# Mortgage Choice and Monetary Policy with John Campbell | Markus Academy | Ep. 111

https://www.youtube.com/watch?v=c9IY2nZEPq0

[00:03] So welcome back everybody to another webinar organized by Princeton for everyone worldwide.
[00:06] We're very happy of John Campbell with us from Harvard University this week.
[00:10] Hi John, good to see you.
[00:13] Hi Marcus, great to be here.
[00:13] Thanks a lot John.
[00:16] John will talk about mortgage choice and monetary policy.
[00:21] What are the implications of different mortgage markets on the monetary policy transmission mechanism?
[00:26] And before we go to John, let me just give you a few opening remarks and summarize also the poll questions are you gratefully filled out.
[00:36] So the first question was, do you have a mortgage?
[00:39] And if so, do you have a fixed rate mortgage or an adjustable rate mortgage?
[00:45] And the answers were, you know, a large fraction of the people don't have a mortgage.
[00:49] 58 have no mortgage.
[00:51] 31 have a fixed grade mortgage and 11 have an adjustable rate mortgage.
[00:59] Of course, John would like to know which country you're in because probably determines that, but we were not able to.
[01:03] Differentiate this at this stage.
[01:07] The second question was, how does a mortgage system with adjustable rate mortgages affect the transmission mechanism of monetary policy?
[01:14] Do you think it's stronger, weaker, no difference?
[01:16] And 78 thought it's stronger.
[01:19] Seven percent thought it's seven is is weaker.
[01:22] And uh fifteen percent thought it makes no difference.
[01:26] The third question is, and that's U.S. focused, does the government securitization activity through the GSE, the government's sponsored Enterprises for conforming fixed rate mortgages, increase the fixed rate mortgage here in the US, yes or no?
[01:40] And the answer was 80-20, so 80 said yes.
[01:46] And finally, in an environment of increasing interest rates, fixed rate mortgages, borrowers paying increased mortgage costs when they try to move, so you lose.
[01:56] Essentially, you have an advantage from a low interest rate mortgage and you give it up if you have to move.
[02:01] And does this lock-in effect make it less, well, significantly more?
[02:05] likely not to move in another location.
[02:08] to take on another job and so forth and.
[02:12] 78 thought yes you know lock-in effect.
[02:14] will actually reduce labor mobility in Gross regions and 28 thought now it will not affect the labor Mobility.
[02:20] so let me give a few thoughts and then we pass on the floor to John.
[02:25] so here if you think about monetary policy if you change the interstate if the central bank is changing the interest rates there's of course a substitution effect and an income effect there's also an endowment effect essentially.
[02:38] but the income effect is really distributive it affects different in a model with heterogeneous agents affects different borders and save us differently.
[02:48] and this has two implications one which is very much focused by the handclap Detroit by the heterogeneous agents new Keynesian models it has an aggregate demand effect.
[02:58] so Savers which have a high marginal propensity to consume are affected differently than save us with our bogus with a low marginal propensity consume.
[03:06] they affect the consumption aggregate.
[03:07] consumption different than Tobin pointed.
[03:09] this out already or clear the reason.
[03:12] some important work my former colleague.
[03:14] Arlene Wong did some important work on this as well affecting some of the aggregate demand management through Interstate changes.
[03:22] but you also have an income effect it also affects the balance sheets of various players through the income effect and that affects then also the risk premium in what the you know price of risk is and how the risk moves around.
[03:36] that affects households and Banks as well and the question is who is affected more the households or the banks and that depends very much on the structure of the mortgage Market as well.
[03:46] so even if they're much prepended to consume is the same across all the uh players in the economy or households in the economy there will be some shifts here as well because some players or some agents in the economy would be more constrained than others and then they might be more essential for certain Financial intermediation.
[04:05] so one way to see how the difference
[04:08] will play out is when the interest rate goes up.
[04:10] if you have adjustable rate mortgages it hits the households in particular indebted households directly.
[04:14] there is less credit demand.
[04:16] if you have fixed rate mortgages and it's given the mortgage are given by the Banks then the banks are hit and then the banks will be under capitalized potentially and they're cut back on credit supplies.
[04:27] so one is much more credit demand uh generally the other one is credit supply.
[04:31] of course the second Channel depends very much how much Market power the banks have on the deposit side and we see this play out dramatically also recently with server with them Silicon Valley Bank and other Banks.
[04:45] the market power on the deposit Market Shields the banks as long as there's no run on these Banks and if they can pay very it policy that is going up but the deposit rate is staying low.
[04:57] finally it also has some very original difference and that's I think what John will allude to in the United States this situation is very asymmetric if you hike the interest rates.
[05:08] it will hit the banks if you lower the interest rate.
[05:10] it fables the households.
[05:12] so I think it's very U.S specific that you know in One Direction it's affecting the banks.
[05:17] in the other direction it's affecting more the households.
[05:21] and my main question I will probably come back to that is if you have a currency area where different countries in the currency area have different structures of the mortgage markets.
[05:33] how do you make monetary policy.
[05:34] it makes monetary policy more challenging and how do you take this into account and can you somehow compensate this if you do some asset purchases in order to balance these two channels that you know some members of your currency area or some countries in your currency area have a different structure mortgage market and you affect the century the transmission mechanism differently.
[05:54] or one could go even so far that one would say that it might not be an optimal currency area if the mortgage markets are different.
[06:01] so with this few opening remarks I'll pass on the floor to John who will know Wayne one has looked at many many data as well and even like that's how this is.
[06:11] I think it's one of the most important elements to understand the monetary transmission mechanism.
[06:14] So John De Flores us and we're looking forward to you know.
[06:18] Okay thank you very much Marcus let me share the screen and hopefully we can get this going nicely.
[06:30] Okay so nicely nicely visible for opening slide.
[06:36] So I want to start by saying how important mortgages are.
[06:41] Uh now the poll results were perhaps a little atypical because almost 60 percent of the people listening to this webinar don't have a mortgage.
[06:49] But that's because this is an unusually wealthy and and elderly audience.
[06:55] Um mortgages are of course the largest household liability in the US more than half the debt of a typical household and that's also true in most other developed countries.
[07:06] Um less so in countries like Germany where there's a low home ownership rate but in most places mortgages are the big.
[07:13] liability that a household has.
[07:16] mortgage rates of course as Marcus mentioned are also the main direct channel through which monetary policy affects household consumption.
[07:25] very obviously mortgage rates have a strong impact on on the rate of Home Building in the construction industry.
[07:32] and problems with mortgage lending were at the heart of the global financial crisis and are again very Salient as we see U.S Regional banks are failing in part because of their large mortgage portfolios.
[07:45] now in the U.S what's what's the U.S mortgage system uh those of us who are based here know it very well at the standard mortgage is a 30-year amortizing nominal fixed rate mortgage or frm firm for short with a refinancing option.
[08:04] when rates fall you can refinance when you do that that's treated as a new loan origination so you have to go through a check for your income credit score of
[08:14] home equity Etc.
[08:16] adjustable rate mortgages or arms for short are also available in the US but
[08:19] the market share of arms in this country is typically low uh that with with occasional spikes.
[08:28] now another feature of the mortgage system that I will emphasize is the so-called points system whereby borrowers can borrow a little extra to cover the closing costs the transactions costs of buying a house without affecting their mortgage balance by paying a higher mortgage rate.
[08:47] then a final important point is that U.S mortgages are typically not assumable if you move you can't pass your mortgage on to the new home buyer.
[08:54] mortgages are also not portable you can't take your mortgage with you.
[08:58] now there is an exception which is that certain government-backed mortgages um are in fact assumable but that's that's a minority that's only going to be about 10 of the U.S mortgage Market.
[09:13] now I want to emphasize that the U.S
[09:15] system is an outlier internationally it is very unusual.
[09:20] so there are of course many countries that rely basically on on adjustable rate mortgages where the rate is fixed for an initial period that might be one year might be two years it won't be more than five years.
[09:34] um Germany has fixed rate mortgages but uh without a refinancing option and the closest international parallel to the US is probably Denmark which is a country I've studied in some detail.
[09:46] Denmark has both arms and firms and the firms have a refinancing option but there are some interesting design features of their system for example in Denmark the option to refinance is an absolute right to do so so long as you're not extracting Equity so it doesn't depend on your income your credit score or your home equity when you refinance you can do so either at market value or face value whichever is more advantageous to the borrower.
[10:13] mortgages are typically assumable there.
[10:16] are no points and then they they fund it differently and the funding system relies on covered bonds so I'll come back to that uh as we go through.
[10:25] now what I want to do um is be a little bit unfair and I want to present a prosecution case against the U.S mortgage system.
