Asian Debt Service Breakdowns
Here is some pretty bad, back-of-the-envelope calculations and estimations I did for the economies in Asia based on yesterday’s information from the Nikkei Asian Review. I don’t know if it gives a good or bad picture of the Asian financing market, but I will use it here as if it does at least provide some accuracy, with full disclosure of my assumptions and why I made them.
First, a few reflections strictly on yesterday’s data:
- Korea has a relatively advanced corporate bond market that could warrant further study for adoption in other East Asian countries, under assumption that corporate bonds can be shifted from bank loans or bonds.
- East Asian countries ex Japan are not too keen on government debt. South East Asian countries are more likely to get government debt in their debt mix.
- Countries with relatively strong bouts of growth behind them have higher consumer debt, generally going above 60% and sometimes approaching 100% of GDP.
Now, for the meat of the post:
What does all this debt cost?
That is obviously a rather tricky question. Let’s start with the nominal financing cash flow to keep from going into default and then build from there.
Nominal Debt Servicing Costs
For the first chart I back-of-the-enveloped, I assumed consumer debt was expensed at the prime rate, same with corporate debt. Government debt is expensed at the government 10-year bond yield, and corporate debt is expensed at an average of the prime rate and the government 10-year rate (meaning I assume that corporations wouldn’t go to the bond market if it is more expensive than the prime rate, and that they are considerably less credit-worthy than their governments). In Japan where inflation is higher than this number, I have added inflation and the government bond rate to arrive at approximate numbers for corporate bonds, but beware that this is forward-looking for Japan in a way that isn’t reflected in any other country due to a recent inflation spike and politics. It might not either be particularly realistic, but given how small corporate bonds are as a percentage of aggregate financing i don’t really mind the induced error which would be one among many.
For computing the nominal cash outflows, I have also modeled the consumer debt mortgaging very simply and roughly. I simply make an assessment of how much of the country’s debt is left on a fixed-amortization schedule on average (where I assume that virtually all of this debt is for housing on a 30-year mortgage, given car ownership rates and costs in Asia and avoidance of credit cards, but this would imply a low-ball estimate of amortization costs). Japan is estimated to average 8 years left to maturity, Korea 17, Malaysia 15, China 25, Taiwan 10, Thailand 12, India 23, and the Philippines 20. It is simply an extrapolation of roughly when the growth spurts have happened in the countries and when credit has been extended to consumers, plus minor adjustments for other debts (such as cars, other changes of the economy on a relatively quantitative basis, and a lot of gut feeling). 30 is then divided by this number, and the result is multiplied by 3.3% to evaluate the 30-year equivalent mortgage equivalent payment rate. Multiplying this rate with the consumer debt fraction of GDP gets me the GDP mortgage cost for each country. This model underestimates the payments in Japan and Korea I think (assuming they have fixed-payment plans), and probably to an extent over-estimates it in China, India and the Philippines under the same (slightly less valid) assumption.
Note also that this model pretty much ignores bond- or debt refinancing. Debt is simply assumed to be rolled in need, as getting a good grasp of the individual government payment horizons, changes in bond prices since these bonds were issued and how these bonds will be refinanced when maturing is so uncertain and speculative that leaving it ignored feels like less of an error than to treat it with a quantitative laziness bordering sinful.
Here’s the visualization of the nominal data:
- Except for China, India and the Philippines, most countries end up with very significant mortgaging outflows.
- Corporate debt is only really significant in Korea, and to an extent in Malaysia. Real costs of corporate debt in Japan are most likely much, much lower.
- China is a pretty standout case in corporate loan interest!
- Japan really doesn’t have too much of any costs of debt anywhere, except for mortgage repayments.
- Indian government debt looks pretty crippling…
Now, how does this change when inflation is (very simplistically) taken into account?
Inflation-adjusted Cost of Debt Capital (CoDC)
This ignores all mortgages as that is more or less a conversion of debt into equity, and that proportion of the debt is already counted as generating value towards the interest rates. The yield rate of whatever debt was simply subtracted by the inflation rate, and the result multiplied to that debt’s share of the domestic GDP (thus over-stating the effect of inflation slightly across the board). I can’t really be bothered running proper bond valuations on this anyway, and I don’t think I (or most of you for that matter) will be involved in buying these bonds anytime in the near future either.
The fundamental problem of this model lies even deeper: the model implicitly assumes that income-generating assets are bought with debt that can just barely be serviced, and then assumes that the value generated or held by those assets follows inflation 1-to-1. if equity is used and a profit is held, then this isn’t true, if wages don’t follow inflation then it isn’t true, if asset prices change at a different pace then inflation then this isn’t true, and if a government runs a deficit then this isn’t true. There are more model problems if you want to look for them. It’s much more of an idealized potential long-term picture, but thankfully a lot of these issues kinda neutralize each other at least.
- The interest expense for the Philippines, India and Malaysia all but disappear!
- Japan goes deep in the “negative”! This model would imply a lot of gains for different people across the economy with these inflation figures, even though wage inflation in the country is lagging. This can probably be explained by asset price increases given the large Japanese savings rates and house ownership.
- China corporate debt is a pretty big cost staple, same with Thai and Taiwanese consumer debts. Korea has corporate loans going roughly toe-to-toe with household debt, and relatively perky corporate bond costs as the only economy under the magnifying glass.
- Government debt costs at the current inflation levels doesn’t look particularly debilitating for any Asian economy. Sure, there’s a pretty big risk later on, but currently the governments are essentially borrowing at below 1% real rates across Asia, and there should be rather little to worry about from a government finance perspective unless growth dies off and/or there are waves upon waves of defaults.
Finally, let’s summarize all this data from the above chart, as well as using a non-inflation-adjusted aggregate CoDC column for good measure:
- Thailand, China and Korea look to bear the largest inflation-adjusted total burdens.
- Malaysia is number 2 when mortgages are considered as they are in this model.
- Japanese mortgages are a majority of the costs in either format, and the current spike in inflation should allow either business activity or loan devaluation to reduce net debt levels in the country by at least [10% + Growth Rate] of GDP on an annualized basis.
- Japan’s large negative inflation-adjusted aggregate CoDC numbers are obviously due to the massive amount of debt in the economy, and shouldn’t really inspire anyone to follow Japanese policy. It does however shed some light on Kuroda’s need to stoke sustainable inflation, preferably of the good kind driven by wages. Still, corporate profits, consumer assets in terms of housing and international financial assets, and government solvency are all getting a pretty big boost here and it will be interesting to follow how this develops in the future.
All data were either estimated to nearest 5% on the Nikkei Asian Review chart, or checked through Trading Economics.
I will try to get some more analysis of aggregate total financing at a later date, but it requires a little bit more intense data analysis and hunting to get the equity market valuations, market overall cost of equity and dividend expenses. No promises.