AIG and the Lunacy of GAAP Reporting
Recently, Tom Gallagher, an analyst from Credit Suisse lowered his earnings-per-share estimate and price target for American International Group (AIG), due to elevated losses from its derivatives business. According to Gallagher, "Virtually all the aforementioned indices have declined this quarter, with non-AAA (rated) being hit the hardest, declining between 3%-25%."
This news bodes poorly for AIG's 3rd quarter GAAP earnings in that they will likely have to take additional mark to market losses on both their investment portfolios, and CDS/mortgage insurance portfolios. Keep in mind though that AIG has tremendous underlying revenue and cash flow to help offset the write-downs but that is not the point of this article.
AIG is an excellent example of a company whose fundamentals and long term prospects are completely ignored due to overdone fears about short term GAAP earnings, which are far different from the financial reality and true "owner's earnings" of the company. ("Owners Earnings" are best described by Bruce Berkowitz as the money that is left in the grocery store's cash register after all expenses and salaries have been paid.) AIG holds a long term outlook on its investment portfolio, as opposed to a Lehman Brothers (LEH) or Merrill Lynch (MER) that partakes in proprietary trading activities and got caught holding securities that they were hoping to unload in the very near future.
By nature of being an insurer, AIG is forced to be relatively conservative in their investments to protect their policyholders. Holding predominantly AAA securities AIG has indubitably been conservative but the dislocations in the credit markets have drastically distorted the reality of their investment portfolios performance.
In addition to the huge investment portfolio that has a tremendous amount of exposure to residential mortgages, AIG also insures mortgages and CDOs which have seen their prices deteriorate in the current environment. AIG's Financial Guaranty business is somewhat similar to that of an Ambac (ABK) or MBIA (MBI) in that these are not securities that they could sell even if they wanted to, but are forced to mark to market because of GAAP.
Instead of trading these securities like an investment bank would, AIG plans on holding the securities until they mature many years into the future. By the nature of how the super senior securities are structured, they are usually protected from the first 15-30% of losses on a given group of CDOs, and are only responsible for interest and principal when they become due. The time value of the payments and the lack of acceleration clauses are some of the attractive attributes of the business.
One key difference from the portfolios of some of the other guarantors is that 71% of AIG's RMBS exposure in CDOs is from 2005 and earlier which are significantly more secure assets then the 2006/2007 variety. These are super senior positions in which there is no available market to correctly assess their current value even if that were truly relevant for a participant that fully intends on holding the security till maturity. Instead of examining the performance and cash flow of the relevant securities AIG has to mark these positions based on Markit's ABX indices which are extremely illiquid, and that are constructed with very different collateral and structures then the super senior positions that AIG and the other Financial Guarantors insure. Further decreasing the relevance of the ABX indices is the fact that many of the large banks with significant residential mortgage backed securities exposure have attempted to hedge their bets by shorting these indices.
These huge institutions are pushing down the prices of the indices to levels that are in no way indicative of the future cash flows that can be expected from somebody holding their linked securities till maturity. Instead of mirroring the performance of residential mortgage backed securities, the indices are creating a sort of reverse Ponzi Scheme in which firms are forced to bet against their own mortgage backed security exposure by selling the indices short, which in turn causes further write downs of their existing positions based on the indices declines as a result of all the selling, forcing capital raises at consistently lower prices injuring existing shareholders.
If this is confusing to you, that is totally understandable as it is befuddling to me as well, but this is what an ardent reliance on GAAP and Fair Value Accounting has created in the midst of the greatest financial panic since the Great Depression. It is obvious that any company with this type of exposure to the residential mortgage market is likely to be seeing losses and AIG acknowledges that, but the clear and obvious reality to anyone who takes an objective look at the marks to market can see that the losses already booked are in no way indicative of the future cash flows that the securities are likely to bring in.