[10:34] um it is conventional in this country for financiers and politicians to boast about how how wonderful our mortgage system is one of the wonders of the world etc etc but I want to present the prosecution case.
[10:47] so I'm going to um emphasize four problems that the U.S system has.
[10:54] the first is that the mortgage channel of monetary transmission is weak.
[10:59] so it makes it harder life gets harder for the Fed.
[11:04] the second is that U.S refinancing rules um worsen inequality because lower income and less sophisticated borrowers are less able and and less aware that they need to refinance and they end up
[11:17] paying higher rates.
[11:19] The third problem is that fixed rate mortgages have long duration.
[11:24] They're long-term bonds, but also that duration is very variable because of refinancing behavior, and that can destabilize the financial system.
[11:33] Which in turn can constrain monetary policy.
[11:36] Finally, I want to show you evidence of locking effects that when rates go up.
[11:42] Um, people with old mortgages will have to give up those mortgages if they move.
[11:46] That can reduce housing market liquidity and labor mobility.
[11:51] And I think these are all four really quite important disadvantages of the current U.S. system.
[11:58] Will you mention some advantages too?
[12:01] Well, I I mean certainly uh uh towards the end I want to talk about what borrowers like.
[12:08] And, you know, there's a sense in which uh uh when when you look across countries, uh borrowers tend to like whatever they're used to, but certainly uh.
[12:18] borrowers in this country are very comfortable with fixed rate mortgages uh.
[12:22] households like the uh insurance that they're buying against increases in their interest rates and their monthly payments.
[12:30] um and so that's the point I will I will get to at the end for most of the talk.
[12:34] the beginning part of the talk I'll just sort of hovering a helicopter over the system and talk about the systemic aspects of it but then we do need to come back down to ground level at the end and ask what people are actually interested in buying.
[12:49] um so I'll go through these four problems in turn uh then I'll talk about some of the evidence on on borrower preferences and then conclude with some policy suggestions.
[12:59] okay so first of all the mortgage channel of monetary transmission.
[13:04] um Marcus gave a really nice introduction to that.
[13:08] um this channel is not about inter-temporal substitution it's not the uh new Keynesian IR equation um you know that that that we talk about.
[13:18] when we when we write down new Keynesian models it's about redistribution across agents.
[13:23] uh a point made Years Ago by Tobin and much more recently by Adrian Eau Claire.
[13:29] so you know when the mortgage rate moves that affects monthly payments by borrowers but it also affects payments received by lenders.
[13:37] and there's only going to be an effect on aggregate demand if borrowers change their spending more than lenders do.
[13:43] and there's two arguments why that might be the case.
[13:46] one is borrowers are domestic residence perhaps some lenders are foreigners and you know we don't care so much about them and they spend on foreign Goods rather than domestic Goods.
[13:57] the second argument is that even in a closed economy where borrowers and lenders are all U.S residents.
[14:05] The Borrowers may have a high marginal propensity to consume or MPC because they're borrowing constrained while lenders have a low MPC because they are unconstrained permanent income consumers.
[14:17] now it's important to note that the
[14:20] second argument only works if mortgage payment changes are temporary.
[14:24] because if they're permanent then lenders will adjust their consumption one for one just the way borrowers will and they'll be a perfect offset.
[14:35] so both Theory and evidence uh suggest that the mortgage channel is stronger it works better for adjustable rate mortgages than for fixed rate mortgages.
[14:47] and I'll cite this paper by DiMaggio Odell for some of the empirical evidence on that and by the way at the end of the slide deck I've got references which one can look up afterwards if you're interested.
[15:01] so the the three Arguments for why this is the case why the mortgage channel is stronger for arms uh on the slide number one.
[15:10] arm payments move with the shot rate.
[15:12] firm payments move with a long-term mortgage rate which typically moves less because long yields typically move less than one for one with short rates.
[15:21] Second point is that arm payments change for all borrowers.
[15:23] Everybody who has an arm when the interest rate goes up or down their payments change.
[15:29] But if you have a fixed rate mortgage, the payments are only going to change for new borrowers who are who are taking out new mortgages and on the downside, people who refinance.
[15:41] And then the third point is that the change in on payments is temporary because it's driven by the short rate.
[15:47] The thread cuts the short rate in a recession temporarily or raises it to control inflation temporarily.
[15:56] But so if if you've got an arm, it's a temporary change in your payment.
[16:01] But if you're reacting to a change in the fixed mortgage rate, that's a very persistent thing because it's a long-term bond yield basically.
[16:09] So it moves very close to a random walk and that means that fixed rate lenders will adjust their consumption more and that will offset the effect on borrowers.
[16:22] okay so I've argued the arms you know.
[16:24] arms are a good thing uh if you want a strong uh monetary transmission Channel.
[16:32] s are still doing these short-term lending and then there's a financial intermediate in between those and the short-term lending.
[16:39] would this change your conclusion for the last bullet point in the previous slide?
[16:44] well I think um the the point I would make is that um um.
[16:51] if we look at this this last Point um we're trying to get an effect on aggregate demand.
[16:55] we need to have a difference between the consumption response of the lender the ultimate lenders and the borrowers.
[17:05] that can be of course a chain of Financial intermediaries in between.
[17:09] but there's got to be some ultimate difference in the consumption response of those who are winning and those who are losing from the change in rates.
[17:18] and if you've got a permanent change in rates um it doesn't matter who's borrowing.
[17:22] constrained and who's not everybody's changing their consumption one for one.
[17:26] and you get a perfect offset.
[17:28] or as that need not be the case won't in general be the case if you have a temporary change in rates.
[17:37] okay now can we can we even do better than arms.
[17:40] um well there are times when the Central Bank may want an even stronger mortgage channel than you can get with an arm.
[17:47] I mean one example is when the short rate is close to the zero lower bound as was the case in the covid uh pandemic.
[17:55] uh the FED could lower the rate up to a point but it ran out of room to lower the rate.
[18:00] so one thing you can do then is to um have a temporary period of forbearance where people don't have to make mortgage payments.
[18:09] and of course this was done exposed in the pandemic.
[18:14] but it should really be done X anti in a way that can be priced in and and understood in advance by both borrowers and lenders.
[18:18] so I I've written a paper.
[18:22] uh with uh Nuno Clara Joel Coco which came out in the Journal of finance a couple of years ago and we we studied this possibility using a structural life cycle model where we were looking not only at the consumption side of things but also the default side of things because of course you the the Central Bank also wants to minimize uh mortgage default or any rate preventer kind of default crisis of the salt that we had uh in 2008 2009.
