Covering a mortgage without losing your ass(ets).

[Nords]

Let's put some numbers to the eternal reciprocated diatribes on rent vs own & mortgage vs debt free.

I'm not claiming that ownership is always better than renting, or that mortgages are always better than debt-free. Those are personal choices and they probably should have a bigger emotional content than a cold-hearted bean-counting financial analysis-- especially if you have trouble sleeping at night.

But if you've been wondering whether you should pay off your mortgage early, then here's another perspective.

We signed the refinance papers today for a credit union 30-year 5.5% mortgage. We'll be making only the monthly payments and no additional principle so this one won't be paid off until the end of 2034 (when I'm 74).

The nicest thing about this refinance (our third) is that the mortgage payments are finally low enough to be covered by my pension with enough left over for property taxes, home/liability insurance, maintenance, and even utilities! Covering the home & mortgage with the pension takes all the uncertainty out of the math. The rest of our living expenses will be handled by <4% withdrawals from the remainder of the retirement portfolio. Spouse's wage income is certainly appreciated but not necessary.

The money that the credit union gave us is invested in the iShares S&P600/Barra Small-cap Value ETF (IJS). The ETF's expense ratio is 0.25%. Dividends (~1%) & cap gains (if any) will be reinvested free of charge. The closing price on 8 October 2004 was $110.06.

Since closing costs will be recouped by the lower payments in just eight months, I'm not going to count that against the returns. The Fidelity brokerage fee is minimal (one really big trade in a frequently-traded account) so I'm not counting that either.

With due regard for Bernstein's four-part <a href="http://www.efficientfrontier.com/ef/998/hell.htm">
"Retirement Calculator from Hell"</a>, we plan to buy & forget with the presumption that past returns are prologue. <a href="http://www.fireseeker.com/">FIRECalc</a> gives this scheme a 74.2% chance of making a profit. (The portfolio's initial draw is 6.81% but inflation is zero since the payments stay the same for 30 years.) If the portfolio survives the 1-in-4 catastrophe, its mean final value is over three times the mortgage's initial balance. (FIRECalc wants a 3.8% mortgage interest rate before it boosts the chance of success to 95%, but Bernstein claims that anything over 80% is meaningless.)

We're comfortable with a portfolio that's high in equities, so this shouldn't exceed our risk profile. But we'll keep an eye on long-term CD rates if we find ourselves tossing & turning at night. It'd be quite sweet to make a large deposit with the credit union and earn more in dividends than we pay in principle & interest. In our 15% tax bracket I guess that'd be >6.5%. (It's too bad that the Treasury no longer sells 30-year bonds.)

If you have a detailed retirement spending plan and if your retirement portfolio is adequately capitalized for a safe withdrawal rate, then a defined-benefits pension may be more than enough to pay off a low-interest-rate mortgage. For us the odds certainly outweigh the risk profile.

Excellent post Nords! You know, I could crunch those
mortgage vs. no mortgage numbers all day long and
end up with a different answer each time. With me,
not having a mortgage is psychological. 50% of my net
worth in real estate with no debt gives me a warm fuzzy.

John Galt

[wabmester]

we plan to buy & forget with the presumption that past returns are prologue. FIRECalc gives this scheme a 74.2% chance of making a profit.

Wow. You've got a *lot* of faith in historical returns.

So, you're doing this knowing that your house is safe because your pension covers the mortgage, right? But didn't you just put your retirement at risk? Assuming your withdrawl was around 4%, and FIREcalc told you that was 100% safe, didn't you just give yourself an overall 25% chance of failure by playing this game?

And aren't you one of those who likes to consider their pension as their bond allocation? If that's the case, do you also consider your mortgage to be a negative bond allocation? Forgetting the pension for a moment, if your mortgage represents 20% of your assets, you're now 120% in stocks. I think that sets some sort of record for risk a retiree is willing to take :)

wab's right.....I don't have much faith in the past repeating either. Hell, I don't have that much faith in the future. Still have complete faith in me though


John Galt

[cute fuzzy bunny]

I just have a couple of questions, that so far I dont think anyone has answered.

Do you own any bonds?

Do they pay less than your 5.5% mortgage rate?

