The funny world of Mr Katzarian

With the world focusing on the Greek debt tragedy, an article in the NY times about one Mr Kazarian, who claims that Greek debt in reality is only one tenth (!) the official size, caught my attention.

Mr Kazarian, founder of Hedge fund Japonica partners, is one of the largest private owners of Greek government debt. He has bought the bonds in 2012 and has made a (paper) killing so far, as they have appreciated in value. He says he is still holding on to his bonds believing they are still good value and – untypical for a creditor – argues for a debt write-off. According to him, the official debt figure (175 percent of GDP, ahem) is a meaningless number since it does not reflect “fair value” and is diverting attention from Greece’s true problems and “unnecessarily suffocating” the country.

My immediate reaction was that this is another version the “we owe it to ourselves” argument made by Keynesians to justify ignoring budget constraints. (For those of you who are unfamiliar with this argument: just remember that it is always helpful to be on the side of government if your goal in life are fat consultancy contracts or a tenured position).

Then I was prompted to ask Mr Kazarian how can the official debt be meaningless and “unnecessarily suffocating” at the same time? I mean, if something really doesn’t matter how can you be bothered with it? Isn’t that illogical?

But, Mr Kazarian is not another Nobel-prize winner without real world decision-making experience. Watching the video I learned he has a quite long and successful track record as a distressed debt investor. As a general rule, I tend to take arguments made by investors who are more experienced than me seriously even if they appear ludicrous. Therefore I decided to delve deeper into the matter.

I am not sure I have captured all of the arguments, but I think the following points summarize his view:

  1. Due to concessions made Greece’s debt burden is sustainable and lower than it appears
  2. Concessions (lower coupon, long maturities…) mean that the debt is worth less than par
  3. Writing of the debt to fair value leads to more realistic figures (“better accounting”)
  4. “Better accounting” leads to more confidence and “better decision making” by politicians

In the following paragraphs I will elaborate on each of these points.

Point 1: Due to concessions Greece’s debt burden is sustainable and lower than it appears

During the Greek debt restructuring 2012 part of the debt was written off, part of it was refinanced. This was a political decision/process, not a market based outcome. Due the fact that the write-off was not high enough (in order not to jeopardize banks who had foolishly lent the money encouraged by government-serving regulation) its debt debt-levels remained elevated and Greece remained shut-off from capital markets access. European tax-cows were forced to refinance the debt via various institutions (“Troika”) which are now the main Greek creditors.

However, at the time of the restructuring periphery debt was yielding upwards of five percent and at these levels Greece’s regular interest payments would have been overwhelming even after the write-off, with the result that Greece’s problems would have been in the news every second day. Now, NOBODY of the bureaucrats involved in the decision could need that. Therefore concessions were made: mostly below average coupon rates and maturity extensions. (The Austrian Finance minister at the time even said this would be a good deal for taxpayers. I don’t think she lied, I honestly think she was/is clueless – there is a lot of negative selection going on in bureaucracies).

So, yes, concessions were granted, but does this mean the debt level is lower than it appears?

Readers will remember the exchange I had with David Einhorn about the true size of Greece’s interest/GDP ratio. There was a material discrepancy between his bottom-up estimate and the official EU commission budget forecast numbers I used. I still have not resolved where difference comes from. Anyway, I have changed my mind and now think that interest/GDP does not matter much.

What matters much more is interest/tax revenue – and here Greece scores poorly. Below you find the tax revenue/GDP ratios for selected European countries.


As you can see, Greece’s tax revenue ratio is much lower than in other countries, even Cyprus’s ratio (39%) is significantly higher. And the difference is material.

Since I could not agree with David Einhorn on the correct interest expense/GDP ratio I have used another source. This post on FT-Alphaville features an IMF projection of Greece’s interest payments of about EUR 10 billion annually, i.e. about 4.75 percent of GDP. This is slightly lower than Portugal’s burden (5%) and close to Italy’s (4.7%) or Ireland’s (the latter figures taken from my post). If we forget for a moment that these countries have debt problems of their own, one could indeed conclude Greece is not extraordinary and wonder what the fuss is all about.

