Counterintuitive thought about mortgages

[Puppycow]

Which is riskier, a fixed rate mortgage or an adjustable rate mortgage? (For purposes of this discussion, I mean just a regular adjustable-rate mortgage without a so-called "teaser rate")

Most people think of a fixed rate mortgage as safer, but is this really true?
For one thing, you pay premium for a fixed rate mortgage. The premium is thought to be the price for transferring the interest rate risk to the lender.

Interest rates go up when inflation expectations go up, but inflation has the simultaneous effect of reducing the real value of debt principal. And when there is general inflation, then average wages should also inflate at roughly the same pace, right?

Putting it another way, a fixed rate mortgage is a bet that the weighted average interest rate over the life of the loan will be higher than the current market adjustable rate plus the interest rate risk premium. I'm guessing that's a bet that the house will win much more often than the customer.

Some people might say that, well, at least the worst case scenario is worse in the case of an adjustable rate mortgage. I wonder if this is really true. Default is one worst-case scenario, and I'm not at all sure that likelihood is really higher for adjustable rates than for fixed rates (assuming that the same person is buying the same house for the same price; all bets are off in case of a so-called "liars' loan," "NINJA loan" (NINJA: No Income, No Job, no Assets) or a "teaser rate mortgage" where someone buys much more house than they can reasonably expect to afford).

Another worst-case scenario is the theoretical total cost of the loan. In case of the fixed-rate loan, there is a nominal upper limit to the total cost. In fact the total cost is fixed on a nominal basis. However, the real cost depends on inflation (or deflation). Therefore, it's not clear to me that the worst-case scenario in terms of real cost is higher for the adjustable-rate mortgage.

IOW, with a fixed rate mortgage, the nominal cost is fixed, but the real cost is anything but fixed. OTOH, with an adjustable-rate mortgage, the nominal cost is not fixed, but the real cost may actually be closer to being fixed (on a real basis) than with a fixed rate mortgage.

[Aepervius]

Well , fixed rate are no bet. You know what you will pay, you know how much you bring in. And barring losing job/health problem for which both fixed and variable rate have the same risk, variable rate is a bet on various parameter independant of the loan itself. Bet are risky by nature, even if the risk is low. So yeah variable are are ALWAYS more risky than fixed rate. AT least for the one taking the loan. Now what you could argue is whether fixed rate are less expansive than variable rate.

[drkitten]

Which is riskier, a fixed rate mortgage or an adjustable rate mortgage? (For purposes of this discussion, I mean just a regular adjustable-rate mortgage without a so-called "teaser rate")

An adjustable-rate mortgage, by any normal definition of risk.


Most people think of a fixed rate mortgage as safer, but is this really true?

Yes.


For one thing, you pay premium for a fixed rate mortgage. The premium is thought to be the price for transferring the interest rate risk to the lender.

That's right.



Interest rates go up when inflation expectations go up, but inflation has the simultaneous effect of reducing the real value of debt principal. And when there is general inflation, then average wages should also inflate at roughly the same pace, right?

Not always. That's why it's a "risk."

A risk isn't a probability of a negative consequence; that's why phrases like "upside risk" aren't meaningless. A risk is an uncertainty.

Since the amount of interest and therefore the payment due is fixed on a fixed-rate mortgage, it's less "risky" than an ARM. In theory at least, the lender did DD and has a reasonable expectation that the buyer can pay it off over the length of the loan. There is no way for that same lender to have the same degree of confidence with an ARM, precisely because he doesn't know what the payment will be over time.

[Professor Yaffle]

Well we got a 5 year fixed rate and have had stable payments throughout a time when they shot up quite high and then went down really low. Although at the moment we are paying more than we would if we had a variable rate, I am definitely happy with the decision we made.

[Yeggster]

We recently dropped our mortgage all together, opting for a standard line of credit.

The way the intertest payments are handled, with the same payment level, it can be payed off vastly faster ... AND the minimum payment required drops every month.

If we have an fananicial difficulties the line of credit will self pay the interest only for months if needed.

It's variable rate interest.

[The Central Scrutinizer]

Which is riskier, a fixed rate mortgage or an adjustable rate mortgage?

Which is better - steak or lobster?

In other words, there is no correct answer.

[Puppycow]

Which is better - steak or lobster?

In other words, there is no correct answer.

