What is the lowest rate you would aggressively pay down your mortgage? (US)

[FINate]

I answered 4% because this is currently the risk free rate. Why base it on the risk free rate? Because a fixed mortgage is risk free (in the sense that it will never change, not talking about the risk of nonpayment). So the closest thing from a risk profile is the risk free rate. Yes, I can get higher returns in equities, but these are higher risk.

A lot of this is contextual though. We hit our FIRE number 10ish years ago and are in a wealth preservation phase. A lot of our NW is already in stocks, so paying off our mortgage was part of our retirement strategy. Renting (or carrying a typical mortgage) on our house would cost us around $40k/year (aka imputed rent). Owning outright means we don't have this expense, plus there's no tax on the imputed rent. This means we can live on a very low income, and so according to the IRS we are in poverty even though we're wealthy. Not only does this mean our income and cap gains are not taxed, but we also get ACA subsidies and other tax credits. Most years we have zero withholding yet get money back on our tax returns.

Whereas if we were still in the accumulation phase I would probably be more included to invest aggressive in VTI instead of paying down a mortgage unless it was closer to 7-8%.

[Wintergreen78]

Here’s how I’ve looked at it, using FIRECalc:

For this example I’ll use a 260,000 mortgage, which works out to a $2,200/month payment, or $26,400/year PITI at 6.65%.
I’ll use $30,000/year as all other spending and assum I’ve got $1,260,000 invested.

Plug $1,260,000 and $56,400 into FIRECalc and I get: minimum $1,260,000, maximum $11,992,150, average $6,508,350
Plug $1,000,000 and $30,000 in and I get: minimum $546,203, maximum $6,585,050, average $2,717,054

You can get fancier and account for tax benefits, other spending changes, etc. My rate is 6.99. I’m in the fence about paying down aggressively.

Maybe I'm misunderstanding, but it seems like you've taken the taxes and insurance out of your spending in the pay-off scenario, but those would be ongoing expenses.  It's only the PI that goes away, not the TI.

Sorry I mis-spoke, but that is what I meant. $26,400 is only PI in this example, not PITI.

I agree with you completely - you need to remember that taxes and insurance still need to be paid after your mortgage is paid off. Also, I’m in California and am very aware that insurance costs could get much higher in the future.

[PMJL34]

Wintergreen, I believe you are flat out wrong....

with those numbers (1.26mil portfolio and 56,400 spend/year) inputed at Firecalc:

"Here is how your portfolio would have fared in each of the 124 cycles. The lowest and highest portfolio balance at the end of your retirement was $-1,379,793 to $6,612,793, with an average at the end of $1,903,289. (Note: this is looking at all the possible periods; values are in terms of the dollars as of the beginning of the retirement period for each cycle.)

For our purposes, failure means the portfolio was depleted before the end of the 30 years. FIRECalc found that 19 cycles failed, for a success rate of 84.7%."

For 1mil portfolio and 30k spend:

"Here is how your portfolio would have fared in each of the 124 cycles. The lowest and highest portfolio balance at the end of your retirement was $546,203 to $6,585,050, with an average at the end of $2,717,054. (Note: this is looking at all the possible periods; values are in terms of the dollars as of the beginning of the retirement period for each cycle.)

For our purposes, failure means the portfolio was depleted before the end of the 30 years. FIRECalc found that 0 cycles failed, for a success rate of 100.0%."

[dandarc]

@PMJL34  - The 26400 needs to not be inflation-adjusted and also fixed at 30 years in any calculator you use for this. That's on the 2nd page of firecalc.

Edit: Actually though . . . I get $-162,229 to $5,864,300, with an average at the end of $2,359,181 for a 99%+ success rate - a bit worse than the 30K / 1M scenario.

Edit again: 260,000 mortgage at 6.65% being $2200 means this is a 15 year term vs. 30. Taking that into account: The lowest and highest portfolio balance at the end of your retirement was $411,618 to $6,849,231, with an average at the end of $2,781,293.

Edit: Conclusion: appears the 6-7% spot is just about the midpoint for whether holding mortgage is good for end results vs. not.

[Wintergreen78]

@PMJL34  - The 26400 needs to not be inflation-adjusted and also fixed at 30 years in any calculator you use for this. That's on the 2nd page of firecalc.

Edit: Actually though . . . I get $-162,229 to $5,864,300, with an average at the end of $2,359,181 for a 99%+ success rate - a bit worse than the 30K / 1M scenario.

Edit again: 260,000 mortgage at 6.65% being $2200 means this is a 15 year term vs. 30. Taking that into account: The lowest and highest portfolio balance at the end of your retirement was $411,618 to $6,849,231, with an average at the end of $2,781,293.

Edit: Conclusion: appears the 6-7% spot is just about the midpoint for whether holding mortgage is good for end results vs. not.

