Korea vs Spain — Household Debt / GDP

How Spain's 47% came about

Spanish household debt stands at 47% of GDP — below the OECD average of 60% and about half Korea's 90%. On that number alone, Spanish households look financially healthy. But the figure is the endpoint of a descent from roughly 85% before 2008.

That decline was not the result of households deliberately saving more and paying debt down. A property bubble burst, house prices fell sharply, loans that could not be serviced were worked out, and new lending was shut off for years. A large part of the statistical fall in debt was loss recognition and credit contraction, not voluntary repayment.

The 47% is therefore less an indicator of "healthy households" than a scar left by an adjustment. The same 47% means something entirely different depending on the path taken to reach it.

What that adjustment cost

Spain's deleveraging was neither short nor cheap. Unemployment at one point climbed into the mid-20s in percentage terms, and youth unemployment ran far higher. As falling house prices eroded net worth, households cut spending and domestic demand stagnated for years.

While the household debt ratio came down, the government debt ratio went sharply up. The fiscal accounts absorbed the cost of cleaning up bank losses and paying unemployment benefits. The proposition seen earlier — that debt does not disappear, it moves ledgers — is confirmed here too.

The implication is clear. If lowering the household debt ratio quickly becomes the goal in itself, total cost does not fall unless you also design where the cost of that process is shifted to.

The options in front of Korea

Placing Korea's 90% next to Spain's 47% makes "it has to come down" look like the natural conclusion. The question is at what speed and at whose cost. A sharp adjustment risks reproducing, in compressed form, the path Spain went through.

The comparatively gradual route lowers the ratio not by shrinking the debt stock but by letting income grow faster than debt. It works on the denominator rather than the numerator; it takes far longer, but the cost of adjustment is not concentrated on particular groups. This route only holds, however, if management of new lending runs alongside it so that debt does not keep climbing in the meantime.

Either way, the precondition is knowing the composition of the debt. Mortgages and merchant credit respond differently to interest-rate and employment shocks, so setting a target for the aggregate alone can leave the genuinely risky part untouched.

Three ways a ratio comes down

There are broadly three routes by which a household debt ratio falls. The first is repayment: households divert part of their income to paying down principal. It looks healthiest, but it also cuts consumption and weighs on growth.

The second is loss recognition. When loans that cannot be serviced are written off, the statistical balance falls — but the loss transfers to financial institutions and ultimately to the public accounts. Much of the path Spain travelled falls into this category. This is where you get stretches in which the indicator improves while the economy deteriorates.

The third is growth in the denominator. If nominal income and GDP grow faster than debt, the ratio falls even with the balance unchanged. Of the three, this carries the lowest social cost — but it is the hardest to achieve in a low-growth phase.

So when you read that "the ratio came down," the first question is which of the three routes did the work. The same directional move in an indicator can point to entirely different economic states. Even if Spain's 47% and some future Korean figure were identical, a different path there should mean a different assessment.

The order to read this table in

First, separate level from path. If you do not know where the 47% came down from, that number cannot serve as a target. Second, when reading a change in the household debt ratio, check the change in the government debt ratio alongside it.

Third, check the composition of the debt (mortgage versus merchant credit) and the rate type (fixed versus floating). Fourth, read on the premise that Spain's figure is provisional and that the OECD average is the central value of a widely dispersed distribution.

Fifth, before adopting a ratio as a target, state which of the three routes down you intend. Repayment, loss recognition and denominator growth require different policy instruments and place the burden on different parties, so a target set without choosing a route tends to drift, at the implementation stage, toward the costliest one.