People aren’t rich because they don’t do the math.
That’s been the lesson, again and again. Not bad luck. Not bad timing. Not bad genes. Just numbers, left unfinished.
You’re taught it backwards from childhood. Saving is virtue. Debt is vice. The bank exists to hold your money, not to lend you theirs. So you spend your life making financial decisions on instinct, on fear, on habits inherited from people whose balance sheets were never large enough for those habits to matter.
Then one day you sit down with a blank sheet and you run the numbers.
And you find out the difference between making it and not making it doesn’t live in your salary, your inheritance, or your timing. It lives in whether you can read a rate table.
Concrete example. The one sitting on my desk this week.
You hold €100k in pledgeable assets. The bank offers you two scenarios.
Scenario A: borrow €50k at 3.5%.
Scenario B: borrow €100k at 4.5%.
The maximum drawdown the bank requires is the same in both cases. If the collateral falls 35%, they liquidate. The trigger sits at the same line. The only difference is how much you’ve taken.
Which one do you pick?
Intuition says A. Lower rate. Less risk. More conservative.
If you stop running numbers there, you’ve already lost.
Assume an 8% gross return on invested capital, roughly the long-term S&P 500. Not aggressive. Realistic. And keep €30k of the borrowed line in cash as a defensive buffer, ready to absorb a margin call without selling.
Scenario A: equity 100k, debt 50k, 30k in buffer, 120k working in markets. 8% on 120k = €9,600. Interest: €1,750. Net: €7,850. ROE: 7.85%.
Yes, I know what you’re going to tell me: if the return is lower than just buy & hold, why should I do it? It’s for illustrative purposes. Obviously, we could have the 30k at 3% in some fund, but it’s just to make the difference a bit more pronounced.
Scenario B: equity 100k, debt 100k, 30k in buffer, 170k working. 8% on 170k = €13,600. Interest: €4,500. Net: €9,100. ROE: 9.10%.
No leverage, no buffer: €8,000 net. ROE: 8.00%.
Look at that sequence. A is worse than not leveraging at all. The drag of carrying 30k at 0% eats the entire spread of the cheap tranche. You paid for a seatbelt without buying the engine.
Now extend it. Each year you re-borrow to maintain the same leverage ratio. Equity grows, debt scales with it, the buffer scales with it. The ROE stays constant. Time compounds.
Year 10: baseline 216k, A 213k, B 239k.
Year 20: baseline 466k, A 453k, B 571k.
Year 30: baseline 1,006k, A 965k, B 1,364k.
B beats the baseline by €358k. A loses €41k to it. The gap between A and B at year 30 is almost four times your starting equity. From the same collateral. Same trigger. Same buffer. Different math.
Here is the structural insight most people miss. The efficiency of leverage isn’t measured in the interest rate. It’s measured in the spread between the marginal cost of borrowed capital and the marginal return on deployed capital. While that spread stays positive, borrowing more is structurally superior to borrowing less.
The nominal rate is noise. The spread is signal.
But the spread alone isn’t enough. It has to be wide enough to absorb the cost of the defense, the buffer, the reserve, the optionality you keep against a margin call. Below a certain leverage, the defense is more expensive than the spread it protects. You end up paying insurance on a position too small to justify it.
This is why partial leverage is the most common mistake of the half-sophisticated. People want the upside of borrowing without committing to the structure that makes borrowing work. They take the cheap tranche, hold a prudent buffer, feel responsible. And they finish the decade behind the unleveraged investor who did nothing.
That is the trap of scenario A. Not that it’s risky. That it’s expensive in a way the rate table doesn’t show.
What does someone who actually ran the numbers do?
They take B. They keep the buffer. They don’t spend the loan on lifestyle. They deploy it into the same long-duration positions they already own, the ones that justified pledging the collateral in the first place. They accept that the worse tail outcome of B exists, and they carry a seatbelt that B is wide enough to afford.
And if they can’t carry that buffer at scale, they don’t take A. They take nothing. Unleveraged at 8% beats leveraged-at-half with a buffer that suffocates the spread.
The bank counts on you not finishing the calculation. That’s why both lines are on the table.
Finish it.