[18:53] so I'll just show you very briefly a couple of figures from from that article um the different colored let's look at the left-hand uh bar chart where it says cyclicality of consumption growth um you want these bars to be you want you want to find a system that has a low bar because then cyclical consumption is not very cyclical uh it's smooth over the business cycle and the different colors correspond to different mortgage systems the first one which is sort of uh mid color blue on the left that's an arm the
[19:26] pink one uh the third one from the left
[19:28] that's a fixed rate mortgage system so
[19:31] you can see the arm bar is a little
[19:33] lower that's great
[19:34] but then the light blue bar which is
[19:36] second from the left that's the arm with
[19:39] a forbearance provision which is
[19:41] automatically triggered in a recession
[19:42] and that's better that that reduces the
[19:45] cyclicality of consumption
[19:47] the right hand side panel here the right
[19:50] figure is the cyclicality of the default
[19:53] rig
[19:54] and this is the this is measuring the
[19:57] extent to which uh defaults happen in
[19:59] recessions again you want a low bar you
[20:02] don't want to have a lot of defaults in
[20:04] recessions
[20:05] um again the the arm mid-colored blue
[20:09] bar is lower than the pink uh fixed rate
[20:12] mortgage bar because lower rates in
[20:14] recessions help Stave off default
[20:17] but you can do much better if you have
[20:20] forbearance in the recession because
[20:22] then people are not going to default in
[20:23] the recession they're not having to make
[20:25] any payments and in fact the the the
[20:28] defaults tend to happen in when you're
[20:31] when you're outside the recession when
[20:33] you have to start making your mortgage
[20:35] payments again and that's why the light
[20:36] blue bar is actually negative in that
[20:39] figure
[20:40] that's the default of the households
[20:42] does it also include the difference the
[20:45] fault of the households which of course
[20:47] then passes through to the uh to the uh
[20:51] mortgage lenders who might or might not
[20:53] be Banks we'll we'll we'll get to bank
[20:55] balance sheets in a little bit okay
[20:56] that's obviously an important issue
[21:00] um
[21:01] so then I can look at the cost you know
[21:03] you don't want to cook up some system
[21:06] that is so unattractive to lenders that
[21:08] the rates will be very high
[21:10] um the left but the left uh bar chart on
[21:13] this slide is showing you the loan
[21:15] premium the spread over the
[21:18] um equivalent treasury rate that is
[21:20] charged by lenders and basically arms
[21:24] firms and the and the forbearance arms
[21:27] all have fairly similar costs the
[21:30] forbearance arm is actually even a
[21:31] little cheaper because lenders like the
[21:33] fact that defaults don't happen in your
[21:35] sessions
[21:37] um and then the right uh bar chart here
[21:39] is showing welfare gains where you you
[21:41] look at the the welfare of borrowers and
[21:44] lenders um and
[21:46] um the light blue bar at the left shows
[21:49] that um
[21:51] this forbearance provision is is welfare
[21:54] uh welfare enhancing so I think there's
[21:56] a case for doing this which we did again
[21:58] we did an ex post in the covert pandemic
[22:01] but it would be better to set this kind
[22:03] of system up in advance
[22:05] just to understand that the the welfare
[22:07] gains that's always relative to the
[22:10] Baseline the Baseline that's relative to
[22:13] the Baseline Arm System and actually
[22:16] this illustration is relative to
[22:18] Baseline autumns or is it yes in this in
[22:21] this paper the arm is the Baseline so
[22:24] um the the positive pink bar in the in
[22:28] this right panel um is showing you that
[22:30] actually borrowers they do like fixed
[22:33] rate mortgages
[22:34] but they like even more unarmed with
[22:37] this extra feature that in a recession
[22:39] they'll be given some forbearance
[22:44] okay so that's the mortgage channel of
[22:46] monetary transmission let's go on and
[22:47] talk about
[22:48] um inequality and I think this is uh
[22:50] really really an important before you
[22:52] jump to that would you also say that the
[22:55] interest rate movements of the FED is
[22:57] hence more pronounced because we have
[22:59] this mortgage system
[23:02] yes it may well be that the FED then has
[23:05] to move the rate more to achieving its
[23:07] goals more and in a more persistent
[23:09] Manner and of course we've just emerged
[23:12] from a period when the Fed was relying
[23:15] very heavily on uh forward guidance uh
[23:18] which is now
[23:20] in a way come back to bite the FED
[23:22] because the FED made strong statements
[23:24] about what it would do and has now done
[23:26] something else because circumstances
[23:27] have changed
[23:29] one reason why the FED had needed so
[23:32] much forward guidance is they were
[23:33] needing to try to move longer term
[23:36] rates longer term mortgage rates
[23:40] okay so that's that that's a great
[23:42] question Marcus thank you okay
[23:45] um so let's talk about inequality
[23:47] refinancing and inequality
[23:49] so let's ask who refinances when rates
[23:52] go down who refinances
[23:54] well in the U.S refinancing requires
[23:57] positive home equity and an adequate
[23:59] income and credit score
[24:02] so what does that mean when the Fed
[24:04] faces a recession and cuts interest
[24:06] rates the impact is weakest
[24:08] in regions with depressed home prices
[24:10] and high levels of unemployment
[24:13] so the mortgage channel of monetary
[24:15] transmission is weakest exactly where we
[24:17] want it to be the strongest
[24:19] and I think that's an important
[24:22] objection to this type of system
[24:25] the second point is that refinancing
[24:27] their even among people who can
[24:29] refinance the extent of refinancing
[24:32] varies with the sophistication of
[24:34] Borrowers
[24:35] now we can measure this effect very
[24:37] precisely in Denmark because in Denmark
[24:39] everybody has an absolute right to
[24:41] refinance so if people don't do it you
[24:44] know it's that they forgot to do it or
[24:46] didn't know they could
[24:48] but the effect is also important in the
[24:50] US and um or some recent work by gerardi
[24:54] Willen and Jang suggest that it helps to
[24:56] explain troubling racial differences in
[24:59] mortgage rates paid by black and white
[25:02] borrowers uh in the US so let me show
[25:04] you a couple of figures
[25:07] this figure shows something called
[25:09] refinancing efficiency using Danish data
[25:12] this is taken from
[25:13] um
[25:15] uh it was a paper that I've uh published
[25:20] um with a team of co-authors in the AER
[25:23] in 2020 based on the Danish mortgage
[25:26] system and the figure shows something
[25:29] called refinancing efficiency which is
[25:32] basically in in our sample period which
[25:34] is a declining interest rate period
[25:37] how much money did you save by
[25:38] refinancing and what is that as a
[25:41] fraction of what you would have saved if
[25:43] you'd refinanced optimally
[25:45] following a particular academic model of
[25:49] optimal refinancing proposed by agawal
[25:51] Driscoll and ladies so The Benchmark the
[25:54] 100 Benchmark is following the advice of
[25:57] academic economist
[26:00] and people don't do as well as that but
[26:02] on average they they in this period got
[26:05] something like two-thirds or seventy
[26:07] percent of what they could have got
[26:10] but what this figure shows is that when
[26:12] you sort households by their Rank and
[26:15] you divide them into quintiles
[26:18] on different characteristics age
[26:20] education income Financial wealth and
[26:23] housing wealth
[26:25] the pattern you see is that refinancing
[26:27] efficiency is higher for people with
[26:30] more financial wealth uh bigger houses
[26:33] more education and more income in other
[26:36] words wealthier and more sophisticated
[26:38] people
[26:40] and it's lower for older people so
[26:44] um that means that there's a big
[26:47] difference some people are leaving a lot
[26:48] of money on the table and those people
[26:50] tend to be people with lower education
[26:53] and socioeconomic status
[26:56] how do you explain the age phenomenon
[26:59] that all the people use it because I
[27:01] think it correlates with sophistication
[27:02] there's a little bit of a hump shape um
[27:05] in age the the age line is the one with
[27:08] x's and it's a bit of a hump
[27:10] um we often see a hump uh in um in age
[27:14] in in household Finance
[27:16] um
[27:17] there's a a very nice paper by agawal
[27:21] Driscoll quebecin labson uh called The
[27:24] Age of Reason which points out
[27:26] middle-aged people tend to do best and
[27:29] and they argue that that's because the
[27:31] young are inexperienced and then older
[27:34] people are in cognitive decline so
[27:37] um
[27:37] the youngest people and senior citizens
[27:40] tend to be doing worse
[27:43] um and I think you know this pattern is
[27:45] roughly consistent with that
[27:49] now
[27:51] I here I want to show you a figure based
[27:53] on uh on on on us data this is taken
[27:56] from the Girardi William and Jang paper
[27:58] and
[28:00] look at the uh the red line and the blue
[28:03] line
[28:04] um these are interest rate differences
[28:06] on mortgages paid by black borrowers and
