What I-S-E-E means for my money

2026-09-03 · by Ash1,250 words

The conventional advice puts earning first. Raise your income, then save some of it, then put the savings to work. The logic is clean and I do not think it is wrong.

I run my own money around four priorities.

Investing has first claim on a dollar. Saving takes what is left after investing. Earning is the engine that produces the dollar. And Experience says what the money is for.

I-S-E-E. Read it aloud and it says I see, which is the point. Most writing about money obscures. If a post here does not leave you saying "oh, I see", I have not finished it.

This is the argument for each part, and I have tried to be honest about which ones the data supports and which one reflects taste.

I, for Invest

Two arguments, and they sit at opposite ends of a life. Neither is "investing is the biggest lever", because during accumulation it is not.

Early, because time is the only input you cannot buy back. You can raise your income at forty-five. You can cut your spending at fifty. You cannot go back and start compounding at twenty-five. Every other lever stays available; this one closes behind you.

Late, because the arithmetic takes over. Past a certain portfolio size, annual returns exceed anything earning and saving can add. I checked this against hundreds of millionaire profiles. Among those stating net worth, income and savings rate, 86 percent have expected returns larger than what they add each year, assuming 7 percent real. At a conservative 5 percent it is still 77 percent.

Solve for the crossover and it lands near the first million. At the medians in that group, returns overtake contributions at a portfolio of roughly 971,000.

The median age at the first million in the same data is forty. That makes forty a rough marker for when returns begin to exceed contributions.

Investing gets first claim because the early opportunity expires and the late arithmetic eventually takes over.

S, for Save

This is the claim I expected the data to reject.

Three quarters of the profiles credit earning as their greatest strength. Ten percent credit investing. If you asked the people in this data, they would call earning the stronger lever.

So I tested it. Split them at the median income of about 207,000 and the median savings rate of 33 percent, and look at when the first million arrived.

Hold income and move from below-median saving to above-median: the date comes in six years. Hold saving and move from below-median income to above-median: five years.

The savings split is one year larger. Narrowly.

The pair I keep returning to is the middle two. A modest earner who saves hard arrives at 38.5. A good earner who does not arrives at 39.5. Those are the same answer.

Six against five is not a landslide and I would not bet a plan on the gap. The practical difference between the two levers is too small for a strong claim. Saving still gets its own place because it is the one I can control on a Tuesday.

I should say the obvious thing: finding a result that fits my framework is the result I would question least hard. Read it with that in mind.

The part that undercuts both claims

Neither lever does much until it is extreme.

Both are mostly flat, then fall in the top bands.

Doubling an income from 100k to 200k moves the date by nothing. Taking a savings rate from 20 percent to 40 percent moves it half a year. Everything that happens, happens at the extreme.

Formally, the correlation with age at the first million is -0.25 for savings rate and -0.21 for the log of income. Both negative. Both weak. Run them together and they explain 9 percent of the variation in the age people arrived, which leaves 91 percent to something else.

And the spread swallows the medians. Sorted by savings rate, the middle half of each band covers about ten years, and a quarter of the lowest savers beat the median of the highest ones. Whatever a savings rate does, it is nowhere near deterministic.

So the difference between saving and earning rests on a narrow margin, inside two variables that between them explain little. I would rather say that plainly than pretend the framework rests on something firmer.

E, for Earn

Earning is the engine and the data agrees: the highest income band arrives five years before the median.

Its place in the acronym does not make it less important. The letters describe my priorities when a dollar shows up, not the causal sequence that produced it.

Which brings up the objection anyone sensible raises first.

You cannot invest before you earn. As a sequence the order is impossible. There is nothing to invest until something has been earned.

So I-S-E-E is a priority order, not a chronology. When a dollar arrives, investing has first claim on it, saving takes what remains, and earning is the activity that produced the dollar in the first place. The useful distinction is what each part does, with Experience made explicit.

E, for Experience

Experience is the part with no data behind it at all.

The usual three pillars are about accumulation. They have nothing to say about what the money is for, which means a plan built only from them optimises toward the largest possible unspent pile. Bill Perkins made this case better than I will in Die With Zero: the goal is not maximum terminal wealth, and money not converted into life is not a win.

I called it Experience rather than Expenses on purpose. Same line in a budget, opposite instruction. Expenses are something you minimise. Experiences are something you buy deliberately.

I have no evidence for this part. It is a value, not a finding, and I am not going to dress it up as one. It is in the framework because a plan without it produces an outcome I do not want.

What is untested

The comparison between saving and earning rests on a one-year gap in a sample where income is reported today and the milestone happened years ago. Someone who arrived at thirty-two has had a decade longer to grow their income than someone who arrived at forty-eight, so the arrow could point either way and this data cannot tell me which.

The investing argument is on firmer ground, because the crossover is arithmetic rather than correlation.

The Experience part is untested by construction.

Everyone in this data already succeeded. There are no failures in it, which means I can see what worked and never what did not.

Next I want the income series rather than the snapshot. The profiles carry thousands of individual income figures, and about a thousand come with an age attached, which should be enough to ask whether the early arrivers were earning more at the time or keeping more of it. That would compare saving and earning more directly.

If it changes the conclusion, I will change the framework and say so.