Monday, December 24, 2007

What is a Tender Option Bond (TOB)?

(Alternative title for long-time readers: There's a meteorite that hit the ground near here. I want to check it out. It won't take long.)

A tender-option bond (from now on TOB) is the municipal bond market's answer to the classic borrow short and invest long. As with many types of leveraged strategies, this one has been getting hit very hard in 2007. It also has some disturbing parallels with the SIV problem which has rocked the money markets.

What is it?
First, here is what a TOB is. You start out with a long-term tax-exempt municipal bond, usually highly rated. Let's say you buy it at par with a 4.5% coupon. 30-year tax-exempt munis are widely available at that price right now.

Then you sell senior notes to some other investor with your long-term muni pledged as collateral. The amount is somewhat less than the value of the collateral pledged. The senior notes have the same maturity as your pledged collateral, but the interest rate floats every seven days. Typically the floating rate is set by a dealer firm (called the "remarketing agent") based on whatever rate will clear the market. The senior note holders also get a put option, also called a tender option. The senior note holders can put their bonds back to the issuer at par at any time with settlement on the next reset date.

If you are familiar with municipal floaters, you know this is a very common structure. It goes by the name VRDO (Variable Rate Demand Obligation) or VRDN (N=Note) or VRDB (B=Bond). It is used by normal issuers who want variable rate debt as well as TOBs. The idea that the bonds would always be puttable at par probably sounds funny to some readers, so think about it this way. It really just like a revolving CP program where the issuer retains the same amount outstanding all the time. While technically in a CP program, the old CP matures with proceeds from the new CP, as long as the capital markets are fully operational, the CP issuer effectively just resets his interest rate.

So back to the TOB. You remember the size of the senior issue was less than the size of the collateral pledged, which leaves us with some residual. That is sold to junior note holders, who in essence have leveraged exposure to the original municipal bond.

These programs can be structured as single name deals, where a single long-term muni has been pledged vs. a single senior VRDO. Or it can be structured where there is a pool of long-term munis pledged against a larger senior VRDO. From a pure safety perspective, the senior holders would obviously prefer the pool, but there are legitimate tax concerned about the pooling structure. Legally, in order for the tax-exempt status of a bond to pass through any kind of structured product, the credit risk of the municipal has to be retained by the investor. Thus the senior/subordinate structure of the TOB can get sticky, especially if its pooled. Some tax attorneys argue that when you pool this kind of program, the senior holders are not actually taking risk on all the municipals in the pool, since its usually the case that several could default without senior holders getting hurt. This senior/subordinate concept is not unlike a CDO structure in terms of how the senior notes are protected.

VRDOs are normally backed by some sort of credit enhancement, which is different than the classic bond insurance. In the case of a VRDO, the backing is normally from a bank, which can come in various forms: a letter of credit, a stand-by purchase agreement, or just a liquidity facility. All of these have similar functions for the investor: it ensures that if investors want to put their bonds back, that someone with capital is there to buy them.

The junior notes are most likely held by a hedge fund. There are many hedge fund for which this is their only strategy, and they spend most of their day creating these TOB structures. Since the junior notes have considerable interest rate risk, the hedge fund generally cooks up some means of hedging this risk. Unfortunately, the world of municipal derivatives is pretty murky, so often the hedge fund winds up using LIBOR swaps or some such as its hedge. This introduces the risk that taxable bond rates move differently than municipal bond rates.

In addition, long-term municipal bonds are usually callable by the issuer starting in the 10th year. The option risk inherent in the municipal makes hedging with non-callable taxable instruments like swaps a real challenge.

TOBs are generally created using AAA municipals as collateral. This has to do with the desires of both the senior and junior note holders. The senior holders, usually money markets, generally want very highly rated securities. Ratings agencies will generally rate the TOB senior piece the same as the underlying collateral. The hedge fund buying the junior piece also wants to avoid credit problems. The arbitrage of a TOB is all about the slope between short-term and long-term bonds. Introducing credit risk merely complicates an essentially simple arbitrage.

Now here is where the problems start...

Current Troubles
Up until now, finding AAA-rated long-term munis was easy, because so many munis were insured by AAA-rated monolines. I believe its around 40% of all municipal issues are insured. But now we're in a world where that AAA rating is imperiled. This has caused the municipal bond market to decouple from the taxable bond market, perhaps not entirely, but to some degree. Whereas historically municipals tend to trade around 80% of Treasury rates, currently the number is above 90%. Long-term municipals are widely available at or slightly above the 30-year Treasury rate.

This means two things for TOBs. First, money market funds are increasingly unwilling to hold short-term TOB debt. Makes sense right? If you were running a muni money market fund, and you could get out of any TOB debt at par right now, wouldn't you? You know that there is a decent chance some of the TOB bonds are about to be downgraded. You also know that the guy across the hall who ran your prime money market fund just got fired over the whole SIV thing. I'd dump those TOB bonds as fast as I could.

