Industrial Production Follows Suit
As Hungary is now an export dependent economy, it is not surprising to see that Hungary's industrial production also plunged on the drop in export demand. Output was down 28.9% year on year in February, according to unadjusted figures, while on a working-day-adjusted basis there was a 25.4% decline, sharper even than the 21% drop seen in January. The statistics office has changed its methodology since the early 1990s so a direct comparison with earlier figures is not really possible, but the office said the February drop is probably the largest monthly fall since the collapse of Communism at the end of the 1980s.
Loan Defaults The Main Threat Now As The Forint Weakens
According to the latest financial stability report regularly published by the Hungarian central bank on condition the economy performs in line with mainstream expectations and contracts by 3.5 per cent this year, with the exchange rate stable at aound 290 forints to the euro, banks would be able keep their capital ratios above 10 per cent of total assets. This is, in theory, comfortably higher than the international regulators’ 8 per cent minimum. But under a “stress scenario”, whereby gross domestic product falls 10.5 per cent and the forint slides by 15 per cent, capital adequacy would reach the 8 per cent level, and banks accounting for nearly half the total assets would fall below the minimum.
The latest IP and export data confirm that the Hungarian economy is indeed deteriorating faster than expected. The forint is currently trading about 6 per cent below the central bank’s baseline forecast, while GDP forecasts of minus 5 per cent or worse are now commonplace. And even this is not the complete story, since as the central bank’s report says: “The calculations described above are characterised by considerable uncertainty . . . [making] credit risk forecasts very uncertain.”
Among the most evident points of uncertainty is the rate of household and corporate loan default. HSBC Holdings Plc, Europe’s biggest bank, expects loan delinquencies to reach 23 percent in Russia before the crisis is over, making the level Europe’s highest rate after, guess who, Hungary where they anticipate bad loans will reach 25 percent of the total.
4 comments:
I would be really interested in exactly what calculations HSBC carried out to get to their 23% bad loan ratio for Russia and 25% for Hungary. Estimates to establish a link between macroeconomic scenarios and loan quality are typically based on reasonably long historic time series. I doubt these in fact exist in any of the cases in question - indeed, I know for a fact that they are not there. For this reason, any NPL forecasts need to be made very carefully if one does not want them to prove simple guessworks. I do not about Russia, but HSBC is certainly not present as a bank in Hungary.
Hello,
Well the Russian case is reasonably straight forward. According to my back of the envelope calculations Russian NPLs will hit 40% next May or June (2010)if the price of oil doesn't go up substantially before, which given the low level of global demand and the high level of indebtedness is hard to see.
The basis for this estimate is a statement by OAO Sberbank Chief Executive Officer German Gref, who reported that NPLs were growing at a rate of 20% per month.
This intrepretation is also back by the mean result of a Bloomberg survey of 17 bank analysts this week gave a mean estimate that Russian bank bad loans will quadruple to $70 billion this year, with non-performing loans rising to 12.8 percent.
In this context the HSBC estimate would see to be cautius, prudent, and rather on the downside.
"Estimates to establish a link between macroeconomic scenarios and loan quality are typically based on reasonably long historic time series."
No. You can't do it like this this time. We haven't got anything to compare this with, arguably, in Europe at least, even the 1930s wasn't as strong as this. So here you need to exercise your own judgement. You need to know the rate of contraction of the economy, the rate of increase in unemployment, the rate of wage deflation, and the rate of commercial bankruptcies.
I don't have data on the last point (if you can point me towards data I would be grateful), but looking at the other three, and the path of the forint (and the role of CHF loans) I don't find the HSBC number exaggeated at all.
Remember, we aren't talking about structured financial products here, but conventional loans in the mortgage book. All this takes much more time to surface, maybe two or three years, but it will eventually surface. I'm sorry, it's just like this.