I don't blame the analysts covering AIG for being focused on the short term GAAP outlooks and complete disregard for the true value of the company. Bold bearish proclamations are easier to sell to their clients and that is also what gets their name in the news i.e. Meredith Whitney in Fortune Magazine. It brings a certain level of celebrity to be the most bearish on a big company like AIG, just like it is sexy to be the most bullish in a bear market like a Henry Blodgett on Amazon. By observing the facts however, an objective analyst can see that this obsession with GAAP is completely distorting the true financial situation of AIG, in addition to many other credit worthy institutions.
Let's break down a few of the numbers to elaborate my point. AIG's Financial Guaranty business has been the main contributor to the mark to market losses of the last few quarters. CDS positions are basically insurance policies on the super senior portions of CDO portfolios, backed by true assets, which in most cases are mortgage backed securities. Through the second quarter AIG has taken $24.8 billion in marked to market losses on this portfolio. AIG's own cash flow driven analysis leads them to believe that actual losses on this portfolio are likely to be around $5 Billion. In their stress case scenario, using extremely conservative default and severity assumptions, they estimate their future losses to be approximately $8.5 billion.
To give an example of their stress case analysis for their subprime exposure, AIG is assuming that 96% of currently delinquent mortgages will default and that net recoveries on a foreclosure will be 28 cents on the dollar as opposed to the 50 cents that is the current national average on foreclosure recoveries. These assumptions seem to be quite Draconian but it takes real work to come up with default and severity rates that in any way can come near to the expected losses that these indices are pricing in, which once again has very little relation to the actual performance of the underlying assets that are insured.
Assuming AIG's stress case estimates prove to be accurate, this would mean that $16.3 Billion of already incurred losses would eventually pass through the income statement and back on to the balance sheet as earnings. This is quite an abstract painting that GAAP is providing in assessing this firm's financial condition.
In its Insurance Investment Portfolio, AIG has suffered from $10.3 Billion in other then temporary impairment losses. 87% of theses charges represent severity charges which are driven from market declines of a certain amount, but that are not at all indicative of the actual performance of the underlying collateral, which has remained AAA or AA rated, and that is performing well. This means that the Investment Portfolio's losses are potentially overstated by approximately $8.9 Billion.
Now I don't believe that anybody could accurately forecast with precision what the final tally of losses or gains will be in these matters. There is no doubt however that the use of GAAP and Fair Value Accounting is creating massive distortions in regards to the true financial position of the Financial Guaranty companies and in this case AIG.
I could go on and on about AIG's business which is one of America's great growth stories of the last half century, but because of limited time I have focused on the largest short term issues just to show why in my opinion analysts are getting the true story all wrong. As time goes by and the true performance of the underlying collateral becomes more important then the temporary views of traders on a completely illiquid market, we will see the real economic fundamentals of AIG. If you have a long term outlook and the backbone to realize it is impossible to pick the bottom, I believe that this is a once in a lifetime buying opportunity of a top of the line insurer with an immense geographic footprint.
At a price of around $20 per share AIG is trading at 2/3rds of a fictionally reduced book value, and I believe that moving forward they should be able to grow book value by 10-13% a year moving forward.
Disclosure: Long AIG, MBI and ABK; short puts on AIG, MBI and ABK.
Related Articles
|
Top Rated Comment Streams:
-
1.Hedged In661
- 2.
-
3.Smarty_Pants399
-
4.cos1000305
-
5.axelrod608296
This article has 23 comments:
-
cranky investor
-
43 Comments
My Website
Aug 27 07:47 AM-
TimButler
-
6 Comments
My Website
Aug 27 10:14 AMFASB mark-to-market accounting rules do their level best to give an accurate picture of a balance sheet at a point in time, namely today. However, the equity value of a company is a function of all of it's future cash flows, discounted to today's present value. The distinction between the two numbers can be, and usually is, large.
Notwithstanding Robert Merton's argument suggesting that capital market stock prices accurately reflect all the information available about the value of a stock today, I concur with the author of this Seeking Alpha article, in suggesting that Wall Street today is paying too much attention to today's mark-to-market book equity, along with the risk of 1 to 2 year sustained EPS losses, rather than focusing on longer run cash flow generation from AIG.
