Convince me to refinance a paid off home

[themagicman]

My personal residence is paid off. With rates dropping lower, I am considering doing a cash out refinance and investing the funds. I understand the market is more than likely going to outpace my 3.8% rate and that I would end up ahead in the long run. The part that makes me hesitate is from a retirement perspective, I do not see how it makes sense. Please let me know if I am missing something with this!

I am a couple years away from retirement. Lets say I refinance $300k of my house and take that and invest it in the market. Even not considering closing costs I feel I still come out behind.

I take that $300,000 and invest it and am able to pull out $12,000 a year (based on 4% and that might even be too high but lets assume 4% swr) .
But my mortgage costs are $1,398 a month or $16,776 a year. So I would need an extra $4,776 a year to be able to retirement which is another $119,000 saved up. This would push my retirement date back two years to do this. Are people doing it because they are going to be better off in 30 years when the mortgage is paid off? Am I missing something?

[honeybbq]

Sorry, no can do. I wouldn't do it given your timeline to retirement.

[dandarc]

You're missing that a large chunk of that mortgage payment over 30 years is going to principal. And that there is an end date on it.

Mortgage is not forever - including the full P&I in the "25x expenses" leads to wrong conclusions. Or more nicely - you're making a more conservative decision than necessary.

25x of other, ongoing expenses + enough to pay off the mortgage if you want is a better way to think about it.

Similar to college if you're doing that for your kids. That $20K / year / kid is not a "forever expense" - 4-10 years and it is done. So planning around needing to pay that $20K every year forever is not doing yourself any favors.

[dandarc]

Also if you get convinced, tell us how the point was made - I need to know that that for getting my wife on board on this issue (1/3rd your numbers - LCOL area win for us there).

[Altons Bobs]

If the market tanks for a few years, will you have enough money to live and still pay for your mortgage or do you have to go back to work?

[dandarc]

Having another $300,000 at the start surely helps with "have enough money to live and still pay for your mortgage".

[FIPurpose]

For retired people, the question revolves around your personal tax situation. Since you'll likely be pulling from an IRA in retirement (though you're only 28 so, not sure how this is true for you), you'll be increasing your taxable earnings in order to pay for the mortgage. So let's say for example the extra mortgage payments bump you up to 10% bracket.

Mortgage interest 3.8%
Income tax 10% (Or you could calculate your difference in effective net rates.) So your effective rate might go up 2-3%.
(lost ACA subsidies?) Another potential 2-3% worth

In retirement the trade-offs change. You have much more control over your tax situation, and often taking the lower tax situation far out weights the benefits of a mortgage arbitrage.

[themagicman]

You're missing that a large chunk of that mortgage payment over 30 years is going to principal. And that there is an end date on it.

Mortgage is not forever - including the full P&I in the "25x expenses" leads to wrong conclusions. Or more nicely - you're making a more conservative decision than necessary.

25x of other, ongoing expenses + enough to pay off the mortgage if you want is a better way to think about it.

Similar to college if you're doing that for your kids. That $20K / year / kid is not a "forever expense" - 4-10 years and it is done. So planning around needing to pay that $20K every year forever is not doing yourself any favors.

But isn't the cash flow portion of it all that matters? Unless you are saying that you could always sell the house/refinance again to recoup the principle put into it?

[themagicman]

If the market tanks for a few years, will you have enough money to live and still pay for your mortgage or do you have to go back to work?

Depends on when I retired. If I refinanced and continued to retire at my current date then, yes I would assume have to go back to work

[themagicman]

Having another $300,000 at the start surely helps with "have enough money to live and still pay for your mortgage".

I guess the point I am saying is, lets say I spend $40k a year, so I will need $1million to retire. If I refinanced and still retired at the same time then I would have $1.3 million and be able to pull off $52,000 a year which would not be enough to pay the mortgage plus keep my $40k a year lifestyle.

[dandarc]

Having another $300,000 at the start surely helps with "have enough money to live and still pay for your mortgage".

I guess the point I am saying is, lets say I spend $40k a year, so I will need $1million to retire. If I refinanced and still retired at the same time then I would have $1.3 million and be able to pull off $52,000 a year which would not be enough to pay the mortgage plus keep my $40k a year lifestyle.

