Citigroup issued $3 billion of a new 5-year senior bond issue on Tuesday. This was Citi's first offering since May, and the first billion dollar sized new issue from a major financial since Berkshire Hathaway came with a 5-year deal on July 30.
Access to the bond market is critical for all large financial institutions. Citigroup's issue proved that access is very expensive. The new issue was sold at a yield spread of 337.5bps over the 5-year Treasury. Prior to the announcement of the new issue, Citigroup's 5.5% bond due in April 2013 was bid at +275. Since the 5.5% bonds have approximately the same maturity and seniority of the new debt, its yield spread is directly comparable to the new issue. Hence the difference in spread levels means that Citigroup had to pay more than 60bps in extra yield in order to find enough investors to buy their debt. Not good.
This highlights some very telling facts about today's market. First, this is an extreme concession for a plain vanilla debt sale of a Aa3 rated bank. In 2006, the concession might have been 5 or 10bps at the most for a new issue. Alternatively, Baa-rated Deutsche Telecom recently brought a new 10-year issue, and the concession was around 15bps. This tells you that while there are buyers of Citigroup debt, they pretty much have to give it away.
Second, at a spread to Treasuries of +337.5, the deal has a very large negative basis to credit default swaps. This means that buyers of Citi bonds could also buy CDS and realize an arbitrage. Citi CDS closed Tuesday at 160bps and 5-year swap spreads closed at +98.5. You own the bond at +337.5bps, swap out the interest rate risk for 98.5bps, and the credit risk for 160bps. You are left with 79bps for free. This arbitrage won't last long, at least not at that level, so expect Citi cash bonds to tighten and CDS to widen in the near term.
Finally, this highlights an important risk of owning bank and brokerage bonds, and the opportunity that new issues offer. Under normal conditions, banks and brokers are routine issuers of new debt. The fact that accessing the bond market is currently so expensive is keeping issuers on the sidelines, but in fact, that's pent up demand for debt financing. As soon as conditions improve, expect a flood of new bond issues, all of which will come at a substantial concession to secondary trading levels.
But its common for the initial trading of the new debt issue to be tighter than the initial offer spread. The idea is that while Citigroup needed to pay a significant yield premium to clear $3 billion in bonds, once that initial supply has cleared, trading in the issue should resume at close to pre-supply levels. Indeed, the new Citi bond is traded initially about 10bps tighter than its original spread, although it now only slightly tighter.
So even if you believe we're already past the bottom for financials, tread carefully into financial bonds, and wait for new issues to make your allocation.
Access to the bond market is critical for all large financial institutions. Citigroup's issue proved that access is very expensive. The new issue was sold at a yield spread of 337.5bps over the 5-year Treasury. Prior to the announcement of the new issue, Citigroup's 5.5% bond due in April 2013 was bid at +275. Since the 5.5% bonds have approximately the same maturity and seniority of the new debt, its yield spread is directly comparable to the new issue. Hence the difference in spread levels means that Citigroup had to pay more than 60bps in extra yield in order to find enough investors to buy their debt. Not good.
This highlights some very telling facts about today's market. First, this is an extreme concession for a plain vanilla debt sale of a Aa3 rated bank. In 2006, the concession might have been 5 or 10bps at the most for a new issue. Alternatively, Baa-rated Deutsche Telecom recently brought a new 10-year issue, and the concession was around 15bps. This tells you that while there are buyers of Citigroup debt, they pretty much have to give it away.
Second, at a spread to Treasuries of +337.5, the deal has a very large negative basis to credit default swaps. This means that buyers of Citi bonds could also buy CDS and realize an arbitrage. Citi CDS closed Tuesday at 160bps and 5-year swap spreads closed at +98.5. You own the bond at +337.5bps, swap out the interest rate risk for 98.5bps, and the credit risk for 160bps. You are left with 79bps for free. This arbitrage won't last long, at least not at that level, so expect Citi cash bonds to tighten and CDS to widen in the near term.
