The comment by Jay Ryan is especially indicitive of the point I want to make. There is a big difference between doing what is legal and doing what is right. Just because a course of action (or inaction) is legal does not necissarilt make it right. Since when has the accounting profession had the luxury of hiding behind simply doing what is legally expected of them? There are moral choices to be made in every undertaking and it is (or so I have been led to believe) that it is the auditors’ role to expose the “less-than-moral” actions of corporate chieftans. If the accountants won’t or can’t do that, who will?

I work in the industry and think it’s important to note that the auditors operated under existing rules and GAAP – Generally Accepted Accounting Principles.

And according to Compliance Week, “companies were still following Financial Accounting Standard No. 140, Accounting for Transfers and Servicing of Financial Assets, to decide when an asset transfer qualified as a sale (that could be kept off the balance sheet) or when it had to be treated as a financing (that remained on)….That rule has since been codified into the Accounting Standards Codification under ASC 860-10.”

In short, they were acting on the laws and practices that were in place at the time.

This is where the problem would be: Both companies are using the same auditor. And the auditor needs to be objective and independent.

You are saying that because the market was failing to “determine a price”, PwC should attest to both AIG’s model and to GS’s without any additional commentary.

So management’s assertion about valuation is one thing that would be producing wildly different answers (VEPCO). I think that problem also, indirectly,affects presentation, completeness and rights & obligations. PwC would at least have to make clear to AIG financials users that there is a dispute with GS about valuation, and to GS financials users that their pricing of the CDS is based on a specific assumptions about the contract with AIG.

Francine is saying there is a potential conflict of interest if PwC remains auditor of both. I don’t think it has anything to do with the amount of time allotted for an audit—because otherwise an auditor could always use time constraints as a defense to a bad audit. If PwC couldn’t understand the math, again, that’s not an excuse. They shouldn’t have taken the assignment.

esa, you’re right that there was no easy resolution of price for that financial “product” slash bet. But how does that absolve PwC? They are supposed to take management’s assertion about valuation and test it. If it is based on assumptions, presumably you have to disclose these assumptions for the users of the financials.

It’s very clear that AIG was not disclosing the downside to the agreements it was making about CDSs. They didn’t disclose—even experts in the regulatory community were surprised about some of the deals AIG entered into. PwC had a hand in that, because they gave over an auditing opinion without requiring much in the way of disclosure about the more exotic products.

I agree with you that in a fire-sale environment, it is hard to find an objective price—one that adequately values future expected cash flows by some reasonable standard. Everyone wants to hold cash and will make great sacrifices on price just to have cash in their hands. Or they are (more likely) subject to margin calls from their lenders. Or they are (similarly) forced to unwind their hedge funds because of redemptions.
The only price available is the one the last panicked seller cashed in for. And with financial products, these prices are both highly unreliable and destabilizing. As we saw in late 2008. And October 19, 1987 for that matter, when the stock exchanges stopped quoting trading prices because there was no “bid”

What I don’t see is how PwC gets to say, “well since the price can really be any number in this environment, we’ll just accept management’s number without comment.” Especially when it is auditing both sides of a transaction.

Are you comfortable with that? Does PwC really feel “objective” with that stance? Would you want to rely on financials made up of assertions like that, without at least footnoted disclosure?

Permit me to illustrate my perception of what happened between GS and AIG. At the end of my analysis, you can either tell me which facts I have wrong or tell me how this matter was supposed to be resolved in the summer of 2008.

Imagine for a moment that AIG sold GS a CDS on a $50 million face value pool of mortgage backed securities in June, 2007. The MBS has 30 years to maturity remaining, yield 5.5% and produce monthly payments of $283,900.

AIG sold the CDS for 50 basis points a year. In other words, GS will pay AIG $250,000 a year. The PV of the CDS premium payments that GS must pay AIG is $3,633,000.