Of course it's a matter of opinion, I just happen to think that Hank Greenberg, Warren Buffet and Robert Willumstad might well agree.
Disclosure -- As of Friday, August 25th, I'm long AIG too. I'm trying to make up for buying AIG Jan 60 '08 Calls at $10 in September '07!
Timing is everything, and differing views is what makes a market. I've placed my bet, and I like to think of it as an investment, though I'll be the first to admit that it certainly does have a speculative component.
-
fxtrader07
-
618 Comments
Aug 27 10:42 AMWhat is the 'correct' current value of AIg based on their expected normalized earnings, normalized adjusted book value and cash flow? Has anyone figured that out with a reasonable degree of accuracy? Can the author tell us anything about that other than he thinks AIG's stock is 'too cheap'? If not, what is that notion of 'too cheap' based upon?
-
eternitus
-
52 Comments
Aug 27 11:11 AMThe thing that's holding me back from buying more shares is uncertainty over how much capital the mentally impaired rating agencies will force them to raise based on these GAAP numbers market to an inefficient, illiquid and frightened market. They've already had to dilute the heck out of investors by raising $20 billion at a low stock price.
-
Jason For Web Site Design
-
3 Comments
My Website
Aug 27 11:53 AM-
1 world currency
-
297 Comments
Aug 27 12:10 PM-
Timothy Travis
-
7 Comments
Aug 27 12:34 PMMerrill's fire sale was from the 2007 vintage which is significantly worse then the 2005 vintage. Even with that being said I think that it was foolish for them to sell those securities at those prices in addition to financing them. I think that it was more of an institutional decision by John Thain to wash his hands of the securites, as opposed to an actual business decision in which they would try to maximize value for their shareholders.
Investment Banks use far more leverage then insurers and have very different capital requirements in relation to mark to market losses then insurers, so it is not really an apples to apples comparison. Just think about the fact that these are secured first lien mortages for the most part. The CES and HELOC exposure is more risky but they also have more protection built into the structures, but on the first lien exposure come up with a default rate and severity that can possibly justify those prices. It is nearly impossible to come up with justification for the current pricing.
-
Timothy Travis
-
7 Comments
Aug 27 12:55 PMI would have touched on some of this but the article was long enough as is. The market cap is $55 Billion as it stands right now. Their stated book is $78 billion to which I would conservatively add back $15 billion to come to about $93 billion adjusted book value.
This provides me with a considerable margin of safety as is and AIG has consistently proven that they can return 11-15% on equity over long periods of time. I'd estimate that their normalized earnings are arond $12-$15 billion per year assuming no growth which is probably far too conservative given the companies track record of growing earnings. Put a 10 multiple on those earnings and you will probably be lot closer to a reasonable market cap then you are right now. The point is the margin of safety is so large that it would be a waste of time and effort to try and come up with a precise DCF analysis. I'd be lying if I said I could accurately forecast exactly how this business will look in 5 years and what the cumulative losses will be. I feel extremely certain however that by observing the information available and making reasonable forecasts given the current strains on the US housing market and projecting further declines, that AIG is at least 50% undervalued under the most dire circumstances. The losses already taken are vastly overstated any reasonable assumptions. The capricousness of the ratings agencies could force a capital raise which would hurt shareholders but would also provide another buying opportunity for dollar cost averaging at absurd prices.
-
jimmy46
-
255 Comments
Aug 27 02:12 PM-
Greg Harris
-
52 Comments
Aug 27 03:08 PMThe only problem I have with this article is I wonder if there needs to be more analysis on what the actual CDOs have underlying them. The real problem with the CDOs is that they're collateralized not by straight mortgages or assets, but tranches and pools of asset backed securities in the first place. A CDO may be an overcollateralization of tranches of securities that are all about to get wiped out, because the securities that back the CDO are early in line for default. If you're collateralizing the CDO with tranches of junk, the overcollateralized super senior portion of the CDO get's wiped out too, because overcollateralization of zero is still zero, or close enough. It's the senior portions of the original AB security that will prove to be collateralized better, not the CDO and it's the double layerering of risk that is the biggest problem, IMO. Until you figure out piece by piece the loss severities on the securities underlying the CDO, it's really hard to know what's going on. That's very interesting a lot of it is vintage 2005. I have a feeling understanding AIG would take enormous amounts of time, which has been my biggest problem. Ultimately I think you're right, but it's a hard fish to fry.