Steak is better. :D

[Puppycow]

Well we got a 5 year fixed rate and have had stable payments throughout a time when they shot up quite high and then went down really low. Although at the moment we are paying more than we would if we had a variable rate, I am definitely happy with the decision we made.

I have to admit that in practice I too have opted for fixed rates each time I had the option to choose. I started on a 5 year fixed rate and when that ended I refinanced with another bank and chose a 10-year fixed rate that they were flogging. 1.9% interest (but there's some up-front fees and you have to pay a guarantee company to guarantee the mortgage for the bank too). Still, I'm saving money because my old bank was going to jack up the interest after the first 5 years.

BTW, interest rates in Japan have been near 0 for almost a decade I think. I probably would have payed less with a variable rate so far.

[Puppycow]

Excerpt from a speech by Alan Greenspan (http://www.federalreserve.gov/boarddocs/speeches/2004/20040223/)

One way homeowners attempt to manage their payment risk is to use fixed-rate mortgages, which typically allow homeowners to prepay their debt when interest rates fall but do not involve an increase in payments when interest rates rise. Homeowners pay a lot of money for the right to refinance and for the insurance against increasing mortgage payments. Calculations by market analysts of the "option adjusted spread" on mortgages suggest that the cost of these benefits conferred by fixed-rate mortgages can range from 0.5 percent to 1.2 percent, raising homeowners' annual after-tax mortgage payments by several thousand dollars. Indeed, recent research within the Federal Reserve suggests that many homeowners might have saved tens of thousands of dollars had they held adjustable-rate mortgages rather than fixed-rate mortgages during the past decade, though this would not have been the case, of course, had interest rates trended sharply upward.

American homeowners clearly like the certainty of fixed mortgage payments. This preference is in striking contrast to the situation in some other countries, where adjustable-rate mortgages are far more common and where efforts to introduce American-type fixed-rate mortgages generally have not been successful. Fixed-rate mortgages seem unduly expensive to households in other countries. One possible reason is that these mortgages effectively charge homeowners high fees for protection against rising interest rates and for the right to refinance.

American consumers might benefit if lenders provided greater mortgage product alternatives to the traditional fixed-rate mortgage. To the degree that households are driven by fears of payment shocks but are willing to manage their own interest rate risks, the traditional fixed-rate mortgage may be an expensive method of financing a home.

So far I haven't seen anyone address my point that with a fixed-rate mortgage, the nominal cost is fixed, but not the real cost, which will depend on inflation and/or deflation of the currency over the life of the loan.

[egslim]

Of course an adjustable mortgage will be cheaper, on average. With a fixed mortgage you basically pay an insurance premium to tranfer the risk of increased inflation to the lender. As usual, the reason to buy insurance against an event is if you can't afford to take the risk, specifically, can't afford the potential loss.

So far I haven't seen anyone address my point that with a fixed-rate mortgage, the nominal cost is fixed, but not the real cost, which will depend on inflation and/or deflation of the currency over the life of the loan.
Except deflation is relatively rare, usually maxes out in the low single digits, and with the Fed's fondness of low interest rates and quantitative easing is unlikely to be a major issue in the US.
While inflation is much more common, and only 30 years ago in the US peaked at over 10%. The inflation/deflation rate is currently at the low end of what's realistically possible.

But more importantly, your salary is also set at a nominal value. So even if a hypothetical high rate of deflation causes the real cost of your fixed mortgage to increase, the real value of your salary increases by the same amount, and your ability to pay won't be in jeopardy.

Fixed mortgage = higher cost, lower risk.

[Puppycow]

But more importantly, your salary is also set at a nominal value. So even if a hypothetical high rate of deflation causes the real cost of your fixed mortgage to increase, the real value of your salary increases by the same amount, and your ability to pay won't be in jeopardy.

Fixed mortgage = higher cost, lower risk.

But salaries are not fixed. Most people get a raise each year roughly in line with inflation. It's much harder to give raises when there is deflation. If you aren't getting a raise that keeps up with inflation, you are getting, in effect, a pay cut. The last time we had high inflation, my understanding is that it was driven by wage inflation.

[egslim]

But salaries are not fixed. Most people get a raise each year roughly in line with inflation. It's much harder to give raises when there is deflation. If you aren't getting a raise that keeps up with inflation, you are getting, in effect, a pay cut. The last time we had high inflation, my understanding is that it was driven by wage inflation.
I don't see the problem.