In June 2008, several things happen. First it becomes apparent that the default rate will on the MBS pool be an unprecedented 8%. Due to the higher default rate, the projected monthly payments are expected to be $261,188 instead of $283,900. Second, the higher default rate has impacted the risk adjusted discount rate which is now 6.5%. If you recalculate the PV of the cash flows on the MBS you see that it has declined substantially from $50 million to $40,860,892. Of that amount $674,000 is due to the fact that Goldman collected 12 months of payments. The remaining decline in the present value of the future cash flows of the MBS of $8,466,518 is due to the higher default rate and the higher risk adjusted rate of return.

At this point, the MBS has suffered a 17% loss and the CDS has suffered a 233% loss. AIG’s projected payments on the CDS is $8,466,518.

So Goldman calls AIG in August and to ask them to post collateral on the CDS. GS wants AIG to post $14.8 million of collateral. Here’s GS’s argument. Mortgage backed securities are trading (when they trade) for 60% of face. At 60% of face, GS calculates that AIG’s liability on the CDS is currently $19,730,000 and 75% of that amount is $14.8 million.

AIG laughs at them. The 8% default rate hasn’t even happened yet. It’s just projected defaults. Plus, a big portion of the loss of $8,466,518 AIG calculated is due to interest rate changes. AIG isn’t insuring GS for interest rate changes. AIG is just insuring against defaults. The biggest part of what GS wants to be paid for though is the decline in the moribound MBS market. Again, AIG never agreed to insure GS for that.

So if you are AIG, what do you pay out? As little as possible of course because this is a once in a lifetime market and it will turn around in 2009. If fact that is true. 2009 was an incredible year for MBS and bonds because they were trading at ridiculously low prices in 2008. No matter how you calculate the value, AIG would have had an enormous gain in 2009 so they were right to hold out and stick with their values.

If you are GS, what do you hold out for? Keep in mind this is a 30 year contract with 29 years left on it. Like AIG you know that MBS aren’t going to trade at 60% of face value for long. This market will never come along again so you have to get out of this trade as soon as possible while it is worth the maximum amount. So you demand that $15 million of collateral to soften AIG up. Then you make the pitch. Forget the $15 million of collateral. If AIG will just pay Goldman $12 million right now, Goldman will cancel the entire contract. Not only that, but if AIG will pay a similar amount, GS will take any other CDS that AIG wants to get out of. Everyone knows the CDS are choking AIG and there are no buyers. So if AIG will just pony up enough money, GS will take away all their misery and their credit default swaps.

Goldman wasn’t really interested in collateral because while they may hold collateral, they don’t own it. They may have to give it back when the market inevitably turns. Goldman made money buying CDS from AIG as long as they could. When the credit markets completely collapsed Goldman knew that there was no more money to be made as a holder of a CDS. The company that was positioned to make money starting in the fall of 2008 was AIG. The only problem was that AIG didn’t have enough capital to stay in the game. Goldman had plenty of capital so if they could get AIG to pay them to take the CDS’s off their hands when the derivative liability was the greatest, GS could make a killing when the market bounced back.

So help me out. It appears to me that AIG was using cash flow models because it knew that long-term the market would snap back to what the models were telling them. It also appears to me that GS was using the market prices for bonds and CDS because it was the strongest negotiating tool at its disposal. Both parties were driven by greed and survival to manage the settlement of a contract in an illiquid market and either maximize their return or minimize the damage they suffered.

Obviously, I’m misunderstanding something. Was there a liquid market for MBS and CDS in the summer of 2008? Were MBS and CDS trading at amounts that were in line with the underlying cash flows? My understanding is that the answer to both questions is no and both AIG and GS took the position that suited them on a derivative that had decades left to run.

As for how the auditors were supposed to resolve that issue, I see no possible explanation for that expectation.