I think the things I want to know regarding AIG are:
1) Is the guaranty business potentially walled off from the rest of the company if they were to just abandon it?
2) If they were to write off those securities to zero, what's their adjusted tangible book value?
3) Assuming 1, what's the value of the rest of the individual components of the company?
4) Not assuming 1, what's the rest of their exposure?
Personally, I think those questions are the root of figuring out the Margin of Safety. I'd say, assume the absolute worst and then get an idea of what's it's worth. Email me offline if you want to discuss more: gregharris_73 .. at .. yahoo dot com. I really don't have enough time to root through as much as I'd want to by myself, but I'm certainly willing to put in some work through it.
-
Greg Harris
-
52 Comments
Aug 27 03:08 PMThe only problem I have with this article is I wonder if there needs to be more analysis on what the actual CDOs have underlying them. The real problem with the CDOs is that they're collateralized not by straight mortgages or assets, but tranches and pools of asset backed securities in the first place. A CDO may be an overcollateralization of tranches of securities that are all about to get wiped out, because the securities that back the CDO are early in line for default. If you're collateralizing the CDO with tranches of junk, the overcollateralized super senior portion of the CDO get's wiped out too, because overcollateralization of zero is still zero, or close enough. It's the senior portions of the original AB security that will prove to be collateralized better, not the CDO and it's the double layerering of risk that is the biggest problem, IMO. Until you figure out piece by piece the loss severities on the securities underlying the CDO, it's really hard to know what's going on. That's very interesting a lot of it is vintage 2005. I have a feeling understanding AIG would take enormous amounts of time, which has been my biggest problem. Ultimately I think you're right, but it's a hard fish to fry.
I think the things I want to know regarding AIG are:
1) Is the guaranty business potentially walled off from the rest of the company if they were to just abandon it?
2) If they were to write off those securities to zero, what's their adjusted tangible book value?
3) Assuming 1, what's the value of the rest of the individual components of the company?
4) Not assuming 1, what's the rest of their exposure?
Personally, I think those questions are the root of figuring out the Margin of Safety. I'd say, assume the absolute worst and then get an idea of what's it's worth. Email me offline if you want to discuss more: gregharris_73 .. at .. yahoo dot com. I really don't have enough time to root through as much as I'd want to by myself, but I'm certainly willing to put in some work through it.
-
brennpa
-
1 Comment
Aug 27 06:49 PMone troubling thing is that if you look at AIG's investment portfolio, it holds a fair amount of 2006 and some 2007 vintage CDOs. i think losses will be higher on this paper even if they hold to maturity. it also seems to me that the presence of this paper suggests that the right hand is not speaking with the left - if the CDS team saw such problems with 2006 and 2007 mortgages, why was this information not relayed to the investment portfolio team? Hank may have been smart enough to run this empire but it seems to me that the inconsistent holdings (don't write 2006 CDS but hold 2006 CDO's) combined with the substantial overexposure to the mortgage business (what percentage of AIG’s total capital based was exposed to the US mortgage market? Why was this allowed?) suggests that this thing is too unwieldy for one person to manage – similar situation over at Citigroup (which also holds a pile of high grade CDOs marked down to $0.60 which has yet to suffer 1 dollar of cash flow impairment).