In case of increased inflation, your salary either stays the same or goes up in nominal terms.
- Your fixed mortgage also stays the same nominally, so you continue to be able to afford it as a percentage of your income - the percentage is reduced.
- Your ARM increases nominally, if it does so much faster than your salary, you have a problem.
Of course if your other expenses increase faster than your salary does, then even the fixed mortgage may eventually pose a problem, but the ARM will be much worse.

In case of deflation your salary is most likely to stay the same in nominal terms. That's because nominal paycuts piss off most people much more than inflationary cuts.
- Your fixed mortgage increases in real terms, but since your salary does so by the same proportion, you continue to be able to afford it.

Either way you're fine with a fixed mortgage. The ARM is much cheaper in case of deflation, but may get unaffordably expensive with high inflation.

The way I see it, your line of reasoning implicitely makes two false assumptions:
1) Inflation and deflation are symmetric.
They're not. Deflation is relatively rare, and only occurs in low, single digit numbers. Inflation is much more common, and may occasionally jump to incredible heights.
2) That people care more about a real change in salary than a nominal one.
Mathematically they should, but reality is different. The vast majority feels better about a 2% rise when there's 5% inflation, than about a 2% cut when there's 5% deflation. That's because inflation is more abstract, while the number of the paycheck is to most people more real.

Even if inflation is wage-driven, that doesn't mean all wages simultaneously increase by the same ratio. Many stay behind the curve, and they get hit hard by the inflation that outpaces them.

[drkitten]

nominal[/B] cost is fixed, but not the real cost, which will depend on inflation and/or deflation of the currency over the life of the loan.


Post #3.

Interest rates and inflation correlate, but do not move in lock-step (in part because interest rates depend not only on how well you do, but how well everyone else is doing). Therefore, not only will the nominal cost vary, but so will the real cost.

So a fixed-rate mortgage has a fixed nominal cost but a variable real cost. A variable-rate mortgage has both a variable nominal cost and a variable real cost.

A variable-rate mortgage is therefore inherently riskier.

[pgwenthold]

How often throughout history has there been net deflation in a 30 year period? It may have happened, but I would suggest that inflation is much, much more likely.

In that case, whereas the fixed rate mortgage does have a variable real cost, it almost invariably gets cheaper.

[drkitten]

How often throughout history has there been net deflation in a 30 year period?

Not especially relevant if you're planning on moving in the next thirty years. If mortgages were only for people who planned to pay them off in full over time, very few professionals would have mortgages.

[Brainster]

There is no denying that a variable rate is riskier. By definition, variance = risk. Now, that's not to say that variable rates are bad; I've had a variable rate on my home for years and it's definitely saved me money compared to what I would have paid on a fixed-rate loan at the time, although I'm sure that I could locate a point in time when I could have switched over to a fixed rate and saved even more.

[blutoski]

How often throughout history has there been net deflation in a 30 year period? It may have happened, but I would suggest that inflation is much, much more likely.

I think you're confusing nonzero interest rates with consistently rising interest rates. It is true that inflation is likely to be perennial, but extremely unlikely that inflation rates will rise consistently. In this post-eighties world, most countries have a national fiscal policy of maintaining control over inflation.

I can't say governments have the same attitude about lending rates.

Real estate has (arguably) been a good investment over time because the inflation of real estate prices is greater than the overall inflation rate, and in particular, greater than growth in real wages. I'm not sure why.

[drkitten]

Real estate has (arguably) been a good investment over time because the inflation of real estate prices is greater than the overall inflation rate, and in particular, greater than growth in real wages. I'm not sure why.

Population growth, IMHO.

The growth of population creates a greater demand for housing and for ancillary services that also require real estate -- like churches, schools, hospitals, police stations, strip malls, restaurants,....

This also explains why inflation in real estate prices tends to be greater in metropolitan areas; cities have been growing in population faster than farmland. Indeed, some of the greatest such price increases happen when farmland is turned into a city (as in Orlando, FL).

[blutoski]

Also:

I did my recent mortgage as a portion fixed and another portion variable. Part of the reason was to mitigate the risk of variable, but there were other factors. (Specifically, the nature of the variable product was that I could do unlimited balloon payments without service charges.)

[oggiesnr]

Another part of the eqation is what the value of your property is going to be, what your credit rating will be and what deals are on offer when your fix expires.

Steve

[egslim]

Another part of the eqation is what the value of your property is going to be, what your credit rating will be and what deals are on offer when your fix expires.
That part of the equation is unknown, since accurately predicting the future is impossible. So you're best off covering yourself against contingencies.