Finally, this highlights an important risk of owning bank and brokerage bonds, and the opportunity that new issues offer. Under normal conditions, banks and brokers are routine issuers of new debt. The fact that accessing the bond market is currently so expensive is keeping issuers on the sidelines, but in fact, that's pent up demand for debt financing. As soon as conditions improve, expect a flood of new bond issues, all of which will come at a substantial concession to secondary trading levels.
But its common for the initial trading of the new debt issue to be tighter than the initial offer spread. The idea is that while Citigroup needed to pay a significant yield premium to clear $3 billion in bonds, once that initial supply has cleared, trading in the issue should resume at close to pre-supply levels. Indeed, the new Citi bond is traded initially about 10bps tighter than its original spread, although it now only slightly tighter.
So even if you believe we're already past the bottom for financials, tread carefully into financial bonds, and wait for new issues to make your allocation.
9 comments:
Why does the stock market look so rosy in the face of this analysis? I believe that Naked Capitalism has asked the same question.
What do you make of the wider cash/CDS basis? Is this due to lack of real money looking to buy bonds? Reluctance to put on basis trades due to counterparty risk on the CDS?
Now remember, I'm not a stock trader, but to me, the stock market doesn't look "rosy" just because it's off its lows. We're something like 18% off the highs? Any time you are looking at the stock market, its a point to point analysis, which really only tells us where we're been, not where we're are now.
The bond market is much better at telling where we are now. You can look at Citi at +330 and know that's an ugly number for a Aa-rated bank. Or Lehman at +400. The fact that Lehman was +450 is all find and dandy, but +400 is still a very wide number.
I guess that's where I come down on the stock market. Right now we're a good bit off the lows, but its like Lehman having rallied 50bps but still very wide.
Absolutely I think counter-party risk plays a big role. It will be interesting to see if they can get this semi-exchange going that is in the works. Right now its obviously not an efficeint market.
Great post. We put on 10mm of this basis trdae today -- I think ist a good one. The other brokers seem to trade at flat/zero basis and we put up 3% capital with our prime brokers on this. I know the basis can go more negative, but still prefer these type of trades to taking a directional credit view in this market. Perhaps a blog post on negative basis in general? Thanks for the idea!
I know I'm rehashing my earlier point, but the basis is all counter party risk. By trying to arb this with CDS you're pay for protection that will only exist if Citi defaults and your counter party is still in business. Is there really a scenario in which Citi defaults and we're not all living in the woods using deer skins as currency? it seems like a better arb would be to by the bond to do the asset swap and then buy the CDS on TIPS of the same maturity. (ie: synthetically write CDS at the ytm of the bond) it seems to me its Citi's spread to treasuries that's too wide not necessarily the basis (ie: odds are who ever sold you the contract is more likely to go under than citi is). I assume CDS on TIPS exists? Or am I completely misguided?
Also, pnl:
per my comment last week, i'm sure lehman can find a way to NPV the premiums on the CDS so it can book the income now. Worst case i'm sure they could sell the strip of premiums to someone. (ie: I think of CDS as a put where the PV of the premiums is the price the put is written for [i think the appropriate discount rate would be YTM - LIBOR, though I need to actually sit down and work it out one day])
Its very hard for me to see a catalyst for credit improving on its own. Short squeezes in the stock market seem to improve credit a bit, but every time its been temporary.
Eventually buyers will start to look past their monthly mark and start making the assumption that most of these banks are going to remain in business. From that perspective, credit is extremely cheap. Again, I don't know what will cause this other than time. So I'm just waiting this out a bit.
Its very hard for me to see a catalyst for credit improving on its own. Short squeezes in the stock market seem to improve credit a bit, but every time its been temporary.
Eventually buyers will start to look past their monthly mark and start making the assumption that most of these banks are going to remain in business. From that perspective, credit is extremely cheap. Again, I don't know what will cause this other than time. So I'm just waiting this out a bit.
Yeah I meant to say do a "Fixed for floating swap" above where I said "asset swap"...
should have read:
"it seems like a better arb would be to by the bond to do the Fixed for floating swap and then buy the CDS on TIPS of the same maturity."
Its just like being short a CDS contract further perceived deterioration of the credit actually helps you as long as citi doesn't default. But you are able write at the of the bond yield rather than the market price of the CDS. Accrued, am I missing something?
Post a Comment