Some Assembly Required

Some of you are still bothered.

Dudley understands.

You read the chapters about spending, Monte Carlo simulations and Betty’s one-pager and thought:

Very nice. But what exactly did you put in the software?

This chapter is for you.

Everyone else may leave. Go outside. Talk to your spouse. Plant something.

What follows contains account balances, inflation assumptions, withdrawal rules and enough financial modeling detail to ruin an otherwise pleasant afternoon.

Dudley considers this recreational.

One boring but important note: these are Dudley and Betty’s planning assumptions. They are not recommendations. Your money, taxes, health, family and tolerance for watching imaginary versions of yourself go broke will be different.

The entire system exists to answer three questions:

What do we have?

What does the life we want cost?

How long will it last?

Everything else is implementation detail.

Dudley loves implementation detail.

This is part of the problem.

The System

Dudley eventually settled on two tools.

Lunch Money tells him what happened. ProjectionLab tells him what might happen.

Dudley created the same major expense categories in both systems. Lunch Money compares each budget with what they actually spend. ProjectionLab carries those expenses into the future and tests whether their money can support them.

Once a year, Dudley reduces the whole operation to one page for Betty.

She does not need to become a ProjectionLab administrator.

Neither does anyone else.

What Do We Have?

Dudley and Betty begin 2027 with approximately $1.2 million in cash and investment accounts.

That includes $30,000 in cash, $50,000 in taxable investments, $230,000 in a 401(k), $830,000 in traditional IRAs and $60,000 in a Roth IRA.

Dudley obtains those balances from their brokerage accounts. He does not separately model the few thousand dollars moving through ordinary checking and savings. That money arrives, pays bills and leaves again.

Giving it a retirement strategy would make Dudley happy and accomplish almost nothing.

The $30,000 cash balance serves as their immediate spending reserve. The rest remains invested.

They also own a house worth approximately $400,000 and automobiles worth about $60,000. Those values come from rough Internet searches, which means they are precise enough to be displayed online and not precise enough to be useful anywhere else.

Their mortgage balance is $230,000. Their auto loans total $40,000.

That puts their net worth at approximately $1.39 million.

Dudley rounds it to $1.4 million because even he has limits.

Betty cares more about the $1.2 million that can support their retirement. They are not planning to sell the living room to buy groceries.

What Does Our Life Cost?

Their current retirement budget is $10,000 per month, or $120,000 per year.

That includes a $1,200 mortgage payment, $800 in auto loans, $1,200 for private health insurance, $2,000 in essentials, $2,400 in discretionary spending, $1,400 for travel and $1,000 for home improvements.

The total is exactly $10,000.

Their actual spending is not.

One month contains travel. Another contains a house repair. Some months include dinner with friends. Others include a Costco purchase that neither Dudley nor Betty can adequately explain.

Lunch Money records what happened. Dudley reviews the transactions monthly, corrects anything categorized strangely and watches the year-to-date trend.

He does not care whether every month equals $10,000.

He cares whether their life continues to cost roughly what they planned.

The same categories appear in ProjectionLab. There, Dudley identifies which portions are fixed and which are adjustable.

The mortgage, auto loans and health insurance are completely fixed. Essentials are 90 percent fixed. Discretionary spending is 60 percent fixed. Travel and home improvements are each 50 percent fixed.

That produces $7,640 per month in fixed spending and $2,360 that can be adjusted.

Calling half of travel fixed may disturb certain financial planners. Dudley is comfortable with that.

The purpose of retirement was not to survive as efficiently as possible inside the house.

The auto loans end in 2031. Medicare begins in 2034 and is modeled at $840 per month in today’s dollars for premiums. The mortgage ends in 2051.

Healthcare grows at 4.5 percent annually. Most other continuing expenses grow at 2.5 percent.

Discretionary spending, travel and home improvements decline by 10 percent every five years beginning in 2031. The reduction applies to the entire category and repeats every five years.

This does not mean Betty has agreed to reduce travel by exactly 10 percent in 2036.

It means the software required a number.

Betty remains independent.

How Long Does It Need to Last?

The projection begins January 1, 2027. Dudley and Betty are both 57 and model the plan through age 89.

Their portfolio begins at approximately 70 percent stocks and 30 percent bonds and cash. Dudley assumes stocks return 7 percent and bonds return 3.5 percent, producing a weighted expected return of approximately 5.95 percent.

He does not separately subtract investment-management fees. They use no-load mutual funds, and the return assumptions are treated as net of the fund expenses already embedded in performance.