Second, the TOB hedge fund's hedge position isn't working. Municipal yields are not moving with taxable yields and this is creating big losses for the TOB. And we know what happens when hedge funds take losses: investors start pulling out. So you've seen TOB programs having to force liquidate bonds.

Bad news. TOBs represent 8% of the total municipal bonds market, or about $200 billion. The muni market isn't known for its liquidity, so if you have a large number of bonds being dumped on the secondary market, the whole market cheapens up. This is exacerbating any credit concerns market participants might be having.

Parallels with SIVs
The parallel with the SIV problem is hard to miss. Money market participants eschew the debt. Vehicle can't roll over. Winds up liquidating into an illiquid market. Exacerbates an already tenuous credit environment.

The good news is that unlike SIVs, the credit quality of municipals remains very strong. While SIVs were involved in the CDO and sub-prime markets, where there has been unquestionable and material deterioration, municipals don't really have this problem. Property tax collections may wane a bit, but the odds of this rising to a level where any cash flow to bond holders is ultimately impaired is remote. The only real problem in munis are the bond insurers. Even if one or more bond insurers were to disappear, the downgrade in most municipals would be from AAA to AA or A. Not good, but not the end of the world.

The bad news is that the bank credit enhancers may wind up owning the TOB bonds. See, if the hedge fund which is operating the TOB program can't make good on the short-term debt, perhaps because it can't liquidate all its long-term bonds and hedges at less than 100%, the bank will probably wind up possessing the underlying bonds. Again, this won't turn out like the SIV problem, where bank sponsors of SIVs are taking bonds onto their balance sheets at steep discounts. But still, cash-strapped banks are likely to just blow out the muni positions, creating more of a supply problem in the long-end of the muni curve.

Sell my Munis? Sell my muni money market?
I'm surprised to find you squeamish, monsieur, that is not your reputation. I think munis are a screaming buy for long-term holders. You are rarely going to get the chance to buy long-term munis at prices about equal to Treasuries. That isn't to say this is the absolute bottom, but I think as we finally resolve the bond insurer issue (one way or another) municipal buyers will come back in. Remember, there is no good substitute for municipals. They are the only tax-free game in town.


Anonymous said...

Again, a flat out four plus fascinating post. Bravo!

The sentence in the whole post that resonated into my addled brain with a vengeance however was:

Property tax collections may wane a bit, but the odds of this rising to a level where any cash flow to bond holders is ultimately impaired is remote.

I sincerely pray that statements such as that are etched in stone.

Anonymous said...

Great post.

Would you mind to post a small example with some numbers to show how a TOB structure would work?

Thx, J

flow5 said...

Your posts are the only ones worth reading 3x

Anonymous said...

Thanks for the info. Just wanted to ask a basic question which crops up every now & then...
what do they mean when it says the CDO is funded /unfunded? do same terminology applies to bonds too?

Anonymous said...

Fred, funded vs. unfunded means that the liabilities are either cash or synthetic (Credit Default Swaps). The investor in the CDO tranches could be synthetic (insurance sold by rated counterparty) or cash collateralised. You generally don't use the same terminology as bonds because bonds are necessarily funded.

To the author, come off it. The paralells between the SIV and the TOB are most strikingly that you have a long term asset, or set of long term assets funded by short term debt. Therefore you are subject not only to interest rate risk which is tractable to manage but credit risk on your seniors which is more difficult to manage. I haven't investigated the back-up provisions that banks have provided to these vehicles but unless they are air tight obligations bet on them to balk if the muni market trends wider and people start charging for senior debt.

Anonymous said...

There will be a piece on HedgeWorld later today (28 Dec) giving the much more pessimistic views of an analyst called Christian Stracke (CreditSights).
Personally, from last Nov, I figured the mortgage-backed issue would reach far wider than just a few lenders/CDO funds, and can still see contagion reaching credit card/auto receivables (HedgeWorld, July).
But like the poster, I think that while many put options on VRDOs are likely to be exercised, there will be a market for them among investors with a longer time horizon. The alternative is that cash-strapped municipalities are hamstrung--possible, but personally I would bet against it. So while in the short term RowdyRoddy may be right, my guess is that the potential liquidity issue will also be resolved in the short term.

Accrued Interest said...

I think the major difference with munis is that the underlying collateral is fundamentally good. In other words, TOBs were created, mostly, with stuff like State of New York credits. Even if the insurance is gone, the State of NY is still rated AA.

On property taxes, states/counties have a lot going for them. In most cases, property tax assesments are not done every year. In Maryland its every 3 years. So normal property tax collection hadn't fully caught up to the price appreciation in 2005 and 2006, so the price depreciation in 2007 is mitigated.

Some localities will have properties that are just abandoned. That's a problem, but cities have dealt with this before, its just normally commercial buildings that are unused during a recession. Besides, do you honestly think cities and states can't cut a little spending or hike taxes? The California problem in 2002 was a MUCH bigger deal than this will turn out to be.