[JoeTheJuggler]

Which is better - steak or lobster?

In other words, there is no correct answer.

I don't think "better" is synonymous with more or less "risky". As drkitten and Aepervius (and probably others) have pointed out, an adjustable rate is obviously more risky.

Whether or not that is "better" for a certain individual in a certain situation is another question.

[pgwenthold]

I think you're confusing nonzero interest rates with consistently rising interest rates. It is true that inflation is likely to be perennial, but extremely unlikely that inflation rates will rise consistently.

No, I was talking about the cost of mortgages in real dollars. We'll just talk about fixed rates now. It applies to adjustable rates, too, but that has variation. If I have a mortgage that costs me $1000 a month, that will still be $1000 a month 30 years from now. Meanwhile, the cost of living will have gone up over that time (inevitably - as I asked, has there ever been a 30 year period with net deflation?). I usually compare it to a rental. So if rentals are running $700, and my mortgage is $1000 a month, then it would be cheaper to rent. However, if I intend to stay (as DrKitten points out, this is not always the case), and if rental rates just follow inflation (call it an average of 3%), then 30 years from now I will still be paying $1000 but rent would cost $1650, so I am getting off way cheaper. And regardless of the inflation rate, as long as there is inflation over 30 years, my $1000 costs less in year 30 than it does in year 1. The only way my mortgage costs more in the end (in real costs) is if there is net deflation. That was what I was talking about.

Now, throw a variable rate mortgage in the mix and it gets more complicated, with my montly payment no longer always $1000. Others have discussed the possibilities, but if, for example, interest rate increases are associated with high inflation, then that hurts the value of my ARM, because you don't get the benefit of the inflation in terms of real cost.

[drkitten]

No, I was talking about the cost of mortgages in real dollars. We'll just talk about fixed rates now. It applies to adjustable rates, too, but that has variation. If I have a mortgage that costs me $1000 a month, that will still be $1000 a month 30 years from now. Meanwhile, the cost of living will have gone up over that time (inevitably - as I asked, has there ever been a 30 year period with net deflation?). I usually compare it to a rental. So if rentals are running $700, and my mortgage is $1000 a month, then it would be cheaper to rent. However, if I intend to stay (as DrKitten points out, this is not always the case), and if rental rates just follow inflation (call it an average of 3%), then 30 years from now I will still be paying $1000 but rent would cost $1650, so I am getting off way cheaper.

Then, not to put too fine a point on it, your mortgage broker is an idiot and doesn't know how to set rates.

The whole point of interest is to compensate the lender for the joint risks of predicted inflation and of possible default in repayment. You can see this from the way that interest rates tend to correlate with inflation; higher inflation means that the principle loses more buying power, so you (generally) pay more in interest to compensate the lender.

If your rental is $700 now, but rent is expected to be $1650 then, I can easily set my mortgage interest rate so that the payment right now is $1650 and will remain so -- and you'll not only not be getting off "way cheaper," you'll actually be paying substantially more over the length of the mortgage, but you'll also be building equity. Heck, I can set interest rates so that you pay $2000 a month, and I end up getting much more in cash -- but you own the house at the end.

There is, of course, a sweet spot, a point where the additional money you pay in interest exactly balances the expected inflation. Ideally, that's where I will set my (fixed)-interest rate, or more likely, a little higher to balance out the chance that you will fail to repay.


And regardless of the inflation rate, as long as there is inflation over 30 years, my $1000 costs less in year 30 than it does in year 1. The only way my mortgage costs more in the end (in real costs) is if there is net deflation.

No. It can also cost you more if you pay a high (fixed) rate of interest.

What will end up saving you money, in the long run, is if inflation rises above my expectations. What will end up costing you money, in the long run, is if inflation drops below my expectations.

Expected, stable inflation is simply priced into loan terms. It's unexpected or unstable inflation that causes problems.

[Brainster]

There was the great deflation from 1870-1890 or so, which isn't quite thirty years but is still a pretty long period of time, which arose from the US government paying off the Civil War debt.

[drkitten]

That part of the equation is unknown, since accurately predicting the future is impossible. So you're best off covering yourself against contingencies.

But it's equally unknown for fixed and variable rate mortgages. So from a comparative perspective, it doesn't matter.