Are they in an amount that is lower than your mortgage principal payoff?

Why do you continue to hold those bonds making less than the one you have issued to the bank?

If you feel that this scenario works, have you secured as much debt as you can possibly garner that is charging a lower rate than the historic 30 year stock return? If not, why not?

What does your 'firecalc' (or any other calculator you fancy) say about your scenario if you pay the mortgage off from that bond pool and leave the rest of your allocations alone (essentially increasing your stock:bond ratio)?

What does your tax man say about the reduced withdrawal rate from not making mortgage payments vs your savings from any possible deductions?

Given that the vast majority of people dont stay in a mortgage for 30 years...in fact a huge majority stay in one for less than 7 and very few go past 10...why does everyone calculate the scenario as a 30 year one (unless of course you are absolutely sure you'll be in the house for 30 years)?

My point here is:

I'd bet that anyone who RE'd has at least enough in bonds to cover their mortgage. They own those bonds to reduce volatility in their portfolio, which is a concern to maintain a certain withdrawal rate, the lions share of which goes to pay the mortgage. In the meanwhile, you're paying more for the mortgage than you're receiving from your bonds.

You're also presuming you'll stay in a house longer than 90+% of people, and that you wont encounter a bear market that will cause a failure of your plan. We've had a lot of 7-10 year bear markets.

Seems to me you're creating a problem, then costing yourself money to solve it, in the hopes that you may come out on the upside if some stuff that you cant control or predict works out as it has in the past over periods of time that probably wont apply to you.

And in the spirit of reciprocal diatribes, I once again and for the 35th time do not dispute the fact that a certain amount of money invested in the stock market over any 30 year period will exceed 5 or 6% return. That isnt the question.

In that narrow and single threaded analysis, holding as much debt as you can get your hands on and investing it all in stocks makes a world of sense. In fact, while rates were low these past years, we should have been taking out loans on our paid off cars, increasing our mortgage to 100%, taking out helocs if possible, and increasing personal lines of credit with rates below 9% to the max.

I'm trying to continue to look at this in a broader financial sense.

This seems to me to be analogous to the guy who wonders why he works all day long to pay for a house he has to leave to go to work, in a car that sits in the parking lot all day, and to pay for the clothes he had to buy only for work, and also to pay for a landscaper, daycare, housekeeper and handyman to do all the things he cant do because he's spending all his time at work.

[wabmester]

I just have a couple of questions, that so far I dont think anyone has answered.

...

Why do you continue to hold those bonds making less than the one you have issued to the bank?

I've answered this one before. There's another way to play the mortgage as a potential future interest rate arbitrage, which Nords alluded to.

For example, you stick the loan in a 1-year CD paying 3%. Then you see what rates do next year. If long-term rates have moved up past 5.5%, you can lock in a winning arbitrage play. Otherwise, you can decide to pay off the loan or play the game again. Your downside is limitted to 2.5%/year.

If you feel that this scenario works, have you secured as much debt as you can possibly garner that is charging a lower rate than the historic 30 year stock return? If not, why not?

None of these bets are a sure thing. As in any bet, the size of your wager is a factor of downside risk, upside potential, and your personal aversion/need for each.

Given that the vast majority of people dont stay in a mortgage for 30 years...in fact a huge majority stay in one for less than 7 and very few go past 10...why does everyone calculate the scenario as a 30 year one (unless of course you are absolutely sure you'll be in the house for 30 years)?

I wonder what the average is for a retiree. I'd be willing to bet it's longer than the overall average.

[cute fuzzy bunny]

I've answered this one before.

To the best of my knowledge, neither you nor anyone else has answered this, or any of the other questions.

In any calculator you choose, using a portfolio size in a reasonable range for someone to be retired early, if you remove a bond allocation, pay off the mortgage and reduce your withdrawal rate by the mortgage amount, your remaining portfolio will result in a higher terminal value, a higher percentage of 'firecalc success', and usually a lower tax burden. Every time.

Using the same stock/bond allocations in both a mortgage and no mortgage scenario results in a nod towards keeping a mortgage.

But I dont know why you wouldnt change your portfolio to suit the fundamentally changed scenario.