Huh - I plugged that example in very quickly. I’m wondering if I had other income left in the second tab, or some other entry.

In any case - the exact numbers were not my point. I was just talking about a way to compare two options. Your actual spend rate, mortgage rate, and investment choices will determine what results you get.

I think the big benefit of using this as a tool is that it shows a range of possible outcomes. In general, the best case outcome is better by not paying off, but when I ran my personal numbers I wasn’t at 100% success if I kept my mortgage, but was at 100% when I paid it off.


Edit: Ha! I think I entered $5,640 for spend in the first example (or something close to that - the numbers almost but don’t quite match). So there is your solution, keep your mortgage, but just don’t pay it, and only spend about $450/month total. That will work great until you get foreclosed on.

[Wintergreen78]

This conversation has got me thinking about other ways to look at this/think about this. One way is to ask “if I pay off my mortgage, what is the probability I’ll be ahead in one year, in three years, or in ten years vs investing in the SP 500”. So, this would involve looking at the distribution of returns.

At 7% per year, one year returns are 7%, 3 year returns are 22.5%, and ten year returns are 96.7%. Looking at these charts it appears basically a coin flip whether the SP 500 will exceed 7% in any year, or 22.5% over three years. Interestingly enough, it seems like the odds are it would be up less than 96.7% over a ten-year period. The long-term average returns of the SP500 are increased by the big returns far out on the right of the curve, but any particular ten-year period isn’t likely to beat 7%.

Here’s the charts I found with just a little searching, I’d be interested if anyone has looked at it more seriously this way, or has a better source for data.

https://klementoninvesting.substack.com/p/the-distribution-of-stock-market

[nereo]

This conversation has got me thinking about other ways to look at this/think about this. One way is to ask “if I pay off my mortgage, what is the probability I’ll be ahead in one year, in three years, or in ten years vs investing in the SP 500”. So, this would involve looking at the distribution of returns.

I think you miss an important component if you try to view it this way. Let’s say you are comparing putting an additional $100k towards investments or towards your mortgage. If it’s put towards your home that money is tied to that intangible asset. The easiest way to access it is thru a HELOC or refinancing, which itself very much depends on the current market. Three years ago I could have borrowed against my home at under 4%. Now it’s around 8%. It’s also quite a bit more involved than selling off shares.

If I had instead invested in a brokerage it’s easier to access but still tied to another market. If it’s gone up, great. But even if it fell the losses might be offset to some degree by tax loss harvesting.

Of course the ultimate goal of accelerated mortgage payoffs is to rid oneself of a sizeable expense. That’s all fine and good provided you reach that milestone. Until you do you have substantially increased your risk, as you have the same expense but less savings should the SHTF.

[Wintergreen78]

This conversation has got me thinking about other ways to look at this/think about this. One way is to ask “if I pay off my mortgage, what is the probability I’ll be ahead in one year, in three years, or in ten years vs investing in the SP 500”. So, this would involve looking at the distribution of returns.

I think you miss an important component if you try to view it this way. Let’s say you are comparing putting an additional $100k towards investments or towards your mortgage. If it’s put towards your home that money is tied to that intangible asset. The easiest way to access it is thru a HELOC or refinancing, which itself very much depends on the current market. Three years ago I could have borrowed against my home at under 4%. Now it’s around 8%. It’s also quite a bit more involved than selling off shares.

If I had instead invested in a brokerage it’s easier to access but still tied to another market. If it’s gone up, great. But even if it fell the losses might be offset to some degree by tax loss harvesting.

Of course the ultimate goal of accelerated mortgage payoffs is to rid oneself of a sizeable expense. That’s all fine and good provided you reach that milestone. Until you do you have substantially increased your risk, as you have the same expense but less savings should the SHTF.

I get that - until you pay off your mortgage you are still at risk of foreclosure and your home’s value is not as liquid as an etf or mutual fund.

In my case I could cash out investments and pay off my mortgage tomorrow. My risk of foreclosure is basically zero. I also have plenty of liquid assets, so relying on home equity for extra cash isn’t likely either.

So, I think it reasonable to think about your odds of being ahead in 3/5/10 years as a way of helping with decision making. I also think this isn’t that critical of a decision. If you’ve bought a house you can afford, if your income and expenses work for you, and if you’re saving and investing reasonably, you’ll be fine either way.

[Telecaster]

It might be more useful to look at the range of outcomes over your investing horizon.  So for example, if your horizon is 30 years, who would paying off the mortgage affect that?   

[JAYSLOL]

Since I’m in Canada, and we don’t have the option of a fixed rate beyond 5 years, I plan to pay extra towards my mortgage regardless of interest rate, for risk management because of the potential that interest rates could skyrocket at some point. I know that looks very unlikely at the moment, it looks like rates will continue to drop, but over a 25 year mortgage, it certainly could happen, and the last thing I need is to be holding a mid-6 figure mortgage and have rates go double-digits like they did in the 80s. 