[28:10] white borrowers in the U.S
[28:12] the red line is the difference in the
[28:14] rate paid on newly originated mortgages
[28:17] so it's about 15 basis points on average
[28:20] and that can be explained basically
[28:22] fully by
[28:23] um
[28:24] characteristics that are relevant for
[28:26] mortgage pricing so
[28:28] Federal Regulation is pretty good at
[28:31] preventing racial discrimination in
[28:33] mortgage lending today and so that's the
[28:36] red line
[28:38] but what's the blue line the blue line
[28:40] is the difference between the rates paid
[28:42] by black and white borrowers on the
[28:44] mortgages they actually have at each
[28:46] point in time in other words instead of
[28:48] the flow of new mortgages which is the
[28:50] red the blue is the stock of outstanding
[28:52] mortgages
[28:54] and you can see that that blue line is
[28:57] higher on average and particularly High
[28:59] towards the right of the figure which is
[29:02] the early 2010s and in general the Blue
[29:06] Line tends to go up whenever rates fall
[29:08] so the the dashed black line at the
[29:10] bottom of the figure is the um a measure
[29:13] of the the the underlying
[29:17] Shadow fed funds rate it it can go it
[29:19] can go negative it's it's it's
[29:21] calculated to avoid the zero lower bound
[29:24] uh so it's a measure of sort of where
[29:26] interest rates are and when interest
[29:28] rates fall that's when you want to
[29:30] refinance
[29:32] um but what happens is white borrowers
[29:34] do that more than black borrowers and so
[29:36] this blue spread widens and in the early
[29:41] 2010s it was 50 or 60 basis points which
[29:45] is really enough to be a very troubling
[29:47] finding
[29:50] okay so because there's one question by
[29:54] George she also used the Danish as a
[29:56] like a sample there's also an FHA and VA
[29:59] mortgages where they supposedly
[30:01] refinancing is very streamlined have we
[30:03] looked at these data as well just to
[30:05] contrast it was Danish I haven't I
[30:07] haven't looked at those data but there
[30:08] are um studies of U.S uh refinancing
[30:12] um decisions for example there's paper
[30:14] by Keys proprint Pope um where they look
[30:17] at a sample of pre-approved uh mortgage
[30:20] refinance office offers and who takes it
[30:22] up and who doesn't and you find similar
[30:25] patterns
[30:27] so in equilibrium what is the effect of
[30:30] this well it's
[30:32] a cross subsidy from the poor to the
[30:35] rich
[30:36] okay so why is that well in a
[30:38] competitive market the extra Revenue
[30:40] that mortgage lenders get from
[30:42] non-refinancers who are paying more than
[30:44] they need to that extra revenue is
[30:46] partly passed on through the competition
[30:49] for borrowers in the form of lower
[30:51] upfront mortgage rates
[30:53] and that implies that sophisticated
[30:55] refinances all of us on this uh webinar
[30:58] we get a cross-subsidy by pooling with
[31:01] non-refinances
[31:03] this is an example of what my colleagues
[31:05] uh quebecs and labson have called
[31:07] shrouded equilibrium
[31:10] um I don't know if there are any
[31:11] um friends of of Monty Python on this
[31:15] webinar but I grew up watching Monty
[31:18] Python and I think of it as a Dennis
[31:20] Moore equilibrium
[31:21] and I'll leave the reference to resonate
[31:24] with those who remember it
[31:27] um but but basically we're we're taking
[31:29] from the poor and giving to the rich
[31:31] we're taking from less sophisticated
[31:32] people and benefiting more sophisticated
[31:34] people
[31:36] what it also means is is that it's
[31:38] harder for innovators to introduce new
[31:40] easier to manage mortgages I mean
[31:42] imagine
[31:43] coming in with an automatically
[31:46] refinancing mortgage which could save a
[31:48] lot of transactions costs and be very
[31:50] easy to use
[31:52] well that's going to be expensive for
[31:54] anybody sophisticated because if you
[31:56] take out that mortgage you're going to
[31:57] lose the benefit of the Cross subsidy
[32:00] but it's the sophisticated people who
[32:02] are the natural early adopters for a new
[32:04] product unsophisticated borrowers aren't
[32:07] going to understand the new product or
[32:09] know why they need it so
[32:11] um you know this we can get stuck in an
[32:13] inferior equilibrium so John you know
[32:16] that the divorce that is doing a lot of
[32:18] these prepayment models one could
[32:20] imagine that uh some some mortgages are
[32:23] sold at cheaper interest rate to the
[32:26] less sophisticated households so if I
[32:29] know this is a less sophistic household
[32:30] they were not three Finance I might give
[32:31] him a lower mortgage rate do you see
[32:33] that or well that this is the strange
[32:36] thing and I think that there should be
[32:37] more research on this we don't actually
[32:39] see
[32:40] um separation of
[32:43] um mortgages into Pools by
[32:46] sophistication or or race for that
[32:48] matter
[32:49] um and therefore we don't we don't see
[32:53] unsophisticated borrowers benefiting in
[32:56] the form of lower rates up front there's
[32:58] a pooling equilibrium now it is a little
[33:00] bit of a mystery why that is
[33:02] um
[33:03] but that's what we seem to see in the
[33:05] data but the Innovation could happen
[33:06] there that it could it could happen
[33:09] there that's that's a very very good
[33:11] point
[33:13] um now I want to focus for a minute on
[33:15] another aspect of this which is I think
[33:18] understudied and that's the fact that we
[33:21] have these mortgage points and you know
[33:23] what are points um if you haven't sort
[33:26] of taken out a mortgage recently you you
[33:28] may not know about them um
[33:30] so what I've put on the slide is a table
[33:33] of
[33:34] um a rate table so
[33:37] the rates are shown in the left hand
[33:40] column
[33:41] and then
[33:42] um the numbers in the body of the table
[33:45] are the
[33:47] um closing costs that the borrower has
[33:49] to pay to the lender as a percentage of
[33:51] the mortgage principal
[33:53] they can be negative uh which is true
[33:55] down at the bottom of the table and that
[33:57] means that the borrower is
[33:59] taking a little bit of more money from
[34:02] uh the lender to cover other costs of of
[34:05] doing the transaction
[34:07] and essentially what we see in the table
[34:09] is that if you borrow more in other
[34:11] words if you go down to these negative
[34:13] numbers in the body of the table you can
[34:15] do that and pay a higher rate the
[34:18] numbers in the left hand column get
[34:20] bigger as you go down
[34:22] so
[34:23] that's the system and
[34:26] I want to emphasize this is weird this
[34:29] is completely weird
[34:30] why because in textbook Finance the
[34:33] interest rate is a single number that's
[34:35] given you know you get it from the bond
[34:36] market it is what it is
[34:38] if you borrow more money that increases
[34:40] the size of your debt but it doesn't
[34:42] change the interest rate
[34:44] in the point system when you borrow
[34:46] extra money to cover these closing costs
[34:48] it leaves the face value of the mortgage
[34:51] the size of your debt unchanged but
[34:53] increases the interest rate
[34:55] well now you might say look it doesn't
[34:57] matter either way the required monthly
[34:59] payment goes up
[35:01] you're just multiplying a rate and a
[35:03] principal balance doesn't matter where
[35:05] you put the numbers
[35:08] um but there is a difference because
[35:09] you're allowed to refinance so if you if
[35:12] you if you take points if you borrow
[35:14] extra promise to pay a higher rate you
[35:17] know you may not have to pay that higher
[35:19] rate for very long because you'll be
[35:20] able to refinance out and walk away with
[35:23] the extra money that you borrowed
[35:27] now the paper by David Zhang uh former
[35:30] Harvard PhD student now teaching at Rice
[35:33] explores the effects of this system and
[35:35] what he shows is that many of the
[35:37] borrowers who take points to cover
[35:39] closing costs are sluggish refinances
[35:42] so they pay they promise to pay this
[35:45] higher rate and then it would be in
[35:46] their interest to refinance but they
[35:48] don't do it
[35:49] that actually lowers the cost of taking
[35:51] points for prompt refinances
[35:54] so it worsens the cross-subsidy problem
[35:57] and he he calculates that in an
[35:59] equilibrium with points calibrated to
[36:02] the 2010s
[36:04] there's a transfer with an average
[36:06] present value of about ten thousand
[36:08] dollars for each mortgage from
[36:09] non-refinances to prompt refinances
[36:12] and furthermore
[36:14] because in this system The Savvy thing
[36:16] to do is take points and then refinance
[36:19] uh that's what people on this webinar
[36:21] will probably do you'll get extra
[36:24] refinancing that otherwise wouldn't have
[36:26] happened
[36:27] the problem with that is there's real
[36:28] resource costs from refinancing about he
[36:32] estimates about fourteen hundred dollars
[36:34] for which
[36:35] uh
[36:36] uh so so for each new mortgage
[36:40] origination at the time of a house
[36:42] purchase
[36:43] so this is uh it worsens inequality and
[36:46] it's actually an inefficient system
[36:49] now finally I've mentioned that this
[36:52] problem uh is not just a fixed rate
[36:54] knowledge problem it is in this country