Adjusted for the lower tax base, however, the picture changes dramatically,


Now Greece looks considerably worse than the rest. For instance, it has to set aside double the tax revenue for interest payments than Spain, another crisis country. And these figures do not yet take into account the recent tax strike of Greece’s citizens.

Of course, you could say that Greece could raise taxes and the problem will go away. This is exactly what the creditors want and we hear lot’s about how Greece needs to improve tax collection these days. Economic ignorance wherever you look! As if increasing taxes has ever solved any problem or led to economic growth (The causality runs exactly the other way: higher wealth allows the political class to extract a larger share before people revolt).

It is absolutely certain that this will not work. The reason being, of course, that higher taxes will stifle what (legal) economic activity is left in Greece. Just think about it: Greece would have to increase average taxes by 30 percent in order to reach Cyprus’s or Portugal’s tax level – impossible!

To sum up: No, Greece’s debt load doesn’t look sustainable WITHOUT new concessions. Even if you ignore the stock and just look at the flows it remains too high.

Ad 2.) The concessions (lower coupon, long maturities…) mean that the debt is worth less than par

This is obviously correct. Without concessions made the bureaucrats could not even pretend that Greece is solvent.

Ad. 3) Writing of the debt to fair value leads to more realistic figures (“better accounting”)

Now, I am not exactly sure what he means by that.

It is certainly true that a write-off would bring much-needed transparency. For instance, tax-payers in the rest of Europe would finally be confronted with holes in their governments’ budget. So far most of the “educated” citizens of Europe have believed their governments’ assurances that all is well. The politicians can get away with this because government accounting doesn’t know about provisions for expected losses and they will be acknowledged only when finally written-off. This is indeed a huge flaw in public accounting and Mr Katzarian is absolutely right to criticise that.

Or take the ECB. A write-off would finally make it clear that it has engaged in government financing, although explicitly forbidden by EU contracts. This certainly would mean the end of the fiscal union debate. As someone who is tired to see government economists and “quality” newspapers argue for that nonsense, I would highly welcome such a development – even if I am aware that it would lead to a big economic and political crisis. (I confess that am a fan of the German proverb: “Besser ein Ende mit Schrecken, als Schrecken ohne Ende!”)

And this is exactly the reason why a write-off will never happen, as long as extend and pretend works. The accounting of government is opaque for a purpose. It has never been otherwise. If you do not believe me, read this book, it will cure you of any illusions you might have.

But, I feel that Mr Kazarian labours under another misconception as regards debt accounting. Warning: this is a bit technical!

He claims that EU accounting is arcane and does not reflect reality, i.e. the “fair value” of Greece’s debt, with the consequence that the parties involved (Troika, IMF, the Greek government) discuss about the wrong figures. Using proper accounting (fair values, not nominal values) the parties would realize that the Greek problem is much smaller. Happy end!

First possibility: Mr Katzarian confuses debtor vs. creditor accounting.

If you are a creditor of someone who is likely to default, the “fair value” of the loan will be below its par value. Now, a conservative management would provision against a likely default and write down the loan to fair value. They would still keep the nominal value on their books, however, for that’s what was lent in the past, but they would book a provision against this claim until the problem is resolved in bankruptcy. For the reasons mentioned before, Greece’s lenders (Troika) are not interested in conservative accounting practices. Heck, this is why they try to avoid technical bankruptcy and continue to pour money into the hole.

As a debtor, however, you are NOT allowed to write down the loan. And this is not arcane, but good and honest accounting. For, when it comes to bankruptcy the creditors’ nominal value represents the legal claim against the debtor’s assets. They form the basis for negotiations and the ultimate court decision.

Now, I would be shocked, if Mr Katzarian did not know all this. I mean that’s the bread and butter of distressed debt investing and he has a 20+ year track record in that field.

Which brings me to the second possibility: he thinks that due to the fact that a lot of the refinanced debt has a negligible coupon, it should be booked like a zero bond.

How does that work?

If you issue a zero bond at, say 50 percent of par, with a maturity of 30 years, it is not the par value that appears on your balance sheet. That would be insane, as it implies the creditors have a pledge on your assets twice as high as the amount they lent you. Instead, the liability rises in line with the interest accrued over time. The more time passes, the higher the creditor’s pledge becomes. It is this sum, i.e. the sum of price paid + interest accrued, which constitutes the claim in bankruptcy.

