Random Thoughts In My Head... Mostly to Do with Bond Insurers
I have been thinking hard about Ambac lately--mostly because I'm planning to take a position. Here are some random thoughts...
ABX Index
Felix Salmon of Conde Naste Portfolio magazine made an interesting trade suggestion for 2008. He thinks going long the ABX index may be worth considering (note that this is not possible for small investors but I found his quote of Alea insightful). For those not familiar, this is an index that tracks CDS of ABS (the one most relevant to us is the home equity ABX index (sub-label of HE)). It likely does not capture the CDS-type insurance that monolines write. I suspect that the monoline pay-as-you-go CDS-type contracts will post much lower losses (insurers typically deal with super-senior tranches, above what the AAA ABX index indicates, and you would also expect insurers to have some expertise in picking out some good MBS assets over bad ones (whereas the index just constitutes of everything that fits the index criteria). Nevertheless it's one data point that's worth looking at.
I suggest you check out the quote in his blog entry. Basically, what I wonder is whether the ABX index is hitting a bottom. There have been so much negative news, not to mention firesales depressing MBS prices (eg. E-Trade), one wonders whether most of the negativity is priced in.
Bond Insurance Not Viable?
I touched on people who think that the bond insurance business doesn't make any sense before. Nouriel Roubini posted an article last month commenting on this lately and let's revisit my argument.
Those arguing that bond insurance doesn't make any sense (eg. Bill Ackman) had a point. Or so they thought. Then something happened: Warren Buffett entered the picture.
Ever since Buffett set up shop, a lot of these arguments became hypocritical (either that or they are arguing against the investment merits of Warren Buffett which many of them say they are not).
What happened to the argument that any industry that "requires" AAA (even the bond insurers don't require it; their market will simply be smaller) isn't a viable industry? Are these people arguing that Warren Buffett's new business is also bogus and has no merit?
Nope. Most have backpadelled and now say that an industry that depends on AAA rating makes sense for the municipal bond market but not structured products. How much do you want to bet that they will backtrack on the structured product comment as well in 2 years?
Everything Depends on the Input Variables
Reggie Middleton has a fairly long article critiquing Ambac. This time he looks at Ambac's non-real-estate insured exposure.
The problem with the debate at this point is that both sides, bull and bear, can be right. We just don't know which one will win in the end. For instance, a big disagreement between me and someone like him is the notion of how accurate market prices reflect reality. He believes that market prices are what really matter, whereas people like me think the market is irrationally pricing many real estate assets, including anything to do with bond insurance. I am generally of the opinion that market prices are quite rational most of the time, but are irrational during market stresses--like now!!!
To see what I mean, consider his analysis of Ambac's asset value if it were based on current market prices. Since the market is very pessimistic, the prices are very low right now. Reggie Middleton picks E-Trade's firesale prices to price Ambac's assets. Needless to say, this yields a massive drop with Ambac's insured portfolio dropping from $69 billion to $9 billion. I don't think I have seen that big of a drop in anything in my life (notwithstanding, GM and their bogus $39 billion loss recently).
The problem is that it's all conjecture at this point. Even though a lot of people are throwing around so-called "hard numbers", none of it means anything because it is dependent on small changes in variables like subprime default rates, loss recovery, and so forth. I don't recall off the top of my head but I believe the key number of subprime default for Ambac (and other insurers) is 19%. That was the number that S&P (I think it was them) used in their stress test. The subprime default rate was something like 14% in mid-2007 (but most of the exposures held by the monolines are older vintages and their default rate is much lower right now). If the rating agency models are somewhat believable--not in their conclusions but in their statistical modelling--then I'm comfortable taking a position in these bond insurers. Ambac (and others) should be able to absorb the losses, assuming they can raise the $1 billion they need... however, if default rates (or other inputs into these models) end up higher, then the mononlines will take more losses and need more capital.
So, everything comes down to the actual losses and defaults. Since no one knows what that will be, anyone can say anything. People like Warren Buffett and Benjamin Graham entered these situations by buying with a steep discount to intrinsic value. Although these bond insurers likely aren't trading at big enough discounts for those classic value investors (or else Buffett would have bought out one of the insurers instead of opening up his own), I'm feeling comfortable given that their price to book value is around 50%.
Potential Future Problems for Ambac
As I have remarked before, anyone taking a position in a bond insurer, or already a shareholder, needs to think about the future losses in other areas. Potential problem areas are credit card debt, commercial real estate debt, student loans, and so on.
For Ambac, student loans are sizeable but I am not too concerned with them for now. Their credit card exposure based on this website (not sure how comprehensive the numbers are) is small ($1.2 billion).
Where Ambac may run into problems is car loans. They seem to have $12 billion of par outstanding, mostly rated BBB by Ambac (internal rating).