[36:56] it's a fixed rate mortgage problem but
[36:58] you can have similar problems in arm
[37:00] systems too so in the UK and many other
[37:03] places including Canada
[37:05] arms are marketed with teaser rates that
[37:08] that you get a low rate up front and
[37:10] then
[37:11] after one to five years the rate jumps
[37:14] up to a much higher standard rate
[37:16] well what what do sophisticated people
[37:18] do of course they refinance jumping from
[37:21] teaser rate to teaser rate and that
[37:23] leaves the unsophisticated people to pay
[37:25] the high rate
[37:26] and again you get this transfer from the
[37:29] poor to the rich or from the
[37:31] unsophisticated to the sophisticated
[37:34] that problem was pointed out almost 20
[37:36] years ago in a report uh on the UK
[37:39] mortgage system by David Miles
[37:42] uh and a recent paper
[37:45] um looking at UK data documents that the
[37:48] transfers continue and they continue to
[37:51] go from poorer to richer mortgage
[37:53] Borrowers
[37:55] so John can ask you a quick question so
[37:57] it could also be that the less
[37:59] sophisticated people are overconfident
[38:01] how sophisticated they are
[38:03] in a sense has anybody done some survey
[38:07] where the behavioral bias is coming in
[38:09] uh or is it like they're overcompetent
[38:12] how quickly they have to move again and
[38:14] re-optimize anyway what drives I agree
[38:18] with you that I think I think sort of uh
[38:21] surveys of mortgage borrowers to figure
[38:23] out what they know what they're thinking
[38:25] about what they anticipate would be very
[38:27] worthwhile I I haven't seen much
[38:29] evidence on that
[38:31] um but I also agree with you that
[38:37] um
[38:38] there's a general tendency in household
[38:40] Finance for people to think that
[38:41] whatever's familiar is safe and
[38:43] straightforward and very often what's
[38:45] familiar is you know risky and
[38:49] surprisingly hard to manage so you know
[38:51] the stock of your employer is a very
[38:53] risky retirement investment even though
[38:55] it's familiar and uh plain vanilla U.S
[38:58] fixed rate mortgage with points is very
[39:01] familiar but also surprisingly hard to
[39:04] manage and people I think don't realize
[39:06] that and in that sense they're
[39:07] overconfident
[39:09] so I forgot to you know there's some
[39:11] questions in in the q r a some people
[39:14] would like to know you mentioned the
[39:16] African-Americans versus whites did you
[39:19] also look at Asians versus whites in the
[39:21] refinance
[39:23] well I'll have to look back at the paper
[39:26] um the main focus is on uh black and
[39:29] white boroughs I think they also look at
[39:31] Hispanic borrowers and I don't remember
[39:33] whether they have evidence on Asian
[39:35] American Borrowers
[39:38] okay
[39:40] um
[39:41] well now let me just quickly talk about
[39:44] um two more problems let's talk about
[39:45] fixed rate mortgages and financial
[39:47] stability
[39:49] um obviously very Salient topic right
[39:51] now
[39:52] so you know fixed rate mortgages are
[39:55] long-term fixed income securities and so
[39:57] they create maturity mismatch when
[39:59] they're held by banks that have deposit
[40:01] financed and that was the undoing of the
[40:04] Savings and Loan industry back in the
[40:06] early 1980s
[40:07] now of course since then we've developed
[40:09] a securitization system which is
[40:11] intended to solve this problem by
[40:13] passing mortgage interest rate risk on
[40:15] to MBS investors
[40:17] but I think what we've learned over the
[40:19] years is that that system works
[40:21] imperfectly partly because
[40:23] the origination process requires Banks
[40:25] to hold mortgages while they're in the
[40:27] pipeline
[40:28] but also because some banks like to buy
[40:31] MBS to earn a term spread you know
[40:33] that's what First Republic was doing in
[40:35] many cases
[40:38] um now arms don't create this problem of
[40:40] course and that's one reason why they're
[40:42] common in countries with deposit
[40:44] financed mortgage origination there are
[40:47] other Solutions in Denmark they use
[40:49] covered bonds for example but
[40:52] um this this financial stability issue
[40:55] is I think pretty important in in this
[40:57] country
[40:58] now it's a little more subtle than that
[41:01] because we because we have refinance it
[41:03] right so
[41:04] that re that refinancing speed
[41:08] um is somewhat random there's there's
[41:10] variation in refinancing rates that's
[41:12] unrelated to the level of interest rates
[41:15] this creates prepayment risk
[41:18] which cannot be hedged in treasury
[41:19] markets and is priced by MBS investors
[41:23] and is a major topic of sophisticated
[41:26] quantitative modeling on Wall Street so
[41:28] it's ironic that the least sophisticated
[41:31] participants in the financial system
[41:32] home borrowers mortgage borrowers by
[41:36] their random Behavior create a source of
[41:38] risk that then is a headache for the
[41:40] most sophisticated player
[41:43] um
[41:43] now of course beyond that interest rate
[41:45] movements also alter the duration of MBS
[41:48] and this creates shocks to the supply of
[41:51] duration that destabilize Bond markets
[41:53] and Banks
[41:54] so it's not just that firms have long
[41:56] duration it's that they have variable
[41:58] duration
[42:00] um so you know what happens when
[42:03] interest rates go up well refinancing
[42:05] goes down so the duration of fixed rate
[42:08] mortgages and fixed MBS
[42:11] gets longer and that contributes to the
[42:13] problems that Banks like First Republic
[42:15] that have a large mortgage portfolio
[42:18] on the other hand if interest rates goes
[42:20] down that stimulates refinancing
[42:23] if if the refinancing happened
[42:25] instantaneously then the refinancing
[42:28] option in the mortgages would restrike
[42:30] and the duration of mortgages would
[42:33] would be restored
[42:34] but actually it happens slowly people
[42:36] refinance sluggishly so when interest
[42:39] rates go down there's a temporary
[42:41] reduction in the duration of fixed rate
[42:43] mortgages and the overall supply of
[42:45] duration in the bond market
[42:47] so my HBS colleague Sam Hansen
[42:50] documented this fact a few years ago and
[42:52] showed how it moves uh bond yields
[42:56] so you know here we are we're in uh
[42:59] we're in an environment where rates went
[43:01] down for a very long time uh look at the
[43:03] red line and they've just recently shot
[43:06] up
[43:07] um so a long-term trend has reversed
[43:11] that's the world we're in now
[43:14] uh but we can see that movements in
[43:17] interest rates uh in this figure taken
[43:19] from Sam Hansen's paper movements and
[43:21] interest rates bounce uh duration around
[43:23] so the the blue line here is the Agri
[43:27] aggregate duration of the overall bond
[43:29] market
[43:30] and if you just had a sort of
[43:32] stock of Treasury bonds sitting out
[43:34] there that blue line would be flat
[43:37] it's not flat it moves up and down and
[43:40] the reason is the contribution of MBS
[43:43] because mortgage duration is always
[43:45] shifting with refinancing behavior and
[43:47] you can see that because the red dashed
[43:50] line is what's moving the green line
[43:52] around
[43:55] okay so you know finally
[43:58] yeah if you were to have automatic
[44:01] refinancing you would prefer that or you
[44:03] prefer an arms Arrangement and can you
[44:05] contrast a little bit the Us versus UK
[44:07] because there are many similarities in
[44:09] the financial sector unlike in Denmark
[44:11] but this is one striking difference now
[44:13] we will come to you I suppose absolutely
[44:15] I mean or automatically refinancing
[44:18] mortgages would um somewhat mitigate the
[44:21] duration shocks that I just showed you
[44:24] but they would leave you with the
[44:25] problem that fixed rate mortgages have
[44:28] um uh generally long duration and they'd
[44:32] also leave you with the problem that
[44:34] with the that duration will tend to rise
[44:36] as interest rates rise amplifying the
[44:40] the problem for the bank balance sheets
[44:43] on the other hand if you have adjustable
[44:45] rate mortgages then you know deposit
[44:47] financed institutions can match
[44:49] um liabilities and assets pretty easily
[44:53] now would you also say it depends also
[44:56] on the market power of the banks on the
[44:58] deposit side or whether this impacts as
[45:00] well yes I mean there are definitely
[45:03] some subtleties here because you know a
[45:05] bank that has a very stable deposit
[45:07] franchise
[45:08] uh where they could just pay a zero
[45:10] interest rate and just sort of sit there
[45:13] um that um
[45:15] that is a kind of a long-term liability
[45:18] and you might choose to
[45:21] um match then a a long-term asset
[45:25] um
[45:27] so the the we've and we've become very
[45:29] aware of this this year that the
[45:31] flightiness of the deposit bases are
[45:33] sort of important
[45:34] um State variable for banks um
[45:39] but let me get let me go back to the
[45:42] sort of implications for monetary policy
[45:43] uh for one moment you know by by
[45:46] encouraging maturity mismatch in the
[45:48] banking sector I think these these firms
[45:51] make life hard for the FED when it needs
[45:53] to raise interest rates to combat
[45:54] inflation
[45:55] so the risk of financial instability is