So is Mr Katzarian right, after all?

Well, I am afraid he is not. First, the refinancing by the Troika was not a zero coupon bond transaction – they really refinanced the whole nominal amount, not some fraction of it. The fact that the “fair value” at the time was significantly below the price paid doesn’t affect the accounting. The correct way would have been to provision part of the exposure immediately on the creditors’ books, but the nominal amount should stay there nevertheless. For comparison, in a real zero coupon transaction there is no need to provision immediately, since only a fraction of par is paid in.

But even if we pretended, it was a zero coupon bond transaction, not much would change for Greece WITHOUT further concessions. Yes, debt/GDP would be lower, but the interest expense would remain the same, with the difference that instead of a coupon payment, the (higher) interest accrual on the zero would need to be budgeted for. I see no way around it: Greece’s debt load is simply unsustainable without significant further concessions.

 Ad. 4) “Better accounting” leads to more confidence and “better decision making” by politicians

Politicians do not want to make good decisions. They want to stay in power, not maximize welfare, or what have you. Arcane accounting is a tool to achieve that (by misleading the public).

But let us pretend for a moment that “better accounting” is indeed achievable. How would it look like?

Clearly the debt would have to be written-off to sustainable levels, whatever that means. If we define the German level as sustainable (defined by interest/tax payments) one would have to cut Greece’s debt load by roughly 65 percent, ceteris paribus, and Greece would be AAA credit according to current rating agency logic.

But is this logic correct? If we talk about “true accounting” would not we also have to record the debts of off-balance sheet liabilities into account? For example, would not we have to account for demographic changes and associated welfare liabilities as well? After all, something similar happens in life insurance/pension accounting.

And Greece’s demographic “off-balance” liabilities are indeed huge, as this article in Handelsblatt argues (unfortunately only in German). The main points:

  • Greece is expected to be the third oldest nation on earth by 2030 (average age 50 years)
  • Greece ranks 55 in innovation (out of 142), behind Uruguay, Serbia and Mauritius (!)
  • Greece’s Math-PISA Rankings show a worrying trend (2006: 28; 2009:39; 2012: 42)
  • International patents (PCT) in 2012: Greece 628 vs. Germany’s 46620
  • Last but not least: with 1.4 births per woman Greece’s population is shrinking

We can safely assume that nobody in Greece’s political class (and the rest of Europe) is interested in honest accounting…


I think that Mr Katzarian gravely misunderstands the situation. He seems to think that by working hard and persuading politicians he can work on a turnaround of the country. He strikes me as somebody completely unfamiliar with bureaucratic logic and political processes. I seriously hope that I am wrong and he succeeds, although I think the chances of this are close to nil. I congratulate him on his paper profit and advise him to sell as much as he can as quickly as possible.

That’s the “nice guy” version.

The “bad guy” version is that Mr Katzarian has realized that he cannot turn his paper profit into cash due to a lack of market depth. There are probably not many people willing to risk large sums of money on Greece staying afloat. In this case, his only chance to get out at close to current levels is a write off of Greece’s debt ratio, such that sustainability is achieved and market liquidity in the bonds increased. The plan is good, there is certainly enough dumb money out there willing to lend money to a serial defaulter (see Ecuador, Argentina, Ivory Coast…). However, this requires that his claim is treated pari-passu with public creditors such as the ECB and the IMF for write-off purposes. I know, I know, that’s what the bond prospectus says. But what’s a contract worth to Mrs Lagarde or Mr Draghi? Not much, I am afraid…



  1. I also had watched Kazarian’s video at the time and summarized my conclusions (see below). I think you don’t give him enough credit even though the man exaggerates wildly and, thereby, damages his credibility.

    My basic premise is the following: the longer the maturities and the lower the interest rate, the more debt assumes the character of equity and no one can ever have too much equity. If all of Greece’s debt were refinanced with a new 99-year bond at a near zero interest rate, Greece’s debt would be tottally sustainable from a flow standpoint and it would dramatically diminish as % of GDP over the next 99 years.

    Secondly, what really matters – actually, the only thing which matters – is the amount of debt service which flows through the budget. Accruals don’t flow through the budget because they don’t get paid. So, if you add a 20 year interest grace period to the above structure, Greece would in effect have a temporary haircut to be followed up with very low interest payments after 20 years.

    A decision would have to be made, namely: what % of tax revenue should Greece be expected to allocate to debt service (interest)? How about half of the weighted average of EZ-countries? (just a thought out of the blue). I think anything below 10% would be fair and anything around 5% would almost be a give-away. BTW, your % of tax revenue figures CANNOT be correct. Greece, I believe, is slightly below 10% but it is quite difficult to calculate that accurately (and Germany, I believe, is much closer to 10%).

    In summary, Greece’s debt can easily be made sustainable by starting out with a sustainable % of tax revenues and working that back into interest rate and maturity structures. Why should that be done instead of reducing the debt via a haircut?

    The loss on Greek debt is already there, it just hasn’t hit the books yet. You can take the loss one-time upront and really hurt tax payers in the donor countries massively. Or you can spread the losses over 50+ years and the tax payers of the donor countries will soon forget about it. If Greece gets what it wants (i. e. minimal debt service through the budget), it should allow its official creditors to get what they want: a pretense for their tax payers that no debt has been forgiven and that there will be no tax increases for losses on Greece.

    Here is some food for thought: suppose that all of Greece’s debt in excess of the 60% Maastricht level were forgiven with the proviso that Greece would have to subsequently pay market rates on the remaining 60%. Depending on what you assume the market rate to be in that scenario, it could quickly cost Greece more than what it is paying now, i. e. before further debt relief as described above (5,7 BEUR last year). Can one expect a country to be able to service its debt as long as it is within the Maastricht level?

  2. Klaus,
    I feel you might misunderstand my position, which is understandable since I have not explicitly stated it yet. Therefore a short summary: I am also for a haircut in order to reflect economic reality (AND the losses in the creditor nation’s budget). Better, I would only have ONE requirement. Just one, no more: Greek Banks will not be allowed to participate in ECB tenders anymore…
    That’s it. Greece can keep the EURO, if so chosses (like Montenegro or (de-facto) Bosnia and Bulgaria-no big deal – just no access to the ECB system
    No negotiations, no humiliation, no troika, no conflict…

    Now to your points:
    It is not entirely true that for liquidity purposes the only thing that matters is the debt service flowing through the budget.
    Because liquidity is NO constraint anyway, for a government that it percieved as solvent (its banks can lend to it and get mone from the ECB).
    It is ONLY when a government is percieved as insolvent – like Greece now – that liquidity is a constraint, as it becomes relevant what “flows through the budged”, i.e. what money is due on short notice
    So we have to ask: what influences the percepton of solvency?
    Answer: many things. And the markets weighting of the different things changes over time. For instance, currently markets are not pricing in demographic off-balance liabilities of Germany and Austria (and other countries). sometimes in the future they will.
    The same for Greece. Just because the interest (of the long term, low-coupon bond) is an accrual (and does not lead to liquidity outflows) does not mean it should not enter solvency calculations.
    This is exactly the reason why unter every possible accounting regime accrual of interest is treated like an ordinary interest expense. There is no liquitiy flowing out, but it surely affects your solvency….
    Now to the figures. How did I arrive at the 15%?

    I took Greece’s GDP for 2013 (USD 242bn ~ EUR 200bn) and used the federal tax number as a percentage of GDP (30%) to arrive at a rough estimate of federal taxes (EUR 60 bn).
    Then I took the IMF interest payment forecast (FT-link in the post, on the bottom there is a table) for 2015, 2016 and 2017 which was around 9.1bn, 10.1bn and 10.8 respectively to arrive at an average of EUR 10billion. Dividing this IMF estimate by my tax estimate I get roughly 15%.
    Please tell me where the error is.
    I double-checked the result with the EU commission budget, mentioned in my discussion with Einhorn, where interest expense of GDP for 2015 was projected to be slightly above 4 percent of GDP. Dividing the IMFs 9.1 billion by the GDP, I arrive at a similar interest expense figure like the EU comission…