[Dicey]

Here's another nuance:

DH and I are FIRE. We occasionally buy ugly houses and make them safe and pretty again. After a couple of years of fruitless searching, we're offering on a property tomorrow. The market's extremely competitive, so we're offering all cash, with a quick close instead of overbidding.

Because we carried long, low mortgages for many years, we have a shitload in equities now. We don't want to sell anything, so if our bid wins, we will just borrow against our investments.

We never could have done this if we had prepaid mortgages instead of shoveling the extra cash into equities first.

OP, and everyone else, you're always welcome over at the DPOYM Club. It gets kind of quiet over there, so new conversations are appreciated. All are welcome.

[Wintergreen78]

It might be more useful to look at the range of outcomes over your investing horizon.  So for example, if your horizon is 30 years, who would paying off the mortgage affect that?

This is exactly what firecalc shows you. In my case the firecalc best case scenario is not paying off my mortgage, which is 6.99 percent. In the worst case scenarios, I’m better off paying off my mortgage. For me, my worst case after 30 years without paying off my mortgage I’m left with about $200k. With paying off my mortgage I’m left with about $1.1 million in the worst case scenario. So, both count as 100% success, but paying off my mortgage makes the worst-case scenario much better.

So, if you make your decision about when to retire based on your worst-case outcome from Firecalc, or the 4% rule, or any of the other “range of outcome” approaches to thinking about retirement risk, and if your mortgage rate is relatively high, you might find that paying off your mortgage moves up your retirement date.

[neo von retorch]

I've got over 29 years of dead pledge (aka mortgage) at 5.95%.

In my custom spreadsheet, having principal + interest (PI of PITI) "forever" is worse than paying it off, assuming 6.8% return on investments every year -- because I need that much more invested.

$446K mortgage translates (currently) to $2830 / month in PI. My custom spreadsheet uses trailing 12 months expenses (don't actually have an accurate read since my most recent move.)

If you're following along, the biggest flaw in my math is not that I assume a 30 year mortgage is forever. The biggest flaw is that mortgage interest will slowly decrease over time. So PI will eventually be smaller. (But there is a problem, because a mortgage keeps forcing me to convert cash into equity. This is especially problematic in my current, hopefully temporary, situation of being unemployed and not yet ready for F.I.R.E. according to the 4% SWR.)

So my forecast says I've reached 62% of my nest egg for having my existing expenses, but 70% if I sell investments to pay the mortgage. (Another naive assumption here is that I wouldn't pay capital gains taxes, but alas, reality bites.)

Overall though the numbers are close, and I hope to someday refinance to a lower rate. In the meantime, I hope my investments continue to beat that 5.95% mortgage rate!

[MrGreen]

We just put a chunk of change down on our 6% mortgage and had it recast. We're left with slightly more than 50% of what we paid for the house. I'm not very optimistic on rates going below 5% so I doubt we'll see an opportunity to refinance. We're playing both sides a little this way.

[Dicey]

Here's another nuance:

DH and I are FIRE. We occasionally buy ugly houses and make them safe and pretty again. After a couple of years of fruitless searching, we're offering on a property tomorrow. The market's extremely competitive, so we're offering all cash, with a quick close instead of overbidding.

Because we carried long, low mortgages for many years, we have a shitload in equities now. We don't want to sell anything, so if our bid wins, we will just borrow against our investments.

We never could have done this if we had prepaid mortgages instead of shoveling the extra cash into equities first.

OP, and everyone else, you're always welcome over at the DPOYM Club. It gets kind of quiet over there, so new conversations are appreciated. All are welcome.

Update: our offer was not accepted, or even countered. We're relieved, because it's an otherwise busy time of year. At least we can watch for other listings in this complex, which wasn't on our radar, as it's slightly beyond our 5 miles from home rule.

I'm loving this conversation. Thanks to all who have participated.

[nkt0]

We are in a similar situation as OP but a little different. Apologies if this is hijacking the thread.

$780k in retirement accounts and post-tax brokerage accounts
$260k in liquid gains from the sale of our previous residence
$260k mortgage at 5.875% on our new primary residence (~$1500/month PI)
$40k annual expenses without PI
$58k annual expenses with PI

Scenario A: We pay off the mortgage in full and carry on with plowing the PI into savings
Scenario B: We invest the windfall from the sale of our house and save less on an ongoing basis

When i plug these numbers into cFireSim and FireCalc, both show the advantage to paying off the mortgage in full.

For example FireCalc says:

Scenario A: The lowest and highest portfolio balance at the end of your retirement was $-1,595,443 to $3,633,084, with an average at the end of $762,565.… FIRECalc found that 35 cycles failed, for a success rate of 71.8%.