The only five situations I can consider where keeping a mortgage is a good idea:

1 - You have bonds or other cash instruments paying more than a mortgage rate
2 - You have no bonds or other cash instruments at all
3 - You're so ridiculously wealthy it doesnt matter which way you do it
4 - You have such a small portfolio that you cant afford to pay off the mortgage at all and stay solvent
5 - You are going to maintain a fixed stock:bond ratio regardless of the financial situation or scenario

#1 is feasible for some people. My dad has some 6 and 7% bonds. #2 is unlikely for an ER but possible for someone young and in the acquisition phase of investing. #3..screw off and quit reading any of this stuff. #4...you're an unemployed speculator, not an ER. #5 WHY?!?

[cute fuzzy bunny]

I wonder what the average is for a retiree. I'd be willing to bet it's longer than the overall average.

Possibly for an older "traditional age" retiree thats already bought their "retirement home". My life today is nothing like what it was 4 years ago. I've moved once in ER and will probably do it again within the next 7 years.

I also see an awful lot of people who buy their 'retirement home' and dont live in it very long, or who move around. My dad lives in a community of 3000 retiree's homes. All of them were built less than 9 years ago, so 100% of them have moved in the last 9 years. His street has 14 homes on it. 5 of them have moved in the last 2 years, and two more are currently planning to sell and move. Small sample size but...

I'll go look for some data.

In the meanwhile, put on a straight face and tell me that you think the majority, or even a fairly strong minority of the ER ranks will stay in their current home for 20-30 years.

[cute fuzzy bunny]

Some data...i'll update this as I read so it doesnt fall out of my head...

Average mortgage length has been dropping steadily. Used to be 7.5 years, its now around 5.

American mortgage terminations result in a move 80% of the time; other way around outside the US; people tend to pay off their mortgages and then stay in the home 80% of the time in europe.

European mortgage length is 2/3 that of americans, but they stay in the home twice as long on average.

Less than 15% of 'real' (over 67) retirees have a mortgage.

...

[wabmester]

The only five situations I can consider where keeping a mortgage is a good idea:

1 - You have bonds or other cash instruments paying more than a mortgage rate
2 - You have no bonds or other cash instruments at all
3 - You're so ridiculously wealthy it doesnt matter which way you do it
4 - You have such a small portfolio that you cant afford to pay off the mortgage at all and stay solvent
5 - You are going to maintain a fixed stock:bond ratio regardless of the financial situation or scenario

I'm not sure why this keeps getting missed, but how about
6 - You can lock in a mortgage at a fixed historically low rate, you think there's a good chance that rates will go higher, so you take a low-risk bet that has a chance to pay you a spread over your mortgage rate in the near term.

[cute fuzzy bunny]

That would be an unfounded speculation variation on #1. That someday you might be able to buy low risk cash instruments that would pay more than the mortgage.

You use the terms "good chance" and "low risk" in areas that I think are highly chancy and perfectly risky.

[wabmester]

You use the terms "good chance" and "low risk" in areas that I think are highly chancy and perfectly risky.

There are some things I will probably never understand. Like how some people think that a maximum downside of 2.5% over one year is "perfectly risky" while they think that stocks are guaranteed to return over 5% pa over 30 years.

[cute fuzzy bunny]

Okey dokey.

What makes you certain that interest rates will rise in the next year? By an amount that will bring a risk free bond into a positive position over current mortgage rates? That would take about a 1.5% rate increase that would be reflected in bonds.

I think we have a very, very good chance that interest rates will remain within .5% of where we are today, which will not raise any risk free cash instrument above the average mortgage rate available today. For at least several years.

Here are some numbers on the above scenario I mentioned, since Nords wanted real numbers.

Using firecalc:

Assumptions:

200k mortgage, for 30 years (just to humor)

Annual payments @5% (well below the average rate available today): $12,883.68

Base withdrawal rate is $30,000.00 per year without mortgage

Portfolio is 1.2 million with mortgage, 1m without.