Anonymous said...

hi RowdyRoddyPiper, thanks for your help!

Anonymous said...

I'm going to state it straight away, I have no idea what the credit quality of Munis is and know nothing of the muni market outside of having one friend that trades them. He claims being a scratch golfer is his most important qualification. I'm just thinking that going off of the ratings seems a little naive, subscribing to the once bitten twice shy school of thought. If an RA can't get subprime right, what makes one believe it can get munis rated correctly?

That being said when you lose the wrap you lose a lot. I have a much different risk profile if I have to have a joint default between two somewhat independent borrowers versus relying on a borrower alone. I don't worship at the altar of free markets exclusively but have a hard time believing that muni issuers would pay for wraps if there weren't a real economic advantage to it.

So you are putting large amounts of leverage on assets that are at least to the casual observer a little hard to get a read on and the leverage you are putting on is short term and subject to repricing. This idea that the assets pledged here are sound credits and that everything in the SIVs was garbage is too black and white. To be honest most of the assets that are pledged into SIVs right now will repay at par, the issue is that there has not been a process of sorting the good from the bad credits yet. I would suspect that if there were a muni credit crisis that most stuff collateralizing these would face a similar lack of discrimination.

SIVs were operating with something in the area of 10X-15X leverage. What is the leverage on these vehicles?

Unsympathetic said...

Speaking of muni's, AI, did you notice this little tidbit?

IMHO this is the grenade with the pin removed thrown into the world of structured finance. If you take away all future muni revenue streams from AMBAC/MBIA, how do they not go BK, let alone retain AAA? And, CDS/CLO finally do act as WMD... just as he kept saying.

Anonymous said...

People who signed mortgages that they either knew they couldn't keep up the payments on (speculators) or didn't know they couldn't keep up the payments on (naive homebuyers) are different from municipalities who have access to the tax power.

Or are they?

It seems to me that an argument can be made that the same model exists for both: risk is mitigated by treating populations instead of individuals.

What did tax revenues do in the 30's?

Accrued Interest said...

Only if you are drawing up a 1930's type scenario should you assume systemic increases in municipal defaults.

Here is a quick list of the General Obligation issuers who have defaulted since 1980:

Anonymous said...

I don't know whether we have to go back to the 1930's or not.

My problem is I don't see any vertical blue stripe on this graph yet:

I worry that 1930 might not be the right reference.

People starve when they can't afford the price of necessities. The price of necessities is going up faster than the shade of Benazir Bhutto is going...well, either up or down depending on who you are.

Once again, I sincerely hope that the quick list stays the same length it is now.

Anonymous said...

Does anyone have an opinion on Andrew Cuomo's subpoena of 18 banks and securities firms in regards to the Auction Rate amrket? Whether their could be a problem here for some of these ARPs having to be taken on balance sheet?

Accrued Interest said...

I'm no lawyer, but I don't see how this could happen. An investment bank performs a sales service on behalf of an issuer. I don't think there is any situation where the issuer could force the investment bank to take securities it sold back onto its balance sheet.

PetRock235 said...

Orance County CA defaulted in 1995 after its investment fund lost billions (invested in derivatives) and they declared bankruptcy.
That was the biggest default since New York City and WPPS.

Vallejo CA declared bankruptcy in May 2008.

Accrued Interest said...


Neither New York nor Orange County missed any payments on their "General Obligation" debt. Neither has Vallejo so far and honestly I'd bet that they don't.

neroden@gmail said...

One of the interesting things, reported recently, is that BBB-rated municipals have been less likely to default than AAA-rated corporates, *overall*.

This means that the 'insurance' and so forth is basically irrelevant. Your munis are going to continue to make their payments, almost certainly.

The question *I* have is about the structure of these deals. I understand that a GO bond is backed direct by tax revenue, and a Revenue bond is backed direct by some specific substream of municipal revenue. But what's with the intervening structure?....

When a SIVs which held *good* underlying assets finds itself in default, or even declaring bankruptcy, due to inability to refinance, a lot suddenly depends on whether the holders of the SIV's debt can walk off with their fractions of the underlying assets, or not.

If a fund is investing in "second-order" securities carved out of good municipal bonds, it is imperative that the fund be able to collect the actual municipal bonds if the whole thing goes under -- otherwise the fund is subject to credit risk and management risk *other* than the municipal government's risk.

I haven't figured out from your explanation whether that is the case or not!

neroden@gmail said...

To explain what I'm thinking a little further...

For a municipal money market fund, the put option is potentially worthless, because it depends on the (likely terrible) creditworthiness of the bank issuing it.

However, the ability to get at the collateral is very worthwhile. If there's no urgency, the fund can just sit on a small amount of long-term debt and collect the interest. If there's a huge run on the fund and stuff has to be sold to cover withdrawals, *someone* is going to be willing to buy that high-grade collateral at par or better -- likely the money market fund's parent company.
So it means: no loss of money.