[45:58] certainly
[45:59] um limiting what the FED can do right
[46:02] now
[46:02] if you have an arm system that's going
[46:04] to put the burden of rising rates on
[46:06] households
[46:08] um which is perhaps where you want it to
[46:09] be because that's where that's how
[46:11] you're influencing aggregate demand you
[46:14] may indirectly stress the financial
[46:15] system if you cause a wave of defaults
[46:17] of course and that's something you have
[46:19] to think about
[46:21] okay now finally let me quickly talk
[46:23] about uh lock-in you know just it is
[46:27] another material so you focus very much
[46:28] every all the rates you showed on
[46:30] nominal rates yeah did you also take
[46:32] inflation into account at the real rates
[46:35] because it might vary a lot whether
[46:36] you're in a circulationary environment
[46:38] or and you know a booming phase where
[46:40] it's due to high demand shocks and I
[46:43] know you worked a lot in that space as
[46:45] well
[46:46] does it have implications also on your
[46:48] mortgage a design
[46:51] with its Supply I mean
[46:55] if if we if we zoom out and think about
[46:57] the risk of really severe inflation
[47:01] another way to look at this is to say
[47:03] nominal fixed rate mortgages are a very
[47:06] strange instrument because they are a
[47:08] massive one-sided battle inflation
[47:11] home borrowers are going to win big if
[47:13] inflation gets out of control so people
[47:16] who bought houses and took out mortgages
[47:18] in the 1950s and 1960s they won the
[47:21] lottery right inflation uh wiped out the
[47:24] real value of their Mortgage Debt
[47:26] and they were left earning the houses
[47:28] and they did very well as a result
[47:30] of course you have to pay for that
[47:32] one-sided bet that option you have to
[47:34] pay for it up front in the form of
[47:35] higher mortgage rates
[47:37] this is one reason why countries that
[47:39] have a history of volatile inflation
[47:41] places like Italy even the UK
[47:44] they tend to not have long-term fixed
[47:47] rate mortgages because those things
[47:49] would be too expensive and that's one
[47:51] reason why they shifted to these
[47:53] adjustable rate
[47:54] uh mortgages
[47:57] so of course an answer to that would be
[47:59] to inflation index mortgages but um
[48:02] alternatively we can just have a fed
[48:04] that's very credible and very good at
[48:05] controlling inflation and then we don't
[48:08] need to worry about that side of things
[48:11] well let me talk now about lock-in
[48:13] briefly so
[48:16] um
[48:17] in a rising rate environment a borrower
[48:20] with an old fixed rate mortgage is going
[48:22] to be reluctant to move because if you
[48:24] move you're going to have to replace
[48:25] your cheap mortgage with an expensive
[48:27] one
[48:28] and this can reduce liquidity in the
[48:30] housing market and also prevent people
[48:32] from moving to take better jobs now in
[48:35] recent decades we haven't worried about
[48:37] that because rates have been falling all
[48:38] the time almost all the time
[48:41] but it's newly relevant uh this year and
[48:43] you know here's it's it's in the news
[48:45] just this week there was a article in
[48:49] the New York Times
[48:50] um emphasizing this point and there's an
[48:53] illustration of the lack of liquidity in
[48:56] the in the housing market people don't
[48:58] want to sell the three people on the
[48:59] left of the cartoon they don't they
[49:01] don't want to sell
[49:03] now
[49:04] lock-in can actually affect uh moving uh
[49:09] even if rates don't literally go up and
[49:11] the reason is that because there are
[49:13] fixed costs of refinancing you don't
[49:15] want to refinance a typical mortgage
[49:16] until the currently available
[49:19] firm rate is about 1.5 to 1.8 percent
[49:22] below the old rate
[49:24] moving of course forces a new mortgage
[49:26] origination so you pay a fixed cost to
[49:28] get a new mortgage rate
[49:30] and that means that the difference
[49:31] between the old rate that you give up on
[49:34] the new rate that you take on does
[49:36] affect the overall cost of moving even
[49:38] if the new rate is below the old rate
[49:41] so um there's a very nice new paper on
[49:44] that by Fonseca and Liu which is a
[49:48] working paper and this is showing
[49:50] the the horizontal axis here is what
[49:53] they call the mortgage Delta which is
[49:54] the
[49:56] um the difference between the old rate
[49:57] and the new rate
[50:00] um people way off at the right are
[50:03] people who are not paying attention and
[50:04] should already have refinanced
[50:06] uh but everybody to the left of about
[50:09] 1.8 or 1.5 uh people who
[50:13] um may or may not be moving but what you
[50:16] can see is the moving rate which is the
[50:18] vertical axis increases with that
[50:20] mortgage Delta
[50:22] and
[50:23] um if rates are rising you're going to
[50:25] be way off at the left in the negative
[50:27] region of the x-axis you're going to be
[50:29] way down uh at the bottom left of the
[50:31] figure
[50:32] and you can see that does have a
[50:34] measurable impact on the moving rate
[50:37] so how big a deal is that this paper
[50:39] estimates that our one percentage Point
[50:41] rise in mortgage rates lowers the annual
[50:44] moving rate by about 70 basis points
[50:47] which is about nine percent of the
[50:49] average moving rate
[50:51] and then they do a rather speculative
[50:54] longer term forecast using forward rates
[50:57] um to suggest that over the next 15
[50:59] years you could see a 1.9 percentage
[51:02] Point decline in the moving rate which
[51:04] would be about 25 of the base moving
[51:07] rate now you know labor market effects
[51:10] are harder to estimate but with that
[51:11] level of immobility they could be quite
[51:14] substantial and you could have a labor
[51:16] market that's kind of gummed up because
[51:17] people are not willing to move to New
[51:20] and better jobs
[51:23] now there's various yes you know just uh
[51:26] of course it also has implications on
[51:29] the housing Supply as you should have
[51:30] shown it so it might smooth the reduce
[51:32] the volatility of house prices now if
[51:34] you go into cycle interest that goes up
[51:36] and housing Supply dries up because of
[51:38] this moving fractions and we go more for
[51:40] renting
[51:41] it might stabilize house prices which
[51:44] example might lead to more house Barbers
[51:46] because it you know it's less likely
[51:47] that house prices go down
[51:53] essentially yeah the problem is not so
[51:56] much the price effects as the allocation
[51:58] the actual matching of people to houses
[52:00] but you're right there could be
[52:02] um interesting effects on on on prices
[52:05] of course
[52:07] you know that the New York Times article
[52:09] that I linked there um was arguing that
[52:12] house prices are being propped up by the
[52:14] reluctance of people to sell
[52:16] but the FED is trying to slow down the
[52:18] economy so perhaps the fed you know
[52:20] wants to limit house price Rises it may
[52:23] be a a bad side effect
[52:26] um
[52:27] anyway if you want to fix lock-in
[52:28] there's lots of ways to do it I mean
[52:30] most obviously arms arms don't have
[52:32] locking but assumable mortgages don't
[52:34] have it portable mortgages don't have it
[52:37] or the Danish system doesn't have it I
[52:40] mean any of these are solutions
[52:42] um and they all avoid the distortions
[52:45] that are created effectively by
[52:47] privileging non-movers over Movers
[52:50] you know economists are used to
[52:53] um we naturally object to things like
[52:56] rent control systems that that adjust
[52:59] rents only when tenants leave or
[53:02] property tax systems that adjust
[53:03] valuations only when properties are sold
[53:05] we recognize the friction that that
[53:08] creates
[53:10] um and I think our mortgage system is a
[53:12] little similar and so I think this is an
[53:15] issue
[53:16] now
[53:17] um so why is this why are we locked in
[53:21] or you will explain us why why we don't
[53:24] see Corrections does the free market is
[53:27] not providing instruments to solve these
[53:29] problems yeah I mean some some of the
[53:31] effects I've been talking about are are
[53:34] um you know externalities sort of policy
[53:36] issues you're going to monetary power of
[53:38] monetary transmission mechanism is not
[53:40] something that uh individual mortgage
[53:41] borrowers are going to think about nor
[53:43] are they going to worry about the
[53:44] stability of the banking system
[53:47] um others involve you know consumers who
[53:50] don't who don't anticipate properly I
[53:52] mean we talked about the overconfidence
[53:53] that people perhaps have or their
[53:56] failure to understand how to manage a
[53:58] mortgage
[54:00] um so these are all reasons why the free
[54:03] market may not solve the uh problem now
[54:07] I know you know time is limited I've got
[54:09] a few things to say about consumer
[54:11] preferences and then some policy
[54:13] suggestions do do I have time to do both
[54:15] of those yes we go a little bit longer
[54:17] okay all right so
[54:20] you know I've I've been talking without
[54:22] regard to what borrowers actually want
[54:24] but it's natural to ask well what what
[54:26] do people want
[54:28] um so one obvious point is that um you
[54:31] know when when the spread between fixed