The model does not assume automatic rebalancing.

In real life, Dudley will rebalance.

Of course he will.

The software is being asked to model retirement, not Dudley’s personality.

Consulting income can range from zero to $50,000 per year before taxes. The $50,000 scenario increases with inflation and continues through the bridge to Social Security.

Dudley has not decided whether he will consult, write or become unusually committed to yard maintenance.

At age 62, consulting ends and combined Social Security begins at approximately $52,800 per year in today’s dollars, increasing with a 2.5 percent cost-of-living assumption.

The $120,000 lifestyle budget does not include income taxes. Consulting income and traditional retirement distributions create tax obligations beyond that amount.

Their tax accountant helps estimate withholding and quarterly payments. Dudley enters the resulting taxes separately rather than pretending that gross income and spendable income are the same thing.

Dudley’s 401(k) qualifies for the age-55 separation-from-service exception. He can use it before age 59½ without the additional 10 percent early-withdrawal penalty, although the distributions remain taxable.

The traditional IRAs are different. Dudley avoids taking money from them before age 59½ because those withdrawals would generally incur the additional penalty.

The basic order is consulting income, operational cash, taxable investments and the qualifying 401(k).

After age 59½, the traditional IRAs become available. The Roth IRA comes last.

The actual withdrawal and tax decisions receive adult supervision from the accountant.

This is one of several areas where Dudley has reluctantly accepted that reading things on the Internet is not a professional credential.

Don’t Be Stupid

The plan assumes Dudley and Betty behave like normal people.

This is difficult to represent in financial software.

If the Dow and S&P 500 fall approximately 10 percent from a recent high, Dudley will notice. The news will become louder. There will be coffee.

Dudley and Betty will talk.

If the decline looks serious, they can reduce spending from $10,000 a month toward the fixed level of $7,640. Travel can wait. Some purchases can wait. The greenhouse can wait.

A broken furnace will not wait.

They resume normal spending when the markets show signs of life. There is no precise mathematical trigger because actual people do not spend their lives standing beside a financial alarm waiting for it to turn green.

They pay attention.

They adjust.

They avoid selling large amounts of stock while prices are down. If they have available cash, Dudley may even buy more.

His technical name for this strategy is don’t be stupid.

It lacks the elegance of a formal withdrawal policy.

It has worked surprisingly well.

Monte Carlo Ruins the Afternoon

ProjectionLab can test more than 400 possible market paths. Dudley also asked ChatGPT to run 1,000.

Some futures begin with strong markets. Some begin with terrible markets. In some, inflation behaves itself. In others, it behaves like inflation.

The test included the scheduled spending reductions, no automatic rebalancing and temporary spending adjustments during major market declines.

With no consulting income, the simplified model succeeded in roughly the mid-fifties before separately estimating taxes on portfolio withdrawals.

At $25,000 of consulting income, it moved into the mid-sixties.

At $50,000, it reached approximately the mid-to-high seventies.

Taxes reduced those results. More aggressive spending adjustments improved them. Changing the definition of “the market shows signs of life” moved them again.

Dudley could make the number better.

He could continue working longer. He could earn more money. He could reduce travel, postpone projects, spend less on dinners and assume Betty would eventually lose interest in flowers.

He could move the result toward 90 percent.

With enough creativity, he could probably reach 100 percent.

The safest retirement plan, Dudley discovered, was to continue working, stop spending and eventually die with an excellent spreadsheet.

This seemed to miss the point.

The software had answered the question Dudley asked. Could they guarantee that every possible future would work without changing anything?

No.

Nobody could.

But that was not how they planned to live. They would watch what they spent. They would adjust during bad markets. Dudley might consult. He might write. Social Security would arrive. The loans would end. They would review the plan every year and replace old assumptions with reality.

They did not need one perfect plan.

They needed a plan they could operate.

Betty did not need 100 percent certainty. She needed to understand whether they had enough, what could go wrong and what they would do about it.

Dudley looked at the simulations again.

Some versions of them ended with very little. Others ended with millions. Most landed somewhere between those outcomes, having traveled, helped their family, fixed the house and purchased an amount of flowers the model had failed to anticipate.

Could Dudley keep working and improve the odds?

Yes.

Was another percentage point worth another year?

Maybe.

Would he ever reach a number that made the future certain?

No.

Dudley closed the laptop.

They had enough money to begin, enough flexibility to adjust and enough sense to pay attention.

Enough was not a guarantee.

Enough was plenty.

Screw it.

Close enough.