoing.

      Presuming that the federal debt ($32 Trillion and counting) all resets to the current Fed funds rate (5.5% or higher) and no new debt is added, total interest expenditures will rise to $1.76 Trillion by 2027 (rollover schedule for Treasury debt is fairly short - about 5-6 years for the entire debt to rollover).

      For context, annualized federal tax receipts amounted to about $2.79 trillion in the 3rd quarter of 2023.

  21. This comment has been removed by the author.

  22. Those 30yr FRMs that were created in the 30's (actually, back then they were 25 years) were created by FHA. FHA insured loans are assumable, so the borrower with a low rate can sell the loan with the house, pocketing the increased value of the loan or transferring it to the down payment on another house.

    If the problem is securitization, why do FHA and GNMA allow assumptions but not the GSEs? Perfectly good question. Anybody? Bueller?...

    Unfortunately, 30+ years of declining rates made people forget this issue exists.

  23. Anyone can buy puts on the TLT ETFs

    US health insurance is guaranteed issue and the high price includes the risks not only of automatic renewal but also people who only buy insurance once sick


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.