CPI for inflation

Run through firecalc with the mortgage of 200k, withdrawal of 30k+12,883.68 with a 60/40 split of stocks and tips @2.5% for 30 years gives you:

100% success, $4,970,027 average terminal portfolio.
Smallest terminal portfolio: $2,291,500 at the end of 1929's 30 year run

Same scenario with $1,000,000 portfolio, no mortgage, withdrawal of 30k, and the 72/28 stock/bond split that results from paying off the mortgage and reducing the bond allocation by 200k (720k stocks, 280k tips @2.5%) results in:

100% success, $5,586,700 average terminal portfolio.
Smallest terminal portfolio: $2,984,528 at the end of 1929's 30 year run

Run it with any portfolio size, any mortgage size, any term of years, you get the same result. At one million dollars portfolio size, the "have a mortgage" scenario falls below 100% while the "no mortgage" scenario remains above 100%. You have to drop below a $700k portfolio before the 'no mortgage' scenario starts to fail. In other words, without a mortgage you can retire on 70% of what it takes to do so safely with a mortgage.

Excepting the 5 conditions above, you are ALWAYS better off paying the mortgage off from your bond allocation. Besides the fact that its financially sensible, it actually REDUCES your risk profile rather than increasing it in the possible hope that stocks will return a certain amount or that interest rates will get to a certain point.

As Cut-Throat frequently says "Sounds like work". And speculation. When its not necessary.

[wabmester]

What makes you certain that interest rates will rise in the next year?

There is very little I'm certain of. But here are some reasons I think it's a good bet (meaning I'm willing to spend up to 2.5% to find out):

-- The CD market is less efficient than the bond market. There were periods *this year* when you could get a 5.5% 5-year CD and a 5% 15-year mortgage. I took advantage of that.

-- There are two components to interest rates: inflation and real rates. Both look like they could be under pressure.

-- Short-term inflation will be impacted by rising energy costs and monetary policy.

-- Long-term inflation will be impacted by the boomer retirement party (read Scott Burns for more info).

-- Real rates are a function of supply and demand. Supply has been low in the corporate sector since corps have been hording cash rather than taking on debt. Demand has been strong in the treasury markets due to foreign governments propping up the dollar. Neither of these trends is sustainable.

-- Our treasury debt will be growing at a good clip. When supply goes up, rates go up too unless demand goes up as well. Where's the increased demand going to come from?

-- The fed has short-term rates below inflation. Unsustainable.

These are a few of my favorite reasons, but even I concede there's no guarantee. It's a small risk I'm willing to take.

[Helen]

"Covering the home & mortgage with the pension takes all the uncertainty out of the math."

Hi Nords,

Is your pension from the Government ? I have two, one as a former Digital Equipment employee and the other from my current job as a Federal Employee. The only one I count on in my retirement planning is the Federal pension.

If your pension is from a private company I'd say your taking on a big risk to pay off the mortgage with pension money.

Just my 2 cents.

-helen

Just paid off my home last Friday. I'm 59, have a couple more years in the present position I'm in and then intend to start another career (I'm lousy at doing the vacation thing). Paying off my primairy residence seems like a no-brainer. It gives me the flexibility (with two pensions plus Soc Sec in a few years) to do any one of a number of things, include ER! Great site by the way. Lots of information here I haven't seen elsewhere.

[Nords]

I agree that analysis tools could be better. I'd love an accurate predictor of future returns and interest rates. But even if Buffett's long-range returns are 6-7% for thirty years (he doesn't say it'd last that long) we'll still beat a 5.5% mortgage rate. And when rates are the lowest in 40 years, it seems reasonable to take out a 30-year mortgage at what could presumably be the bottom of that cycle. A small arbitrage over 30 years really adds up.

I'm not trying to say that this arrangement is good for every investor or every family. I am trying to say that paying off the mortgage is a bromide subject to financial & risk analysis. Instead of mindlessly declaring "Debt bad, payoff good", we should look at the amount of risk we're willing to take. In this case we've figured a way to wangle a chunk of money and play odds that are far better than any Vegas house. "Necessary"? That depends on your risk tolerance. We feel we have the basic expenses covered, and if we don't need the rest then it doesn't seem imprudent to take a long-term risk.

I think most financial recommendations are subject to harsher analysis under one's own lifestyle, savings, & spending parameters. But let me go into the nitty gritty in the next couple posts.

[Nords]

Quote: