I suppose we have to talk about Greece. Things occurred yesterday! The main things are here and here, basically the Troika is pleased that Greece has done everything right, to some approximation, and therefore they are disbursing some new loans and revising the terms of their old loans to make things even better, and all will be well by the time the aid program concludes in, I believe I have this right, 2057.1
The particular things that are or might be happening are, officially:
- Greece is getting €43.7bn in new loans from the European Financial Stability Fund,
- Its old, direct loans from other EU countries will have a 100bps lower interest rate than they used to, and its old and new EFSF loans will have a 10bps lower "guarantee fee cost."
- The old and new bilateral and EFSF loans will have be extended by 15 years, and the EFSF loans' interest will be deferred for 10 years.
- The other EU countries will give Greece the profits from Greek bonds they bought at a discount in previous support programs.
- Greece will try to buy back some public bonds in the open market at prices "no higher than those at the close on Friday, 23 November 2012," which means about 35 cents on the dollar (16.3% yield) for the ten-year.
This structure - except for the last part, which is just fun2 - is designed to accomplish the two perennial goals of:
- reducing Greece's debt, and
- saying you haven't reduced Greece's debt because it's, like, "illegal" or something to reduce principal of official loans.
I have made my own proposal for reducing the principal that Greece owes while not reducing the principal that its lenders are owed, because that is a thing you can do if you have levers like massively reducing the value of a debt through interest cuts and maturity extensions, but it seems to have been ignored. Maybe next EU conference! That's in, what, three months?
Felix Salmon discusses this magic of reducing the value of the loans by extending them past the halfway mark of our young century out of desire to avoid principal reduction, then adds "I don’t know if anybody’s done the math to work out what the effective NPV haircut is here." So: it's tempting to do so, because (1) I live to serve etc. and (2) a thing I actually like to do is pretend that things that happen in the official sector happen in the real world, and apply NPV math to them, and be all, what's up, NPV math. This is not, it seems, a very productive endeavor, but here we are.
It's particularly unproductive for Greece because of the confusion and obfuscation involved, but here is a stab in the dark; take this as less "here is what happened" and more "here is a template for doing really rough math if you had better data and were so inclined." But with that and many other caveats,
- if you take the €170bn of official-sector loans (but not including SMP bonds, IMF loans, or a few miscellaneous things) that I can readily get traction on,
- prior to yesterday, they were worth in a super-rough ballparky way €66bn-ish,
- and now they're worth €36bn-ish, for a haircut of €30bn, 18% of par, or 46% of what they were worth yesterday.
- Assumptions, errors, etc.!3
These are all fake numbers both because the data is patchy and because they're sort of not the point; the point is to return Greece to a sustainable path where it pays off all its debts, its borrowing costs come down, and when we all look back on this in 2057 we'll laugh about how pessimistic our discount rate assumptions were. Of course this is true in any restructuring: you don't understand, the borrower says, you'll actually get paid more this way, because we can actually pay you this reduced/extended amount, whereas the other way, man, good luck.
I suppose giving Greece ten years before they have to pay back any money, and forty-five years before they have to pay back all of it, is a helpful way to get to that point. Whether it is more helpful than keeping the schedule the same and just lopping €30bn off the principal amount of their debt is beyond my pay grade. There's no particular reason to think that those two approaches are equally helpful for Greece: you might care more about cash flow, or more about sheer amount of debt, and that would influence which approach you'd take. I'm just here to do the fake-o arithmetic, and the fake-o financial engineering, to show that on a certain narrow view of the world they're the same thing.
Eurogroup statement on Greece [EU]
Euro zone, IMF reach deal on long-term Greek debt [Reuters]
How the official sector restructures, Greece edition [Reuters / Felix Salmon]
A ‘broader concept of debt sustainability’ for Greece (feat. OSI-lite) [FTAV]
A thing that is almost like Greek debt math [Google Docs]
1. That's a year. You probably knew that? I dunno, it looked strange to me. 2057. I feel like I'm writing science fiction. Boring, boring science fiction.
2. Would you believe that that bond closed just over 35 today? You would? Would you expect it to close much below there between now and the end of Greece's buying efforts?
3. Oh gosh those. Really, better ideas or data would be welcome. So there are two main sources of funds, let's talk about them separately.
First there are EFSF loans, of which here is a sample document, which are at EFSF's cost of funds plus (i) some margin, which for Greece has been zero for a while, plus (ii) a "guarantee commission fee," which was lowered from 10bps to zero yesterday. Here is a table with amounts disbursed, maturities, etc. on those loans. Another €43.7bn was unlocked in the current deal but the terms of that are opaque to me; I am assuming a 2027 bullet maturity (pushed out to 2042) for no reason other than that that's what the last couple of tranches were.
Then there are the bilateral loans - the Greek Loan Facility - and related IMF loans. These are sort of described here but not in a way containing any information. The bilateral loans are at floating rates, basically Euribor plus some margin; that margin was 150bps and was reduced by 100bps so, I guess, it's 50bps. Their interest payments were not deferred, unlike the EFSF interest payments, though their maturity was pushed back.
For basic assumptions, because I don't know (1) future Euribors or (2) the EFSF's cost of funds, I've just used (1) Euribor swaps for the bilateral loans and (2) EFSF extant bonds for the EFSF loans, roughly matching maturities. For a discount rate, I've used 16.25% based on 10-year Greek bonds, for basically no reason.
For what changed, I assume that there was a 10bps reduction in commitment fees on the EFSF loans, a 100bps reduction in interest on the bilateral loans, a 15 year maturity extension on all of them, and a 10-year deferral of interest (followed by immediate lump-sum payment of deferred interest??) on the EFSF loans. I also assumed that the floating and floating-ish rates would be 12bps higher, which is just eyeballed from the difference between 30, 40 and 50-year eurozone swaps (Bloomberg IRSB, Euro Zone). Just for fun I also lowered the discount rate for the new world order to 13%, based on the Greek 30-year bond, which seems like a super-rough way to give credit for the theory that this will make Greek debt more sustainable etc.