You're taking the 4% rule too literally. In retirement, you pull what you need to meet your cash-flow needs. You need $56K to pay for lifestyle + the mortgage, so that is what you'd draw. Also note the mortgage P&I does not increase with inflation - your other spending does, at least it does in the research that backs up the 4% rule. That $16K of mortgage payments in year 30 is a lot cheaper than the first year's $16K.

You have 18 years worth of mortgage payments in that extra 300K, and you're expecting to make a net of $15K or more each year, plus the compound return on that over the life of the mortgage - that adds up quickly. That's like having a minimum-wage worker just giving you all of there income every year. Even if year 1 is a 50% correction, you've got 9 years of mortgage payments by having more money on hand, and in all liklihood even if that happens you'll have some big returns in the next few years that will off set it. The risk is not very high and the reward is potentially huge.

Even if the mortgage is "forever" - meaning you are willing to refinance any time it makes sense to do so, 4.3% isn't that different than 4% in terms of safety. Take a look at the "broke vs. dead" charts. We agonize over the sliver of "broke", but that's actually a really small liklihood. Strongly considering pulling the plug when we get to 5%, and taking out a mortgage on our house will help us get there faster in all likelihood.

Caveat - depending on how your assets are structured, the requirement for more cash flow could impact your tax situation in a fairly large way - that's more of a thing to worry about than "4% won't pay my P&I!", but assuming the $16K doesn't have a huge tax impact for you the best play is to have a low, fixed-rate mortgage even during retirement. Take as much cheap money as the bank will give you.

I'll finish with my "shit, that was dumb and I need to fix this" story. I paid my house off in 2016 - that was a mistake. Would have $40-50K more today in investments, net of the mortgage balance, if I had been smarter in the 2014-16 time range. Times 3 for your example - my house cost $125K when we bought it, so an 80% mortgage would only be $100K. That's in just 3 years. That's the kind of money we're talking about with these decisions - 6 figures without a doubt and possibly 7 figures in 30 years. #1 thing on my list to do once we move back into the house in October is to re-finance and fix this massive problem going forward. I'm lazy and this requires my wife to be on board is why it wasn't done earlier.

[PathtoFIRE]

I'd only consider the mortgage interest as an addition expense, so rather than $16,776 per year of additional expense, it'd instead start out at an annualized $11,400 in the first month, and steadily drop over the next 359 months (assuming 30 year mortgage).

So your expenses would be $51,400/yr (and the mortgage interest portion of that would drop a little every month), or $1.285M needed, covered by your now $1.3M retirement portfolio. Yes, you would be drawing down $56,776 from your portfolio that first year, but about $5,400 of that initially would really just be a conversion from the stocks/bonds you had in your portfolio into real estate (1 undiversified property, yes, but you'd certainly have more diversification than if you kept the house paid off with all $300k tied up in it).

In other words, as long as the mortgage rate is less than or equal to your withdrawal rate, the numbers would match up, with the added complexity that you're obligated to convert a small portion of your stocks/bonds into real estate each year (initially 0.4% of your total portfolio per year, ramping up to 1.2% per year by the end of the 30 year mortgage, assuming your portfolio stays around $1.3M; obviously history says its more than likely to grow over 30 years even after withdrawals). And you would likely still come out ahead even if your mortgage rate exceeds withdrawal rate, given the likelihood that you're portfolio grows faster than your withdrawal rate.

[Telecaster]

Having another $300,000 at the start surely helps with "have enough money to live and still pay for your mortgage".

I guess the point I am saying is, lets say I spend $40k a year, so I will need $1million to retire. If I refinanced and still retired at the same time then I would have $1.3 million and be able to pull off $52,000 a year which would not be enough to pay the mortgage plus keep my $40k a year lifestyle.

As a proud member of the DNPOYM Club I agree with your analysis.  You don't (probably) don't want to increase your WR early in the term due to sequence of returns risk.  You'd be going from 4% to 4.3%, which isn't terribly risky, but it is riskier.    The upside is that you are much more likely to wind up wealthier at the end of 30 years than you are to wind up broke.   

Others have mentioned that due to inflation, your WR will actually go down over time (because the mortgage is fixed).  And that's absolutely true.  But you will still have a higher initial WR, which is the riskiest time. 