ABX Index
Felix Salmon of Conde Naste Portfolio magazine made an interesting trade suggestion for 2008. He thinks going long the ABX index may be worth considering (note that this is not possible for small investors but I found his quote of Alea insightful). For those not familiar, this is an index that tracks CDS of ABS (the one most relevant to us is the home equity ABX index (sub-label of HE)). It likely does not capture the CDS-type insurance that monolines write. I suspect that the monoline pay-as-you-go CDS-type contracts will post much lower losses (insurers typically deal with super-senior tranches, above what the AAA ABX index indicates, and you would also expect insurers to have some expertise in picking out some good MBS assets over bad ones (whereas the index just constitutes of everything that fits the index criteria). Nevertheless it's one data point that's worth looking at.
I suggest you check out the quote in his blog entry. Basically, what I wonder is whether the ABX index is hitting a bottom. There have been so much negative news, not to mention firesales depressing MBS prices (eg. E-Trade), one wonders whether most of the negativity is priced in.
Bond Insurance Not Viable?
I touched on people who think that the bond insurance business doesn't make any sense before. Nouriel Roubini posted an article last month commenting on this lately and let's revisit my argument.
Those arguing that bond insurance doesn't make any sense (eg. Bill Ackman) had a point. Or so they thought. Then something happened: Warren Buffett entered the picture.
Ever since Buffett set up shop, a lot of these arguments became hypocritical (either that or they are arguing against the investment merits of Warren Buffett which many of them say they are not).
What happened to the argument that any industry that "requires" AAA (even the bond insurers don't require it; their market will simply be smaller) isn't a viable industry? Are these people arguing that Warren Buffett's new business is also bogus and has no merit?
Nope. Most have backpadelled and now say that an industry that depends on AAA rating makes sense for the municipal bond market but not structured products. How much do you want to bet that they will backtrack on the structured product comment as well in 2 years?
Everything Depends on the Input Variables
Reggie Middleton has a fairly long article critiquing Ambac. This time he looks at Ambac's non-real-estate insured exposure.
The problem with the debate at this point is that both sides, bull and bear, can be right. We just don't know which one will win in the end. For instance, a big disagreement between me and someone like him is the notion of how accurate market prices reflect reality. He believes that market prices are what really matter, whereas people like me think the market is irrationally pricing many real estate assets, including anything to do with bond insurance. I am generally of the opinion that market prices are quite rational most of the time, but are irrational during market stresses--like now!!!
To see what I mean, consider his analysis of Ambac's asset value if it were based on current market prices. Since the market is very pessimistic, the prices are very low right now. Reggie Middleton picks E-Trade's firesale prices to price Ambac's assets. Needless to say, this yields a massive drop with Ambac's insured portfolio dropping from $69 billion to $9 billion. I don't think I have seen that big of a drop in anything in my life (notwithstanding, GM and their bogus $39 billion loss recently).
The problem is that it's all conjecture at this point. Even though a lot of people are throwing around so-called "hard numbers", none of it means anything because it is dependent on small changes in variables like subprime default rates, loss recovery, and so forth. I don't recall off the top of my head but I believe the key number of subprime default for Ambac (and other insurers) is 19%. That was the number that S&P (I think it was them) used in their stress test. The subprime default rate was something like 14% in mid-2007 (but most of the exposures held by the monolines are older vintages and their default rate is much lower right now). If the rating agency models are somewhat believable--not in their conclusions but in their statistical modelling--then I'm comfortable taking a position in these bond insurers. Ambac (and others) should be able to absorb the losses, assuming they can raise the $1 billion they need... however, if default rates (or other inputs into these models) end up higher, then the mononlines will take more losses and need more capital.
So, everything comes down to the actual losses and defaults. Since no one knows what that will be, anyone can say anything. People like Warren Buffett and Benjamin Graham entered these situations by buying with a steep discount to intrinsic value. Although these bond insurers likely aren't trading at big enough discounts for those classic value investors (or else Buffett would have bought out one of the insurers instead of opening up his own), I'm feeling comfortable given that their price to book value is around 50%.
Potential Future Problems for Ambac
As I have remarked before, anyone taking a position in a bond insurer, or already a shareholder, needs to think about the future losses in other areas. Potential problem areas are credit card debt, commercial real estate debt, student loans, and so on.
For Ambac, student loans are sizeable but I am not too concerned with them for now. Their credit card exposure based on this website (not sure how comprehensive the numbers are) is small ($1.2 billion).
Where Ambac may run into problems is car loans. They seem to have $12 billion of par outstanding, mostly rated BBB by Ambac (internal rating).
Siv -
ReplyDeleteIt sounds like you're about ready to pull the trigger. Good luck!
I know you like history, recalling Buffett and AMEX. I came across this from 1991, which has eerie similarities to what we're going thru now.
UKHousingBubble