[54:35] and adjustable rate mortgages moves
[54:36] around people do react to that interest
[54:39] costs do matter and one of the reasons
[54:41] why people like arms is that when the
[54:44] short rate is below the fixed rate it's
[54:47] cheaper initially
[54:49] um if that yield curve inverts that will
[54:51] tend to push you the other way that's
[54:53] not surprising
[54:55] but what about the cross section so
[54:57] who's doing what
[54:59] now it's well known that in the US the
[55:02] arm share is higher for subprime
[55:04] mortgages and jumbo mortgages
[55:06] and it's lower for so-called conforming
[55:08] mortgages which have Prime borrowers and
[55:11] where the mortgage is not too large
[55:13] now that's often attributed to the
[55:15] implicit government subsidies to these
[55:18] mortgages that are provided by the
[55:19] activities of the gses Fannie Mae
[55:22] Freddie Mac and so forth they provide
[55:25] credit guarantees and they securitize
[55:27] conforming mortgages uh when which are
[55:30] fixed and that may lower the rates and
[55:32] on this Pat on this interpretation the
[55:34] cross-sectional pattern is evidence of a
[55:36] distortion in the mortgage Market
[55:39] what I want to say is that research I'm
[55:42] doing now looking at Danish data shows
[55:46] very similar patterns there but in
[55:48] Denmark there are no gses there is no
[55:50] Distortion so the reason must be
[55:52] different
[55:55] so
[55:56] um in this work with Anderson Coco
[55:58] hansman and romaterai uh what we do what
[56:01] we find is that the arm share in Denmark
[56:03] is higher in two very different kinds of
[56:05] groups
[56:06] young borrowers first-time home buyers
[56:09] they have low Financial assets
[56:11] and then the other group is middle-aged
[56:13] borrowers with large houses and large
[56:15] Financial portfolios
[56:18] and what we think is that arms are
[56:20] appealing differently to these two
[56:21] groups the first group is borrowing
[56:23] constrained and they just want a low
[56:25] current interest rate which will permit
[56:27] higher current consumption
[56:29] they're aware that rates may go up in
[56:31] the future that's the risk they take but
[56:33] they believe that their income will be
[56:35] higher by the end and they'll be able to
[56:36] handle it
[56:38] the other group
[56:40] uh middle aged households with large
[56:43] houses and large Financial portfolios
[56:45] and they We Believe are using arms as a
[56:47] cheap form of Leverage if you're an
[56:49] individual and you want to lever up
[56:51] you know collateralize borrowing against
[56:53] your house is the cheapest way to do
[56:55] that and an arm is typically cheaper
[56:57] than a firm and in effect what you want
[57:00] to do is borrow at the short rate and
[57:01] then buy bonds and stocks now if rates
[57:05] go up in the future these households can
[57:07] always optimally deliver they can
[57:09] actually pay off the mortgage if they if
[57:11] they need to so they're not concerned
[57:13] about interest rate risk in the same way
[57:17] so I've got some slides with evidence on
[57:19] that but I think maybe it would be more
[57:22] interesting to the audience to skip the
[57:24] the those slides and go to my my policy
[57:28] suggestions and then we can see what
[57:30] questions and comments or another 10
[57:32] minutes
[57:34] all right um so this is just showing
[57:37] that um this is just some evidence from
[57:39] Denmark
[57:40] um the left panel shows that people with
[57:43] arms are more likely to move that's that
[57:45] makes complete sense because they only
[57:47] care about the current rate if they're
[57:48] going to move
[57:49] the right panel shows that the poorest
[57:51] people and the richest people are the
[57:53] most likely to use arms
[57:55] and then if we can look at a
[57:56] multivariate analysis we could see that
[57:58] people with big mortgages on the left
[58:00] and uh tend to use arms and the the
[58:06] effect of financial assets to income is
[58:08] u-shaped which is these two groups the
[58:10] the people with almost no Financial
[58:12] assets who are constrained and then the
[58:14] people with big Financial portfolios who
[58:16] are using arms to remember
[58:20] Okay so
[58:22] let's pull it all together and let me
[58:25] leave you with some thoughts about
[58:26] policy
[58:27] um
[58:28] I've got three slides on that okay
[58:31] so
[58:33] first thought is you know different
[58:34] countries have very different mortgage
[58:36] systems
[58:38] but the nature of these systems is
[58:40] highly persistent over time so if I come
[58:42] on Marcus's Academy and I say let's
[58:44] radically change the US mortgage system
[58:46] you know that's Pie in the Sky that's
[58:49] just not going to happen these things do
[58:50] not change quickly there's both
[58:52] political resistance to institutional
[58:54] change
[58:55] look at the fact that the government
[58:57] still is running the gses even though
[59:00] they were taken temporarily into
[59:02] conservatorship 15 years ago
[59:05] and then there's borrower resistance to
[59:07] novel mortgage products so things change
[59:09] slowly
[59:11] having said that
[59:13] I think the traditional U.S fixed rate
[59:15] mortgage does not deserve the strong
[59:17] political support that it's received in
[59:19] this country I think the US might be
[59:21] better off if our system were to shift
[59:23] no doubt gradually in the direction of
[59:26] say the Canadian system where mortgage
[59:28] rates are fixed for only five years not
[59:30] 30. monetary policy would be more
[59:33] effective and the banking system would
[59:35] be more stable
[59:39] now
[59:41] here are some more thoughts I think we
[59:43] learned in the covid-19 pandemic that
[59:45] mortgage forbearance in a recession can
[59:47] be powerfully stabilizing it it relieves
[59:50] household budgets it stimulates
[59:52] consumption
[59:54] but it would be better to build that
[59:56] into mortgages xnt and price it with
[59:58] explicit provisions
[01:00:01] another point is refinancing options
[01:00:04] which we take for granted in our system
[01:00:06] are they're hard to manage and they
[01:00:08] benefit sophisticated borrowers at the
[01:00:11] expense of lower income unsophisticated
[01:00:13] Borrowers
[01:00:14] so I would say that mortgage policy
[01:00:16] should favor either plain vanilla arms
[01:00:19] no teaser rates
[01:00:21] or automatically refinancing fixed rate
[01:00:23] mortgages over traditional fixed rate
[01:00:25] mortgages wherever possible we should
[01:00:27] lean in the direct in those directions
[01:00:30] now I particularly want to criticize
[01:00:33] points I think it's a particularly
[01:00:35] pernicious system
[01:00:36] so we could easily change the system
[01:00:39] such that if you borrow your closing
[01:00:41] costs you increase the mortgage balance
[01:00:44] in a way that could be excluded from LTV
[01:00:46] limits
[01:00:49] but but then would not change the
[01:00:51] interest rate and would not create this
[01:00:53] incentive to refinance
[01:00:56] so I that's perhaps low hanging fruit if
[01:00:59] there's something sort of feasible that
[01:01:00] I'm suggesting it would be to do
[01:01:02] something about points it's just a bad
[01:01:03] system
[01:01:05] limits of the primary drivers of this
[01:01:08] point system
[01:01:11] um well what what many people do is they
[01:01:13] go up to say um you know an 80 LTV or
[01:01:18] some threshold level or ATV LTD there's
[01:01:20] a lot of clustering there and then they
[01:01:22] may need a little more money they borrow
[01:01:24] the closing costs using points what I'm
[01:01:26] suggesting is that the you could
[01:01:29] um
[01:01:30] have the borrowed closing costs add to
[01:01:34] the loan size
[01:01:35] um but in a way that where you relax
[01:01:37] that LTV very slightly
[01:01:41] I also think that we should uh concern
[01:01:43] ourselves with lock-in I think policy
[01:01:45] should favor mortgages that don't lock
[01:01:47] in that means either arms assumable
[01:01:49] mortgages or portable mortgages
[01:01:51] basically within our system
[01:01:53] and other countries have these you know
[01:01:55] in the Canada and the UK mortgages are
[01:01:57] normally portable so we could do it too
[01:02:02] finally uh this is my very last slide I
[01:02:05] we I I hope it's been clear from these
[01:02:08] remarks that you know we can't assume
[01:02:11] that the mortgages we have are the best
[01:02:12] that can be designed
[01:02:14] there's a lot of historical accident in
[01:02:16] the type of system that we have there's
[01:02:18] many weird features of the system
[01:02:21] um I think policy should try to make
[01:02:24] space for Innovative mortgages and there
[01:02:26] are some things I haven't really
[01:02:27] discussed so for example inflation
[01:02:29] indexation of principle
[01:02:31] you there are often rules against
[01:02:33] negative amortization but that if you
[01:02:37] have such rules it should be a rule
[01:02:39] against Real negative amortization you
[01:02:41] should certainly allow nominal negative
[01:02:43] amortization
[01:02:44] you may also want to index principles to
[01:02:48] home values shared appreciation
[01:02:49] mortgages which are that you know
[01:02:52] there's a lot of
[01:02:53] um new mortgage lenders uh in the
[01:02:56] fintech space who are beginning to offer
[01:02:58] these types of contracts and I think we
