FREE NEWSLETTER
Go to main Forum page »
Jonathan used to chastise me for saying that I thought a good goal for retirement income was to replace 100% of base pay or salary- excluding overtime and any form of bonuses.
I was making a suggestion, opinion, not suggesting a requirement because given most people don’t reach that goal, it is obviously not required even while desirable.
However, that’s the way Connie and I live. In fact, between my pension and our combined social security, gross income is a bit more than 100% replacement- based on my salary in 2008. Ah, there’s the rub. The value of that income has been eroded 49% so are we actually now living on 50% replacement. I ask myself, where would we be if we started at 75%?
Who could be neutral in this debate? Who else but AI, so I asked ChatGPT?
The answer was most people need 70% to 80% – conventional wisdom from AI. It said, “Most People Don’t Need 100% Income Replacement in Retirement.”
So, I asked why. Here are the reasons given.
1. Work-related expenses disappear
When you retire, you stop paying for:
This alone can reduce required income by 5–10%.
Well maybe, but just as maybe, some of that may not apply. Dues, fees, commuting costs or childcare didn’t apply to us.
2. You’re no longer saving for retirement
This is one of the biggest reasons 100% replacement isn’t necessary. A portion of income likely goes into:
In retirement, you stop saving — you’re withdrawing instead. That frees up 10–20% of your gross income.
Again maybe, or perhaps probably, but anything above 10% may be a stretch for many people. We never saved above 10% unless you count my non-cash compensation. Plus in retirement we still save – in 529 plans (by choice) and into an emergency fund.
3. Taxes are usually lower in retirement
Retirees often pay less in taxes because:
For high earners, losing the 6.2% Social Security payroll tax alone makes a big difference.
Again maybe. But RMDs can hurt. Tax laws change. Investment income can vary. In some cases the FICA taxable wage base may be a minority portion of wages.
4. Mortgage is gone
For many households, the mortgage is 20–30% of spending.
That could be a biggy – if the mortgage ends just before or upon retirement. In our case the mortgage was paid more than five years before I retired so there was no impact.
5. Kids and major obligations are gone
People often spend less because:
Even affluent households see expenses drop by 10–15% here.
I have two problems with this assumption. There may not be children. Plus by the time a person retires these expenses probably have long been gone, unless you have children in college when your 65.
There it is. AI expertise on retirement income. The question is how will all this apply to your situation.
It’s still my opinion and only my opinion than as a buffer against inflation and unpredictable events that ..%…is a desirable goal.
So AI has spelt it out for you but you’ve found a reason to argue against every single point.
How about as in the many, many times you’ve tried to get your pet theory to stick to almost universal disagreement you, for once, consider looking at the problem from the other side – that it is not at all what your income in employment looks like that matters: it is what your expenditure is that matters?
A $ is a $. It doesn’t matter one jot whether it came from income, capital, interest, dividends, betting wins, found on the sidewalk when you spend it.
If you insist on having some rules re income I’d suggest that “if you’re still spending every cent of income you’re not really in a position to retire” is a better and more useful one for the majority of people. If people don’t understand the variable applications of their money then replacing income doesn’t help them – they are still living hand to mouth.
Tracking expenses in the years before retirement yields an answer you may need, but is the work necessary?
Look at gross income, subtract taxes to get net income, subtract all savings and haven’t you spent the balance?
Then subtract major expenses guaranteed to go away upon retirement. Isn’t that a valid starting point for income needed?
You could add new spending for things you didn’t spend before retirement like significant travel.
Why do people feel more comfortable tracking expenses in detail?
Have you ever watched any “personal finance guru sorts out people’s lives” type show?
Of course your method of calculating total expenses works but it doesn’t really provide any insight. The trope for those type of shows is the patients always say “We earn X but at the end of the month there is nothing left/overdrawn – where is it all going?”. Then the formula is to track expenditure for a month and present the insights and opportunities for easy wins back to them.
You clearly have more income than you need by the amount you give away. So your rule doesn’t even work for you even at a high cost time of life. Above a certain level of security people clearly have the ability to trade future income for time in the present, just as they, in their working lives have sacrificed time for current and (hopefully) future financial needs. Far better to encourage people to do their own planning and trade offs around that then hectoring them with some arbitrary “rule” that you’ve devised because you have an aversion to some fairly basic maths.
I would tend to agree with Chat GPT. Not everyone’s income is based on straight salary. For much of my career I had a base salary with monthly commissions paid based on my monthly production. My annual commissions were usually 5-8x of base salary (and sometimes more). We usually live on about 3x salary, and saved the rest. Our approach to income needed for retirement was to estimate/categorze our expenses for 3-4 years before retirement, and then determine what expenses would change during retirement, along with adding in a healthy travel budget. 8 years retired now, and our projections were generally spot on. The main challenges were budgeting healthcare premiums and out of pocket medical expenses for several years prior to Medicare enrollment. We generally estimated high for that line item in the budget, and were usually happy to have extra $ in the budget not being utilized for medical premiums/copays/deductibles. An intangible with the high cost health plans was being able to fund HSA accounts which gave us generous tax credits. Those accounts are now being used to fund our Medicare premiums. Those premiums should be covered by the HSAs until around age 80 or so.