[blutoski]

No, I was talking about the cost of mortgages in real dollars. We'll just talk about fixed rates now. It applies to adjustable rates, too, but that has variation. If I have a mortgage that costs me $1000 a month, that will still be $1000 a month 30 years from now. Meanwhile, the cost of living will have gone up over that time (inevitably - as I asked, has there ever been a 30 year period with net deflation?). I usually compare it to a rental. So if rentals are running $700, and my mortgage is $1000 a month, then it would be cheaper to rent. However, if I intend to stay (as DrKitten points out, this is not always the case), and if rental rates just follow inflation (call it an average of 3%), then 30 years from now I will still be paying $1000 but rent would cost $1650, so I am getting off way cheaper. And regardless of the inflation rate, as long as there is inflation over 30 years, my $1000 costs less in year 30 than it does in year 1. The only way my mortgage costs more in the end (in real costs) is if there is net deflation. That was what I was talking about.

Now, throw a variable rate mortgage in the mix and it gets more complicated, with my montly payment no longer always $1000. Others have discussed the possibilities, but if, for example, interest rate increases are associated with high inflation, then that hurts the value of my ARM, because you don't get the benefit of the inflation in terms of real cost.

Hm. Variable rate usually keeps the fixed monthly payment at $1,000 regardless of interest rate changes. What a higher interest rate does is mean the $1,000 covers more interest than before. Ultimately, the total amortization is increased. Lower interest rate means you're paying more principal in the $1,000.

Normally, I think it's advised to set your monthly payments to what you can offord. That's the starting point.

[pgwenthold]

Hm. Variable rate usually keeps the fixed monthly payment at $1,000 regardless of interest rate changes. What a higher interest rate does is mean the $1,000 covers more interest than before. Ultimately, the total amortization is increased. Lower interest rate means you're paying more principal in the $1,000.


So it's also a variable term, then?



Normally, I think it's advised to set your monthly payments to what you can offord. That's the starting point.

What I could afford in year 1 is very different from what I will be able to afford in year 30. For a variety of reasons. But that would even be the case if I were working in the same job and had only cost of living raises tied to inflation.

Of course, you are also free to pay down the principle if you can afford to pay more, so that you don't make it to year 30, but you don't have to.

[pgwenthold]

No. It can also cost you more if you pay a high (fixed) rate of interest.


More compared to what?

I am talking about the cost of my mortgage NOW compared to the cost of the mortgage in 30 years. Because of the fixed payment, my mortgage payment costs me more now than it will 30 years from now in real terms. That is true regardless of the interest rate, but is only true if there is net inflation.

If you are making some other comparison, then it is not what I am talking about.

[drkitten]

I am talking about the cost of my mortgage NOW compared to the cost of the mortgage in 30 years.

Ah. I'm talking about the comparison you made, between rental costs and mortgage costs, which I think was your implicit comparison in "30 years from now I will still be paying $1000 but rent would cost $1650, so I am getting off way cheaper."


Because of the fixed payment, my mortgage payment costs me more now than it will 30 years from now in real terms. That is true regardless of the interest rate, but is only true if there is net inflation.

That's true, but not necessarily relevant, because you pay interest up front on the balance of the mortgage. So your payment thirty years from now will be $1000, but only because you already paid $650 (or its equivalent) in interest on the mortgage in the first few months. So the pre-paid interest plus the payment you make at the end of the mortgage term still ends up costing you the same in real terms as if you had set up a schedule of increasing instead of fixed payments.

[blutoski]

So it's also a variable term, then?

That's very rare.

You'd have the same $1,000 payments, and the portion that goes toward principal is usually low at the beginning, but rises over time.

When you roll over the mortgage or obtain a new one, there will be a remaining principal, and if interest rates rose, it will be higher than you originally calculated. But you're still out of your mortgage in 5 years.



What I could afford in year 1 is very different from what I will be able to afford in year 30. For a variety of reasons. But that would even be the case if I were working in the same job and had only cost of living raises tied to inflation.

Yes, but for some reason, you were comparing rental with some sort of monthly mortgage payment. Generally, you can afford $1,000 for housing or you can't. The rental and mortgage payments will be the same for a particlular person. It's a question of what you get for your $1,000, and that includes a lot of intangibles such as pride in ownership, and capital gains.

Since housing prices are rising faster than the rest of inflation and also rising faster than wages, it is always logical to buy based strictly on these three factors. But there are other factors, so I can't say that it's always logical to buy.