So it is effectively a product of expert judgement, based on the judgement of other experts and on thin anecdotal evidence in the Russian case, and perhaps on even thinner evidence in the Hungarian case. Well, this is hardly better, I am afraid, than a weather report saying that in winter, it is likely to be cold or even colder. Honestly, I have simpathy for your claim that the current extreme conditions do not allow to use methods that are standardly described as the best industrial practice, but then we should warn readers that the whole frenzy about stress testing, which is tremendously popular in these days, may be pointless, and that the results we quote are based on something substantially more simple and approximate. On Hungary, by the way, the only factual thing we know is that the NPL ratio was 2.8% at end-2008, but of course this does not tell a lot about the future. The anecdotal evidence is that the bad loan stock is currently growing quite significantly, but the earliest time to know more details will be in about a month time, when the Q1 data is out. Losing eventually all extra debt caused by the HUF weakening seen so far would mean losing 10% of the total loan book of local banks, but surely that would happen over several years, given the existing repayment schedule.
Hello again,
Thank you for your comments, although I resist the idea that all of this is like trying to forecast the weather in six months time.
There are some constants. Hungary, for example has a declining and ageing population. This is a given, and we can assume the position won't change much over the next couple of years.
But then this has quite considerable implications for what we can expect to see from domestic Hungarian GDP growth, even if most bank analysts don't see the connection.
On GDP, I think it is reasonably clear that the economy will shrink by 5% plus this year, and at least 3% next. More, I would agree, we cannot see, since it depends on things which happen between now and the end of 2010. For example, what will be the forint euro, what will be euro-dollar, what will be the level of the Hungarian CPI, what will be the Hungarian unemployment rate, what we be the degree of recovery in German exports?????
Without this kind of data, I would agree, most "predictions" are more or less useless.
So I would accept your point that putting a number on the level of Hungarian bad loans may well be a fools game at this point, and I certainly wouldn't defend the HSBC number, since I have no idea what methodology they have been using.
I also agree with your scepticism about "stress testing" since at least here in Spain, all the results I have seen severely underestimate the extent of the problem, largely becuase they refuse to contemplate the size of the macro correction which is taking place. So really even the "worst case scenarios" employed could be considered somewhat benign.
But where I would dig my heels in would be over the central issue of defaults in this crisis. I still think most people have been mislead into complacency by the "super nova" characteristics of the US melt down, and have not stopped to think that equivalent deflation pressures over a period of two, three or more years can have the same consequneces. It is just that it takes longer to see.
Spain's contraction started really in the summer of 2007 (ie nearly 2 years ago), and the level of bad loans is still below 5%, but just give it two more years of the same (or worse) and we will get there.
Of course, I have no idea how high the level of NPLs will go in Spain, or what the critical level that will blow out the banking system is going to be. As you would point out, no one does. It is an empirical question, and we will know it when we see it.
Basically, I see the Hungarian economy (like the Spanish one) as caught in a self perpetuating downward spiral at the moment, with the contraction putting pressure on the budget deficit, and the budget deficit putting pressure on the forint, and the forint putting pressure on the NBH to keep interest rates high.
This is the classic 1930s situation, I think.
Basically, the economy contracts, the budget deficit grows, the government cuts spending and the economy contracts more, meanwhile the rating agencies cut, and the NBH can't.
So you have a down the bath plughole process, where there is absolutely no stimulus (either monetary or fiscal) available, and there is no homeostatic endpoint till:
A) Either via forint weakening, or wage and price deflation
B) plus a rseurgence in German exports
you can get enough export GROWTH to stop the tailspin.
We are at least two or three years away from this point at present, and so in the interim the economy has become a huge machine for producing NPLs (of whose size incidentally we have no real idea, since banks in all countries and in all crises have a well known tendency to hide NPLs for as long as possible to protect their share price and capitalisation, this is why the deterioration when it comes is normally very rapid).
So, as I say, the main output of the Hungarian economy at this point is the NPL, and this is not (as assumed by the IMF and the EU Commission) simply a balance of payments crisis. And the sooner people with the authority to take decisions do so the better for the entire Hungarian people.
Basically, I have no idea what the finally level of loan default might be in Hungary, or when we would get there. Basically, I'd rather not wait to find out, and would prefer someone did something now so we don't get there.
Post a Comment