I mostly agree with Mark.
In my view, the goal is not X % of my pre-retirement income (however defined), nor X % of my pre-retirement spending. Rather, is the best estimate I can make of my IN-retirement spending, plus some margin of safety.
My pre-retirement spending is the starting point. Then subtract things I won’t spend on anymore, but increase or add new things I will. There’s the basic estimate.
Add the margin of safety that lets one sleep at night. Then determine wither what guaranteed income assures it, or the portfolio needed to generate it (in the latter case, maybe add more margin of safety because it’s not guaranteed).
There you have it. The goal is some combination of guaranteed income and/or assets that should provide it. Yes, it requires making some estimates and doing some math, and is harder than saying X % of income. In my opinion it is also better.
Regardless of how we get there, our income in retirement will be some percentage of pre-retirement income. I don’t know how anyone can accurately predict they will need 70-80% replacement, but I do know if 100% is achieved, not much else matters. There may be excess income under some circumstances, but unlikely to be a shortfall and likely a hedge against inflation based on my own experience at least.
I’m interested to know if I’m the outlier here, but 15 years after retirement, I’m spending more than twice what I spent before retirement. Yes, lots of it is on family vacations, eating out, etc.
I don’t know about twice, but certainly not less including discretionary spending and that’s the way I wanted it.
I was trying to think yesterday about this and it hurt my head. It is no wonder a lot of people retire and hope for the best. We did a lot of catch up in the last 10-12 years before retirement. It was when our income was the highest. It was when the kids were on their own and college was over. It was when we moved to a lower cost of living area and bought our retirement home. It was when we paid that home off so we wouldn’t have a mortgage in retirement. It was when I was able to save a good deal. I am guessing there are a lot of people like us. Chris
I’m with Mark on this one. I’m looking at expenses, not current income. At the moment, we save 36% of our gross income, and another 14% goes towards college fees.
Once that goes away, our fixed monthly expenses would easily be covered by Social Security alone, so anything we pull from our portfolio would be for travel and other extras.
I’m sure I could spend a current year’s income plus SS if I had to (living the high life!), but it’s certainly not necessary.
Surely you know your situation is a considerable outlier. Saving 36% is quite exceptional especially given your basic living expenses can be covered by SS benefits alone. That’s an equation I just can’t comprehend, but more power to you.
My net SS payment doesn’t cover our HOA fees and property taxes. Net meaning after income tax and premium withholding.
Can you not see that’s because you’ve chosen to live in a high cost area? I suspect many many retirees would be horrified if they didn’t have the baseline security of basic living expenses covered by SS (or alternate pension equivalent).
In my opinion, targeting a percentage of income replacement in retirement is the wrong approach—it’s asking the wrong question entirely.
The real problem you should be solving is determining how much income you actually need. The better approach is to track your spending in the years leading up to retirement—getting that big picture view of your actual expenses—and then work backwards to calculate the asset value required to fund that amount. Essentially, it’s a bottom-up approach rather than a top-down one.
I agree with your approach and did essentially the same thing. It seems logical and perfectly tailored to our spending habits. Anyone can do this.
My wife and I are the same age. My goal was to retire at 65 (she had already retired 4 years earlier) and for us to continue consuming after-tax cash at the same rate, once retired. I figured we’d spend more on some things, less on others but maintain the same overall consumption rate, adjusting for inflation.
I calculated the gross amount of cash plus SSA benefits needed to provide the desired amount of after-tax cash. This would be our withdrawal amount for year one.
Bill Bernstein observed that the two most important determinants of successful portfolio withdrawal strategies are “burn rate” and horizon. With this in mind, I chose a horizon of 30 years and a burn rate of 3%. I multiplied the withdrawal amount needed in the first year by 33 to determine how much I needed to accumulate.
I recalculate this annually using the same horizon and burn rate, providing a margin of safety. Takes a couple minutes to do so.
I find it to be the right approach for us. I don’t specifically track our income and expenses. I do know that our income is about the same as when we worked and if many that say expenses go down in retirement are correct so we do have excess every month that we continue to invest and save
Do you really see tracking past spending as necessary as opposed to looking at net income, subtracting what you know for sure you will not spend after retiring and there is your future spending needs?
Not at all. What you’re now describing is a very big picture version of bottom up planning. What you originally suggested was taking a stab in the dark and estimate a percentage replacement based on rule of thumb…are you really sure your not practicing to become a politician? 😉
The idea of an income replacement percentage is common here. People ask “how much will I need.” And the answer comes back 70-80%. Apparently after running the numbers that is the common result so why not take the easy road😁
My method is even easier.
I don’t disagree with you—your suggested method is a large-scale, bottom-up approach, and it’s definitely better than just asking an AI about average percentage replacement rates.
Think about it this way: the mean elevation in the US is 2,500 feet. But that information is completely useless if I’m planning to climb Mount McKinley, the highest point in the country. Averages can be meaningless when what you actually need is specific, relevant data.