[01:03:00] want to allow experimentation to take
[01:03:03] place
[01:03:04] um a useful concept is that of the
[01:03:06] regulatory sandbox where you have a
[01:03:09] light touch for new products that have a
[01:03:12] limited market of course if these
[01:03:14] products start to take off then you're
[01:03:15] going to have to move them out of the
[01:03:17] sandbox and into the the main playground
[01:03:20] and then you may need to regulate them
[01:03:21] more carefully
[01:03:23] um so this reminds me actually of Bob
[01:03:25] Schiller's push to have this macro
[01:03:28] markets and also the you know having
[01:03:30] them indexing of their mortgages based
[01:03:33] on how the shared mortgages is this yeah
[01:03:35] so I mean it's it's a it's a really
[01:03:38] interesting question as to whether
[01:03:41] um who is it that should be a house
[01:03:43] price risk right and should that be
[01:03:48] um investors deep pocketed investors or
[01:03:50] should it be actually homeowners and the
[01:03:53] ambiguity is that if you're living in a
[01:03:56] house for a long time
[01:03:58] the the house price variation the house
[01:04:01] actually Hedges you against rent risk it
[01:04:04] may not be risky in the relevant sense
[01:04:06] and in that in that situation the
[01:04:07] homeowner is actually the natural holder
[01:04:09] of the of the of the house price risk
[01:04:12] of course that's not true if you're
[01:04:14] going to be downsizing or moving to
[01:04:16] another area with uncorrelated house
[01:04:18] prices
[01:04:19] so there are many subtle issues there as
[01:04:21] to whether
[01:04:23] and how much uh homeowners should
[01:04:26] transfer the health risk uh to to
[01:04:28] lenders
[01:04:30] uh but I certainly think we want to
[01:04:32] allow some experimentation in this space
[01:04:36] so let's suppose so we go for some of
[01:04:38] these regulatory changes or this new
[01:04:40] Innovations I guess for each Innovation
[01:04:43] there will be a transition phase you
[01:04:45] probably can't just do it immediately
[01:04:47] another certain changes but I would say
[01:04:49] the transition phase has to be much
[01:04:51] longer 10 15 years and others we can
[01:04:53] introduce in you know two-year
[01:04:54] transition phase
[01:04:56] well I would I would have thought
[01:04:58] um again I was sort of beating up on
[01:05:00] points and I I think that that's um a
[01:05:03] system that one could reform quite
[01:05:05] rapidly
[01:05:07] um you know and with minimal disruption
[01:05:09] actually
[01:05:11] um other things might might take longer
[01:05:14] um so after global financial crisis
[01:05:17] there was this huge wave everybody
[01:05:18] wanted to adopt the Danish system and
[01:05:21] you also contrasted very much with the
[01:05:23] Danish system
[01:05:24] I would you use very much in the favor
[01:05:27] of the Danish system as well putting
[01:05:29] everything together I would say that the
[01:05:31] Danish system is
[01:05:35] um
[01:05:37] it fixes many of the problems with the
[01:05:39] US system
[01:05:40] so for example you don't have locking in
[01:05:43] Denmark
[01:05:45] um
[01:05:47] you don't have people frozen out of
[01:05:50] refinancing because they're it you know
[01:05:52] they've lost their job or their house
[01:05:54] prices Fallen
[01:05:56] that's a good thing
[01:05:58] um however the Danish system uh still
[01:06:02] relies on discretionary refinancing by
[01:06:05] the borrower it's not automatic
[01:06:06] refinancing and so you still get these
[01:06:09] inequalities which I
[01:06:10] showed you
[01:06:13] um the funding system is good because
[01:06:15] it's this covered Bond funding mechanism
[01:06:17] which ensures that you really do pass
[01:06:19] the interest rate risk through to
[01:06:22] end investors who are well positioned to
[01:06:24] bear it so it's good from our financial
[01:06:27] stability point of view
[01:06:29] I would say if you could take the Danish
[01:06:31] system and make the refinancing uh more
[01:06:34] automatic that would be the best
[01:06:36] possible fixed rate mortgage system
[01:06:39] uh there's still a question though as to
[01:06:41] whether it's as good as a more
[01:06:43] adjustable system which would have uh uh
[01:06:46] uh you know a stronger monetary
[01:06:49] transmission mechanism
[01:06:53] so let me just the final question so
[01:06:55] thanks a lot for very insightful talk
[01:06:57] and very you know laying out the
[01:06:59] groundwork for many policy changes here
[01:07:02] come back to the inflation thing so you
[01:07:05] made a very important remark earlier
[01:07:06] that is saying countries which
[01:07:08] experienced a lot of high inflation or
[01:07:10] volatile inflation went more for an
[01:07:12] adjustable rate mortgage system
[01:07:16] s other way around the fact that you
[01:07:19] know that I have this
[01:07:21] adjustable rate mortgage system might
[01:07:23] induce the Central Bank than to allow
[01:07:25] more inflation volatility can it go both
[01:07:28] ways and the fact if I don't have the
[01:07:30] same the fed or the central bank is more
[01:07:32] concerned about inflation volatility and
[01:07:34] has to control it therefore because it
[01:07:36] causes more distortions you know I don't
[01:07:39] I I don't think uh um
[01:07:43] I mean in an arm system inflation
[01:07:46] volatility really does bite on
[01:07:50] households in a very fundamental way
[01:07:52] because when inflation goes up
[01:07:55] what happens is inflation is eroding the
[01:07:58] real value of an orange and the borrower
[01:08:00] the lenders are compensated for that by
[01:08:03] uh the interest rate going up now the
[01:08:06] the the mortgages if you like a floating
[01:08:09] rate note so it has a stable Capital
[01:08:11] value that's all good the problem is
[01:08:13] that borrowers are in effect being asked
[01:08:15] to repay more quickly when inflation
[01:08:17] goes up
[01:08:18] now if they're unconstrained they can
[01:08:22] just adjust other borrowing and it's all
[01:08:24] fine that they can smooth it all out but
[01:08:26] if the borrowers are borrowing
[01:08:28] constrained at the time when inflation
[01:08:30] goes up
[01:08:31] it really hits their budgets now the
[01:08:33] Central Bank may like that if that may
[01:08:36] be slowing down the economy which is
[01:08:37] what the Central Bank wants to do
[01:08:40] it can be very painful for Borrowers
[01:08:42] and certainly there are political
[01:08:44] implications
[01:08:45] uh you know I grew up in England went to
[01:08:48] college in England and my cohort of um
[01:08:51] friends you know all moved to London and
[01:08:53] all bought houses and all had families
[01:08:55] in the 1980s around the same time
[01:08:58] they took out adjustable rate mortgages
[01:09:01] and then in the late 1980s the UK
[01:09:04] government lost control of inflation a
[01:09:06] little bit and rates shot up and house
[01:09:09] prices fell and you know my generation
[01:09:12] of college friends were all stuck uh
[01:09:15] with
[01:09:16] um a big big hit to their household
[01:09:18] budget
[01:09:20] um that of course slowed down the
[01:09:22] economy very effectively so monetary
[01:09:24] policy worked but it was a very painful
[01:09:27] experience and it was very politically
[01:09:29] difficult for the government it may you
[01:09:31] know it um it put a lot of pressure on
[01:09:35] the government to improve the conduct of
[01:09:37] monetary policy and not to do that to
[01:09:39] the population which then of course led
[01:09:43] to the experimentation with the European
[01:09:46] exchange rate mechanism and you know
[01:09:48] then that ended in tears and then later
[01:09:51] in the 90s uh the central the bank of
[01:09:53] England was made independent
[01:09:55] um
[01:09:56] so I think I would argue adjustable rate
[01:09:58] mortgage systems
[01:10:00] um you know monetary policy is effective
[01:10:03] but it's also becomes a very politically
[01:10:05] difficult thing because
[01:10:08] the the the a very large way that the
[01:10:11] population is having their budgets uh
[01:10:14] directly affected by what the Central
[01:10:15] Bank does
[01:10:17] I guess we saw it also last year when
[01:10:19] the guilds was jumping because of the
[01:10:21] ldi crisis and suddenly the mortgage
[01:10:24] Market froze
[01:10:25] for a few weeks yes yes
[01:10:29] okay so it's a first order importance I
[01:10:33] think it's probably very important that
[01:10:34] you know clients like you study this
[01:10:36] very carefully and uh I guess we have to
[01:10:39] have further insights and I hope that
[01:10:42] you will push the U.S mortgage Market in
[01:10:44] a better Direction and uh you know over
[01:10:46] with the transition phase we can end up
[01:10:48] with a better mortgage structure okay
[01:10:50] but also in other countries thank you
[01:10:53] and I I hope that you know the the large
[01:10:55] audience for for this uh webinar series
[01:10:58] will will uh carry on both the research
[01:11:01] and and the policy in life
[01:11:05] thanks a lot John so I really appreciate
[01:11:07] it and I would like to invite all the
[01:11:09] viewers to also attend next week where
[01:11:11] we talk about how we can use chat GPT
[01:11:15] for economic research so you will learn
[01:11:17] a lot of tips and tricks how to better
[01:11:20] use it with Kevin Bryan who is an expert
[01:11:22] on jet GPT
[01:11:25] bye-bye and thanks for joining us