One other thing – I think mark-to-market clearly has problems but this is rule. For that reason, it is not impossible that a further move downward in ABX causes further mark-to-market hits, leading to rating downgrade, additional collateral, and then further equity raise. Stock is so cheap it won’t matter if you hold the Company for several years but if you’re managing money for clients, it is much tougher as you could conceivably endure 20% hit for another follow-on offering!…my two cents
-
Tom Armistead
-
211 Comments
My Website
Aug 27 07:45 PMAIG sounded a lot like ABK when they started talking about why they increased the stress case losses on the insured CDO portfolio from a range of 1.2 -2.4 billion to a range of 5 - 8.5 billion. Apparently they didn't understand the structural considerations created by downgrades of the collateral. They do now.
I worry about the rating agencies mandating a capital raise under adverse conditions, so I am keeping the AIG position small until I can get some clarity on the issue.
-
WEBISKING
-
173 Comments
My Website
Aug 27 08:58 PM-
beatindex
-
1 Comment
My Website
Aug 28 12:55 AM1. Our financial system is quaking from the huge deleveraging happening because of the huge amounts of derivative securities tied to the mortgages/real estate industries as they themselves unwind. As this unfolds, all those gamblers who are on the hook will be penalized severely, and might disappear, because of the power of leverage, which cuts both ways with a 10x or 20x effect.
2. AIG does not simply hold long CDOs or CDS. They were 'insurers' and sold short CDS ie were the insuring party on a lot of credit default swaps tied to mortgages/real estate. So, their potential losses from this are unlimited and leveraged.
3. So, as hedge funds or whoever got their CDS, CDO insured by AIG gains on a leveraged basis due to house prices declining, AIG, MBIA etc will bear the losses on the other side of that zero-sum game. A 10% decline in house prices could potentially reflect itself in a 100% decline in shareholder equity amounts on which AIG, MBIA etc insured those 'credit derivatives' (CDO, CDS, ABS, CLO...)
4. Therefore, potentially these stocks have negative or zero future shareholder equity. If so, how can anyone place a multiple on the fair book value on an expected basis. Remember, trailing book value will look huge and you must not value these financials now on trailing P/B basis, but only on an expected P/B basis.
-
fxtrader07
-
618 Comments
Aug 29 05:42 AM-
JasonC
-
367 Comments
Aug 29 04:00 PMGo to their August 10Q, go to page 101, see the table "contractual obligations". There are $1 trillion in total. Over $700 billion on one line, their insurance and investment contracts. Check the times and you will see those stretch out over 30 years into the future at the average run rate.
When credit gets worse, interest rate spreads rise, and this creates mark-to-market losses in current value accounting. But as every bond investor knows, rates moving higher reduces the actual value of the position only up to the duration point in the future. At the duration point, it has no impact. And beyond it, the sign of the impact reverses.
Well, liabilities also change in present value when interest rates change. Significantly higher rates will dramatically lower the present value of a liability stream that is $700 billion in size and spread over 30 years into the future.
Can AIG "mark" its liabilities to reflect the fact that a much smaller investment in bonds at today's huge spreads, will fully cover those future obligations, than it would take to meet them a year ago?
One entry and one side of the sheet accounting is occurring.
It is only justified if the *entire* increase in spreads since last summer shows up as higher loan losses, dollar for dollar. They won't. Instead, higher spreads will cover the higher loss rates and then make up much of the losses so far, as well.
I leave aside how ridiculous it is to worry about the liquidity of a company with $700 billion in float for 30 years, and $100 billion a year in operating cash flow.
The present value of AIGs bondholdings has indeed fallen, because you can go buy bonds or preferreds paying 8% to 10% right now, not the 6% of last year.
But being a huge bond-owner with cheaply borrowed money (much of it essentially free, if the underwriting is done remotely correctly) is a much better business to be in when bonds yield 8-10%, than when they yield 6%, let alone the 4% of a few years ago.
-
JasonC
-
367 Comments
Aug 29 04:34 PM-
WEBISKING
-
173 Comments
My Website
Aug 30 04:56 AM-
Buprestid
-
16 Comments
Sep 07 12:28 PM-
E Nuff Sed
-
146 Comments
Sep 09 01:58 PM-
Orca
-
30 Comments
Sep 11 09:38 AM-
Buprestid
-
16 Comments
Oct 17 10:29 PM