Scenario B: The lowest and highest portfolio balance at the end of your retirement was $-2,807,463 to $4,421,510, with an average at the end of $635,341.… FIRECalc found that 46 cycles failed, for a success rate of 62.9%.

cFireSim has similar results. Scenario A success rate 75.2%; Scenario B success rate 70.4%.

I guess in a way this is a question of risk tolerance. Do i want a guaranteed return of ~7% from my house or a variable return of ~9% in the stock market. I'm oversimplifying but i think this captures my situation. Thoughts/critiques?

[dandarc]

That's not quite right on the data entry - you don't just add the $18K to annual expenses because the mortgage P&I does not inflate (assuming this is a fixed-rate mortgage).

Edit: results aren't shifting in firecalc from that additional spending / income screen vs. without the additional spending. Unclear why that is to me.

[joemandadman189]

For me it'd probably more a function of personal circumstances and what proportion is taken by housing costs, and more specifically housing interest payments.

I woudn't rely on just using a a straightforward "overpay x% when rates are at y%" rule because context matters, and real stock market returns don't really have any correlation to rates anway - which makes sense, as stocks know what the risk free rates is and price themselves accordingly (most of the time).


That said, if I was holding a large mortgage (which I am) then I would be less inclined to hold bonds in my investment portfolio, as the mortgage gives me plenty of interest rate risk as it is.

This is where we are, mid 40s with a pretty new mortgage @ 5.5% on a 30 year term, to us not having the mortgage payment in our monthly expenses would be a great help in stepping back, hence our interest in making extra payments, lump sum payments while also maxing out retirement and contributing to brokerage accounts.

[Tigerpine]

I'm starting to think that the best plan for us is to spread out extra mortgage payments so that we pay off the mortgage early, coinciding with our planned retirement date (4.75%).  Over the long term, we might be able to get a higher return in the stock market, but it also increases the SORR risk to have a higher monthly liability once we hit retirement, since that is a sum that must be paid at risk of losing our residence.

[dandarc]

That's not quite right on the data entry - you don't just add the $18K to annual expenses because the mortgage P&I does not inflate (assuming this is a fixed-rate mortgage).

Edit: results aren't shifting in firecalc from that additional spending / income screen vs. without the additional spending. Unclear why that is to me.

@nkt0 - Figured out what I was doing wrong - firecalc ignores that additional income / spending data if it starts in a year in the past, and I entered 2024 (default is 2028, so does need to be changed to do this analysis).

So I get 33 failures for scenario B (DPOYM) vs. 35 for Scenario A (POYM) - similarly slightly better on the range of outcomes. That tracks better since upthread found that "breakeven" from this calculator was in the ~6% range - 5.875 is getting awfully close to that, so makes sense that results would be close.

[nkt0]

@nkt0 - Figured out what I was doing wrong - firecalc ignores that additional income / spending data if it starts in a year in the past, and I entered 2024 (default is 2028, so does need to be changed to do this analysis).

So I get 33 failures for scenario B (DPOYM) vs. 35 for Scenario A (POYM) - similarly slightly better on the range of outcomes. That tracks better since upthread found that "breakeven" from this calculator was in the ~6% range - 5.875 is getting awfully close to that, so makes sense that results would be close.

Ah that's a good catch. I forgot that the PI doesn't inflate (although technically i got a 3/1 ARM so it _could_ inflate or deflate!). Still seems pretty close between the two scenarios.

[JupiterGreen]

Haven't had a mortgage in a minute but will be closing on a house this month or next. 30yr 330k mortgage at 6.375% fixed I believe our strategy will be to pay that dollar eater off unless rates go down and we can refinance, not holding my breath. We'll have a bit over 200k in home equity at closing, otherwise we have about 1.8m in investments.

We might still feel like paying off a mortgage if it were in the 5% or even 4% range, wouldn't be in any rush with 3% and I'd leave alone 2% mortgage rate or lower. 

[dandarc]

@nkt0 - Figured out what I was doing wrong - firecalc ignores that additional income / spending data if it starts in a year in the past, and I entered 2024 (default is 2028, so does need to be changed to do this analysis).

So I get 33 failures for scenario B (DPOYM) vs. 35 for Scenario A (POYM) - similarly slightly better on the range of outcomes. That tracks better since upthread found that "breakeven" from this calculator was in the ~6% range - 5.875 is getting awfully close to that, so makes sense that results would be close.

Ah that's a good catch. I forgot that the PI doesn't inflate (although technically i got a 3/1 ARM so it _could_ inflate or deflate!). Still seems pretty close between the two scenarios.

Yeah - if you have no confidence in being to re-up the arm at a similar or lower rate, then paying off makes sense. Was simulating a 30 year fixed rate in the calculator.