[pgwenthold]

Ah. I'm talking about the comparison you made, between rental costs and mortgage costs, which I think was your implicit comparison in "30 years from now I will still be paying $1000 but rent would cost $1650, so I am getting off way cheaper."


Nah, that was just my way of showing that you are paying less each month for housing at the end of the term than you would be if you were just paying rent during an inflationary term.

The comparison of rent vs mortgage costs is a lot more complicated, and making up numbers for a starting point is not a basis for serious analysis.



That's true, but not necessarily relevant, because you pay interest up front on the balance of the mortgage. So your payment thirty years from now will be $1000, but only because you already paid $650 (or its equivalent) in interest on the mortgage in the first few months. So the pre-paid interest plus the payment you make at the end of the mortgage term still ends up costing you the same in real terms as if you had set up a schedule of increasing instead of fixed payments.

But if I had a schedule of increasing payments, I wouldn't be talking about the cost 30 years later. Shoot, when would that be paid off? Year 10? 12? Probably somewhere in there.

However, I think my rent comparison kind of shows your point. Yeah, I am paying a lot less for housing in year 30, but because I paid more up front. In the scenerio I gave, it would take an inflation rate of 1.25% to break even on the end of the term payment. However, it needs to be 2.5% to break even on the total cost over the full term, although that is not taking into account the equity. If this is a $100K house, and still worth $100K at the end of the term, then the inflation only needs to be .25% for me to come out ahead. And if the housing market increases at the inflation rate, that rate only needs to average 0.125% over the 30 years.

Of course, this is why rental property can work as an investment, even if the rent charged is merely enough to cover the mortgage.

[drkitten]

But if I had a schedule of increasing payments, I wouldn't be talking about the cost 30 years later.

Why not? You can set up a schedule of payments for any timeframe you like. Borrow $300,000 and pay it off at $1000/month (increasing at 3% per year) for thirty years.

Shoot, when would that be paid off?

Year thirty, by construction....

The "cost of the mortgage in real terms" is a fairly complicated calculation, precisely because with a typical mortgage, you're paying more than the house is worth in the first few years, and substantially less in the last. The payment schedule is set up to reflect this, with the expected inflation rate figured into the interest rate that you're charged.

The end result, of course, is that you pay much more than the nominal cost of the house --- and typically much more than the nominal cost of the house, even at the end of the mortgage. This more than compensates the lender for whatever rate of inflation he expects over the term of the mortgage.

Where the trouble comes in is if there is unexpected inflation. If interest rates are set assuming 3% inflation, but there is "really" only 2% inflation, then the mortgage holder's real costs will be substantially higher than expected, and of course, vice versa. So you can end up losing substantial amounts of real money on a fixed-rate mortgage even without deflation.

[drkitten]

That's very rare.

You'd have the same $1,000 payments, and the portion that goes toward principal is usually low at the beginning, but rises over time.

When you roll over the mortgage or obtain a new one, there will be a remaining principal, and if interest rates rose, it will be higher than you originally calculated. But you're still out of your mortgage in 5 years.

I'm not understanding, then.

I take out a 30-year ARM with a monthly payment of $1000. Next month, the bottom falls out of the market and interest rates triple.

I'm still paying $1000 per month. I'm still out-and-done in thirty years (360 payments). So my total payment is $360,000 regardless of interest rates.

What "adjusts"?

My understanding (http://www.moving.com/articles/adjustable-rate-mortgage-ARM.asp) is that when interest rates go up, your payment goes up. (" 1-Year Adjustable Rate Mortgage : This is a 30-year loan in which the rate (and therefore your monthly payment) changes every 12 months on the anniversary of your loan.") The only loans I've seen where the payment is fixed are the "balloon" loans where you pay what you can for five years and then skip town just before the bank tells you that you still owe $108,000 on your $100,000 house.

[pgwenthold]

Why not? You can set up a schedule of payments for any timeframe you like. Borrow $300,000 and pay it off at $1000/month (increasing at 3% per year) for thirty years.



Oh, that's a different ballgame. I thought you were talking about increasing your payments on your initial $100K mortgage (our $1000/month scenerio). That explains better your comment about loading interest payments up front in a fixed payment schedule.

Do companies do this type of "real cost fixed" mortgage?

[pgwenthold]

I'm not understanding, then.

I take out a 30-year ARM with a monthly payment of $1000. Next month, the bottom falls out of the market and interest rates triple.

I'm still paying $1000 per month. I'm still out-and-done in thirty years (360 payments). So my total payment is $360,000 regardless of interest rates.