[Telecaster]

I'd only consider the mortgage interest as an addition expense, so rather than $16,776 per year of additional expense, it'd instead start out at an annualized $11,400 in the first month, and steadily drop over the next 359 months (assuming 30 year mortgage).

So your expenses would be $51,400/yr (and the mortgage interest portion of that would drop a little every month), or $1.285M needed, covered by your now $1.3M retirement portfolio. Yes, you would be drawing down $56,776 from your portfolio that first year, but about $5,400 of that initially would really just be a conversion from the stocks/bonds you had in your portfolio into real estate (1 undiversified property, yes, but you'd certainly have more diversification than if you kept the house paid off with all $300k tied up in it).

I don't think that's valid though.   Yes, you get to keep the principle, but it doesn't go back in your stache.  It is locked up in your house where you can't spend it.  On the flip side, you still have to come up with the full payment each month, and that does come from your stache.   So you still have to use the 4% Rule (or whatever % you like) to calculate the SWR for all your expenses including the mortgage payment.     

[dandarc]

Gonna get myself banned if I keep posting to this thread, but PathtoFIRE has the most right way of looking at it presented so far. The P part of P&I is not really an expense.

[Telecaster]

Gonna get myself banned if I keep posting to this thread, but PathtoFIRE has the most right way of looking at it presented so far. The P part of P&I is not really an expense.

You can call it whatever you want, but the money for the P part comes out of your stash each month, and that increases your withdrawal rate.  And the money you pay for the P part doesn't go back into your stash, so you can't reap the investment rewards or otherwise spend it. 

[dandarc]

Gonna get myself banned if I keep posting to this thread, but PathtoFIRE has the most right way of looking at it presented so far. The P part of P&I is not really an expense.

You can call it whatever you want, but the money for the P part comes out of your stash each month, and that increases your withdrawal rate.  And the money you pay for the P part doesn't go back into your stash, so you can't reap the investment rewards or otherwise spend it.

It is not as easy to get the money out of your house as it is to get it out of your 'stache, but it is far from impossible. You could borrow against the house - that should be part of your plan if rates remain low anyway. There is also the "sell the house" option.

Think about what you're saying though - put $300K into your house right now so you'll need less in your portfolio and avoid putting $300K into your house over the course of 30 years. That doesn't pass the smell test for the "paid off house is better for liquidity" argument.

[PathtoFIRE]

Thanks Telecaster, you're right that you're moving assets from more liquid to less liquid, but as dandarc mentioned, the OP already has a large portion of their net worth (or expected net worth, I temporarily forgot they aren't quite FI yet). I think the point is that the principal portion that you pay each month is really just a forced asset allocation change. It's very small, but persistent and will eventually move the needle, but over 30 years, and we know that really it's those first 5-10 years that are important for sequence of returns risk.

But you do infer a good point, it is a forced selling of stocks/bonds, at least each year, maybe more often depending on how you cash flow FIRE, so it should be pointed out that that carries a slight risk of selling low. But in the context of the overall net worth, it's small, and of course if the forced low selling lasts more than a few quarters, then there must be major economic problems, and likely the OP would be feeling pressure irregardless of whether their house is paid off or leveraged.

[Telecaster]

Think about what you're saying though - put $300K into your house right now so you'll need less in your portfolio and avoid putting $300K into your house over the course of 30 years. That doesn't pass the smell test for the "paid off house is better for liquidity" argument.

No, he's taking $300K out of his paid off house right now.  That increases his WR at the beginning of his retirement.  Full stop.  Increasing your WR increases your sequence of returns risk.   For the OP, the only reason to do this is if he wants to wind up filthy rich in 30 years--which is the most likely result.  But he's taking a correspondingly higher risk of a busted retirement as well.  Most people would prefer to avoid the busted retirement, which is why the 4% was discovered in the first place. 

Now, if the OP was ten years prior to retirement or ten years after, we'd be having a different conversation.   The issue the effect on his initial WR.   For the record, I have no problem with someone taking a 4.3% vs. 4% WR.  Just recognize there is additional risk. 

