2. Tax deferral: the tax you do not pay today

Costs 5: Tax as a cost

This is the central module of the level, and its idea fits in a single sentence: a deferred tax is not a smaller tax, but it produces the same effect as if it were, because for every year it takes to be paid that money stays invested in your name and goes on earning a return. The practical consequence, measured in euros over thirty years, is far larger than almost anybody's intuition suggests.

As in the rest of the level, there is no tax advice here and no recommendation about how to trade: there is a calculation. The specific percentages used to illustrate it are Spain's at the time of writing and may well have changed by the time you read this; what does not change is the arithmetic.

Two identical portfolios and two calendars

Picture two investors who start with 100,000 euros and get exactly the same 7% gross a year for thirty years. Same market, same selection, same luck. They differ in one thing only:

• Portfolio A. Every December 31 it sells the whole position, settles the tax on the year's gain and the following day buys the same thing back with what is left.
• Portfolio B. It touches nothing for thirty years and pays once, at the end, on the accumulated gain of the whole period.

To isolate the effect, we apply the same 19% rate to both. There are no commissions, no spreads and none of the costs from the earlier levels: only the calendar.

What happens to the portfolio that turns over

Portfolio A does not compound at 7%. Every year it hands over 19% of its gain, so what actually stays inside and goes back to work is 81% of that 7%, which is to say 5.67% a year. That is not a one-year penalty: it is the rate at which it grows for three decades.

At that rate, 100,000 euros turn into some 523,060 euros after thirty years. And that money is already clean: it owes nothing to anybody, because it settled up every tax year.

What happens to the portfolio that holds

Portfolio B compounds at the full 7%, because nothing leaves the vehicle. After thirty years it is worth some 761,226 euros. Now the bill arrives, and it is not small: the accumulated gain is 661,226 euros and the tax comes to some 125,633 euros, a figure portfolio A was never presented with in one lump.

Even so, once it is paid there are still some 635,593 euros left.

The difference between the two is around 112,533 euros, more than the initial capital all over again, and there has not been a single differing investment decision between them. The same sum over twenty years gives 301,324 euros for portfolio A against 332,444 net for B: a gap of some 31,120 euros. Look at the pattern, because that is the important part: going from twenty years to thirty did not multiply the distance by one and a half, it multiplied it by more than three. The gap between two series compounding at different rates widens exponentially, not linearly.

The interest-free loan

There is a more honest way of describing what is going on. In the first year, portfolio B generated 7,000 euros of gain and did not hand over the 1,330 euros of tax that its twin did pay. Those 1,330 euros are not a gift or a rebate: they are a debt. They will be paid in full, at the same percentage, the day the position is sold.

But in the meantime they are inside your portfolio, invested, earning for you. It is the state's money working in your name and charging you nothing at all for the privilege. Hence the classic description: an unrealized gain works in practice like a loan with no interest and no fixed repayment date. In the example that loan had grown to be worth 125,633 euros after thirty years, and everything it earned along the way stayed with the investor. It is the same intuition John C. Bogle used to repeat about low-turnover index funds, and the one Robert Arnott and Andrew Berkin developed when they measured how much return evaporates in portfolios that are forever replacing positions: the toll is not the payment, it is the compounding that payment takes down with it.

What this calculation does not say

It would be dishonest to leave it here without marking out the conditions under which the advantage narrows or reverses. The decision is yours; here is the map:

• If the rate that applies on final settlement turns out to be much higher than the one in the intervening years, the saving from compounding can be wholly or partly absorbed. A rising ladder of bands means that concentrating an enormous gain in a single tax year can land it on higher rungs.
• If the rules change between the moment you defer and the moment you pay, you are taking on a regulatory risk the portfolio that settles every year does not run.
• If the asset you hold turns out to be worse than the alternative, the tax advantage does not make up for a bad investment. Holding on to a position purely to avoid a settlement is letting the tail wag the dog.
• If you need the money sooner, the whole calculation shortens, and the advantage with it.

The worked example, in one line

One hundred thousand euros, 7% gross, thirty years, the same rate for both: 523,060 euros turning over every year against 635,593 euros holding on, already net of a final tax bill of 125,633 euros that looks brutal and nonetheless leaves the investor 112,533 euros ahead. When the next level talks about turnover, this is the figure worth keeping in your head.