What "adjusts"?

My understanding (http://www.moving.com/articles/adjustable-rate-mortgage-ARM.asp) is that when interest rates go up, your payment goes up. (" 1-Year Adjustable Rate Mortgage : This is a 30-year loan in which the rate (and therefore your monthly payment) changes every 12 months on the anniversary of your loan.") The only loans I've seen where the payment is fixed are the "balloon" loans where you pay what you can for five years and then skip town just before the bank tells you that you still owe $108,000 on your $100,000 house.


Yeah, if you are still "out of your mortgage" after a fixed term, what it would have to mean is that you might be left with some balance. But that still means that the balance is due at the end of the term. You may have to get another loan to pay off the balance.

BTW, what is the distinction between a mortgage and other types of loans? You can have cars that are more expensive than houses, so what makes one a mortgage and another just a loan?

[drkitten]

Oh, that's a different ballgame. I thought you were talking about increasing your payments on your initial $100K mortgage (our $1000/month scenerio). That explains better your comment about loading interest payments up front in a fixed payment schedule.

Do companies do this type of "real cost fixed" mortgage?

Not typically, no. The default risk would be unacceptably high, lenders prefer to get their money up-front (avoiding the discount problem) and consumers wouldn't go for the ever-increasing schedule of payments. It's purely a thought experiment.

But it's a very useful thought experiment precisely because it shows how one overpays in the early stages of a conventional mortgage and underpays at the end, irrespective of interest rates. Therefore, neither the real cost of the first payment nor the real cost of the last payment reflects much about the actual real cost of the mortgage as a whole.

So there are two parameters involved. The seller typically sets the nominal price of the house, because the seller gets his money at closing, when the real price and the nominal price are identical. The lender sets the real price of the house, because the lender is the one being paid back over time, and he uses the interest rate to control what the real price is.

If the seller and the lender are the same person, you can get all sorts of bizarre things going on. I could, for example, sell you my house for ten dollars, at a hundred thousand percent interest. If my calculations are correct, you would pay about $850 a month, all of it interest, and have a final payment of $10 thirty years from now. And I think we'd both be happy with that deal. Indeed, you might be even happier with that deal than you would be with a conventional mortgage. You now have a bill of sale for $10 on the house, which you can use to lower your property taxes, and all of your house payments are deductible as mortgage interest.

But all this really shows is that "inflation," per se, has little effect on the value of a mortgage, if it's properly taken into account at the start. I can set the (planned) real value of the mortgage independently of the nominal value. It's the difference between the planned real value and the unplanned real value (i.e. what actual inflation is) that causes problems all around.

[drkitten]

BTW, what is the distinction between a mortgage and other types of loans? You can have cars that are more expensive than houses, so what makes one a mortgage and another just a loan?

Linguistic. mostly. The same reason that old men have grey hair while old women are platinum blondes.

There's also a whole body of case law dating back to medieval times about mortgages that obviously do not apply to medieval automobiles.

One of the big (legal) issues, for example, is "who owns the [mortgaged] land"? There are two obvious answers, both obviously wrong. For example, the land might be owned by the lender until the debt is paid off. Unfortunately, this means that (for example), if the lender goes bankrupt ten years from now, his creditors now own the mortgaged land, and who preserves the rights of the borrower (what the creditors want to keep the land)?

The alternative (which is more common today) is that the borrower actually owns the land and the lender has a lien or something on it. The problem with this, of course, is fraudulent conveyance. I mortgage the properly, turn around and sell it for ten bucks to my cat, and default on the mortgage. You can't get the money from me, but you can't get the land back, either, because I no longer own it.

So, way back when, the main difference between a mortgage and a loan was that a mortgage was written using the first legal theory, a conventional loan with the second. Since this made so little sense that even the legislature eventually noticed, most jurisdictions have cleaned this matter up.

[drkitten]

How odd. The fairies are evidently posting for me.

[blutoski]

I'm not understanding, then.

I take out a 30-year ARM with a monthly payment of $1000. Next month, the bottom falls out of the market and interest rates triple.

Maybe it's different in Canada: I've never seen a 30-year mortgage here. I have seen a 5-year mortgage, with payments based on a 30-year amortization at the prevailing rate. If it's a fixed rate, neither the payments nor the amortization will change. If it's an adjustable rate, the payments stay the same, but they're covering less principal and more interest, so 5 years in, your new mortgage may require a 26-year amortization to stay at $1,000/mo payments.