[bacchi]

Think about what you're saying though - put $300K into your house right now so you'll need less in your portfolio and avoid putting $300K into your house over the course of 30 years. That doesn't pass the smell test for the "paid off house is better for liquidity" argument.

No, he's taking $300K out of his paid off house right now.  That increases his WR at the beginning of his retirement.  Full stop.  Increasing your WR increases your sequence of returns risk.   For the OP, the only reason to do this is if he wants to wind up filthy rich in 30 years--which is the most likely result.  But he's taking a correspondingly higher risk of a busted retirement as well.  Most people would prefer to avoid the busted retirement, which is why the 4% was discovered in the first place. 

Now, if the OP was ten years prior to retirement or ten years after, we'd be having a different conversation.   The issue the effect on his initial WR.   For the record, I have no problem with someone taking a 4.3% vs. 4% WR.  Just recognize there is additional risk.

The $300k can be considered separately since it doesn't need to exist beyond 30 years. It also has a different floor than the 4% rule; that is, the withdrawal is always $16776 and never 4% + inflation each succeeding year.

Then it becomes a question of whether the $300k can safely handle a withdrawal of 5.5% over 30 years.

The % withdrawal is also wrong since the withdrawal amount never changes. It doesn't look like there's a good cfiresim mode for this (even non-inflation adjusted mode somehow withdraws $33k/yr ???).

[dandarc]

What Bacchi just said.

Looking at CFiresim a non-inflation adjusted $16,676 spend against a $300K starting balance goes to 0 or less only a couple of times ever in history - actually fails less over 30 years than the 4% inflation adjusted withdrawal rate we're usually talking about.

So about 98.5% of the time you're better off by at least a little and often by a lot by having the mortgage and only if you were unlucky enough to retire in one of the worst 2 years since 1871 you'd be worse off.

[dandarc]

The % withdrawal is also wrong since the withdrawal amount never changes. It doesn't look like there's a good cfiresim mode for this (even non-inflation adjusted mode somehow withdraws $33k/yr ???).

Worked fine for me - I put 300K for the portfolio and chose non-inflation adjusted spending of 16,776.

Maybe the spending chart is confusing? Inflation adjusted spending level - that deflationary end of the 1800's looks strange.

[themagicman]

Think about what you're saying though - put $300K into your house right now so you'll need less in your portfolio and avoid putting $300K into your house over the course of 30 years. That doesn't pass the smell test for the "paid off house is better for liquidity" argument.

No, he's taking $300K out of his paid off house right now.  That increases his WR at the beginning of his retirement.  Full stop.  Increasing your WR increases your sequence of returns risk.   For the OP, the only reason to do this is if he wants to wind up filthy rich in 30 years--which is the most likely result.  But he's taking a correspondingly higher risk of a busted retirement as well.  Most people would prefer to avoid the busted retirement, which is why the 4% was discovered in the first place. 

Now, if the OP was ten years prior to retirement or ten years after, we'd be having a different conversation.   The issue the effect on his initial WR.   For the record, I have no problem with someone taking a 4.3% vs. 4% WR.  Just recognize there is additional risk.

The $300k can be considered separately since it doesn't need to exist beyond 30 years. It also has a different floor than the 4% rule; that is, the withdrawal is always $16776 and never 4% + inflation each succeeding year.

Then it becomes a question of whether the $300k can safely handle a withdrawal of 5.5% over 30 years.

The % withdrawal is also wrong since the withdrawal amount never changes. It doesn't look like there's a good cfiresim mode for this (even non-inflation adjusted mode somehow withdraws $33k/yr ???).

Yep, this is what I was not seeing. Looking at it separately and having a non inflation adjusted spending plan and seeing the percentage of success.

[themagicman]

What Bacchi just said.

Looking at CFiresim a non-inflation adjusted $16,676 spend against a $300K starting balance goes to 0 or less only a couple of times ever in history - actually fails less over 30 years than the 4% inflation adjusted withdrawal rate we're usually talking about.

So about 98.5% of the time you're better off by at least a little and often by a lot by having the mortgage and only if you were unlucky enough to retire in one of the worst 2 years since 1871 you'd be worse off.

This makes sense to me. It allows for a higher withdraw because it is not inflation adjusted and this pot of money is looked at "separately" . I was not thinking about it this way