    Now, it is entirely possible that the numbers are wrong and the EU comission as well as the IMF have difficulties computing interest payments for Greece for the next three years. But, since it is more likely than not that the figure is correct (and I do not have time to investigate the matter) I rely on the official figures. As I said in the post, I do not know why Einhorn’s (and apparently your) estimates vary that much. You might look for differences here

    Furhtermore, it is simply NOT correct, that IFRS requires the DEBTOR to record the debt at fair value.
    It is another matter for the creditor (here you have a difference whether the debt is classified HTM or AfS)
    I have laid out the mechanism of IFRS accounting in the post. I think it absolutely makes sense.
    I do not know about this public accounting stuff Mr Katzarian mentions…
    Yes, there are so called CVA adjustments (maybe you mean that), but this practice is highly questionable and this practice arouse from the need to match debt and assets for perfectly hedged positions. Weak companies have seized on the opportunity and extended this principle to massage their numbers. But it is a dodgy practice. So dodgy that even regulators understand that: Bank regulation requires banks to deduct gains made from writing down of debt for the capital regulation…

    If Greece were indeed solvent, as Mr Katzarian claims, the market would be happy to refinance its current obligations to international lenders. It would realize, that it has much less debt and just labors under a liquidity shortage. Yes the market might be inefficient sometimes, but to think that it doesn’t grasp the numbers five years into the crisis does not happen in my experience…

    Last point, and many people do not understand this: the fact that there is the IMF and the ECB who are willing to lend an nutty rates, does not mean there is autimatically “more equity” for private bondholders. In case of a haircut, you can be sure the IMF and the ECB will NOT be treated pari-passu, i.e. bondholders are super junior here….
    Liebe Grüsse

    1. Just quickly regarding the %-of calcualtion: I did not come close to the details you outline. I only looked at the MinFin’s bulletin (pages 11&12) which shows total government revenues for 2014 of 76.109 MEUR (of which tax revenues 43.665 MEUR). Total interest expense is shown as 5.613 MEUR (gross; before ECB rebates and not counting interest income). So the interest expense was 7,3% of gross government revenues (or 12,9% only of tax revenues).

      Click to access GENERAL%20GOVERNMENT%20MONTHLY%20BULLETIN%20DECEMBER%202014.pdf

  3. Klaus,
    you link shows tax revenues of 43.6bn and social security revenues of 19.1bn. The sum comes pretty close to my back of the envelope calculation (60bn) I arrived using the tax revenue to GDP calculation. Social security payments are just taxes, nothing else The relationship between them and taxes labeled as taxes varies from country to country for political reasons…
    the 73 bn further include transfer reciepts of 8 bn, which I prefer NOT to include for the following reasons:
    * the transfer reciepts might be linked to specific items (protection of EU border, humanitarian aid…) which might not be available for debt service
    * they might be one time, i.e. not sustainable
    * there are also non-negligible transfer payments (3bn) Greece has to make.

    Now to the interest payments:
    the Greek fin ministry has expenditures of 5.6bn for 2014, whereas the IMF figures I refer to have EUr 7.6bn in 2014.
    Why the difference?
    The answer: cash vs. accrued interest.
    Look at pages 45 and 46 in the following link:

    Click to access cr14151.pdf

    There in Table 3 you have the IMF accrual of 7.6 for 2014, whereas Table 4 shows you the IMF cash interest figure which is, unsurprisingly, equivalent to the fin. min. figure of 5.6bn…
    Since accrued interest has to be paid back in the future, I consider it relevant in assessment of government solvency…liquidity managment is a separate exercise..
    BTW: the 10 bn in interest expense I use is due to the projections for 2015 and beyond, where said item increases. I understand this is due to the fact that some concessions made already have expired….
    Given that I want to estimate Greeces solvency under a going concern scenario, I think it makes sense to use projections…
    To sum up: Greece’s debt is far from sustainable even taking into account generous concessions already made….

    PS: Given that more than 50% of its gvt. bonds trade at negative yields Germany’s interest/tax reciepts ratio is headed to 0…..

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s