My understanding (http://www.moving.com/articles/adjustable-rate-mortgage-ARM.asp) is that when interest rates go up, your payment goes up. (" 1-Year Adjustable Rate Mortgage : This is a 30-year loan in which the rate (and therefore your monthly payment) changes every 12 months on the anniversary of your loan.") The only loans I've seen where the payment is fixed are the "balloon" loans where you pay what you can for five years and then skip town just before the bank tells you that you still owe $108,000 on your $100,000 house.

Ah. You're talking about a mortgage that covers the entire amortization period. In this situation, I can understand why they'd have to increase the payments if the interest rates change.

However, I've never met anybody who has agreed to a 30-year term. That is extremely risky, because you're locked out of re-negotiating for better terms down the line. As a consequence, it's very rare these days. They were very common in our grandparents' era. I think what happened is that the high interest rate phase in the '80s put too many people out of their homes because they increased the payments instead of extendig the amortization, so people are wary of that type of agreement.

The strategy with which I'm familiar is to sign up for between 1 and 5 years at a fixed or variable interest rate, but the payments don't change. Increases in interest rates mean you pay off the mortgage later; decreases in interest rates mean you pay off the mortgage sooner.

[drkitten]

Ah. You're talking about a mortgage that covers the entire amortization period. In this situation, I can understand why they'd have to increase the payments if the interest rates change.

However, I've never met anybody who has agreed to a 30-year term. That is extremely risky, because you're locked out of re-negotiating for better terms down the line.

Ah! I see the issue, then.

In the States, thirty-year fixed-rate mortgages are probably still more common than not. But almost all mortgages allow for pre-payment (in fact, I think they are required by law to allow that) so if you can lay your hands on a quarter million in cash, you can just pay it off here-and-now.

Which makes it relatively easy to refinance; you just borrow the quarter million in cash on whatever the new mortgage lenders terms are and pay off the old one.

Similarly, thirty-year ARMs are fairly popular -- they will typically come with a lifetime interest cap, so you are guaranteed that the interest will never rise above (some obscenely large nominal sum).

[blutoski]

Ah! I see the issue, then.

In the States, thirty-year fixed-rate mortgages are probably still more common than not. But almost all mortgages allow for pre-payment (in fact, I think they are required by law to allow that) so if you can lay your hands on a quarter million in cash, you can just pay it off here-and-now.

You can make what are called 'balloon payments' in Canada, too, but the mortgage may have terms that limit the amount before you get dinged a service charge. This is why we did 50:50 with fixed and variable - the variable mortgage had unlimited balloon payment options, but the fixed mortgage was limited to 15% of the principal per year before early repayment charges applied. We felt we would be able to make significant balloon payments over the next 5 years.



Which makes it relatively easy to refinance; you just borrow the quarter million in cash on whatever the new mortgage lenders terms are and pay off the old one.

That'll work, but I guess I'd be frustrated if the concern was about quality of service, and I couldn't get a new mortgage for the remaining amount to close out the original.



Similarly, thirty-year ARMs are fairly popular -- they will typically come with a lifetime interest cap, so you are guaranteed that the interest will never rise above (some obscenely large nominal sum).

Well, colour me educated. As a comparison, have a look at the largest mortgage lender in Canada's current offerings:

http://www.rbcroyalbank.com/products/mortgages/view_rates.html

Anything longer than 5y is closed, and the closest to a 30y is a 25y with current rate of 8.05%. I don't imagine they sell a lot of these.

[Typicallucas]

... I started on a 5 year fixed rate and when that ended I refinanced with another bank and chose a 10-year fixed rate that they were flogging. 1.9% interest (but there's some up-front fees and you have to pay a guarantee company to guarantee the mortgage for the bank too). Still, I'm saving money because my old bank was going to jack up the interest after the first 5 years.

If you are obtaining 5 and 10 year "fixed rate" mortgages that get jacked up after 5 and 10 years... isn't that just a 5 or 10 ARM - adjustable rate mortgage??

It was always my understanding that a fixed rate mortgage is fixed for the life of the loan. Not fixed for the first 5 or 10 years.

I don't know the answer to your original question. I tried to figure it out but I think it will take someone who is smarter with math than I am.

I suspect that it will depend on the terms of the two loans, which one will cost less over all. Also you would have to make some assumptions about what will happen with interest rates and inflation, which may or may not be close to what happens in the real world.