Is it not enough that PwC was clearly torn between two clients (and maybe more who would have been impacted) who held enormous financial sway and lost its independence and objectivity? I think PwC finally succumbed to Goldman Sachs, selling out AIG while still tippy-toeing around the necessity to finally say which one was closest to complying with standards. Actually taking a consistent stand would potentially implicate other clients such as JP Morgan and Bank of America as well as Freddie Mac in a mark-to-model or rather “mark to make it happen” scandal?

Her view is that PwC inevitably sided with the analytics as run by Goldman Sachs, which would have given the higher annual yield and the lower bond price — 9.34% in my example.

I think it’s very interesting that PwC was in a situation where they would have to attest to a yield-to-maturity valuation in one circumstance, and a mark-to-market valuation in another. You are correct that both of these could be acceptable GAAP. But I wonder what would be missing from the one or the other in the “VEPCO” framework:

Valuation
Existence
Presentation (and disclosure)
Completeness
Obligations and rights

The point is that PwC would have to attest to management’s assertions for VEPCO of the one client, and then go to the counterparty, and attest the same. But they can’t disclosure that either counterparty values the same security differently, or that the one counterparty has a claim for more collateral from the first, and that the first has a potential liability to the second.

If PwC’s auditing puts one of the two in shaky circumstances PwC loses the business.

So probably PwC was not independent at some point, as to their relationship with GS and AIG. They need not have brought the parties together to ascertain an acceptable price. Only that their sign-off on financials could have been influenced by the consideration that they would lose a multi-million dollar revenue stream if they raised doubts about one or the other’s valuation of the controversial CDS asset.

This is why I say the Big Four are a cartel. It is harder to be independent–or objective—when your client base contains so many interlocking financial companies. The collapse of one could affect the financial health of others of your clients.

I disagree with the post on the notion that mark-to-model is inappropriate, if the model is a discounted cash flow of expected revenue streams. One of the tragedies of the financial crisis was this whole veneration of “mark-to-market” as some kind of earthly deity. In a market panic (mass selling of securities), the market is not rationally or objectively pricing the paper that is trading. They are simply looking for a price to dump it. So these prices are hardly “objective” and the discount rates are hardly credible. I am very much in favor of a mark-to-model if the model is credible. Hold-to-maturity transactions would fit that model.

Francine McKenna is bringing up very important concerns that few people seem to be thinking about. Would we be better off in a Big Twelve auditing environment, instead of Big Four? If there were multiple auditing firms–or a rotation—it’s less likely that one auditing firm would be auditing two companies with the same, material, transactions between them. That should be an audit concern.

Breaking up the Big Four into threes, for example, would hurt no one except for those at the very tip-top of the corporate pyramids in each. These people have already abdicated their responsibility to the general public in exchange for partnership multi-millions, and so no one should cry a tear for them.

You and I are in complete agreement. There are an infinite number of ways to value what you refer to as the ESA bond and the related credit swap. During 2008 and most of 2009, the range of possible values was all over the place because the market and the valuation inputs where a bit like the instruments on an aircraft in a 1960’s movie about the Bermuda Triangle; a fog enveloped the plane and every needle on every guage just started spinning. My example was intended to illustrate that the issue is not cut and dry. There were many answers to the question of “How much is the CDS worth?”It wasn’t something that PwC could have been expected to resolve. Furthermore, the code of professional conduct forbids them from violating client confidentiality and sharing such sensitive information between clients so they weren’t in a position to force a resolution as Francine suggests.

What was their obligation? To audit the estimate that AIG recorded as a liability in accordance with the professional standards. Do those standards require them to get the counterparties sign-off on the estimate? No. Is the auditor culpable if estimate is wrong? No. Do auditing standards and accounting standards require that the two audit clients record the same amount? No. Do auditing standards indicate that if such an unresolved matter exists between two companies, the auditor has a scope limitation? No.

These situations exist on many audits and Francine’s expectation for how this one and all the others should be resolved is just that: her expectation. It has no basis in GAAP and no basis in GAAS. I know she disagrees but her view simply isn’t reality. A matter such as this can’t be resolved with a few phone calls between an auditor and his clients during the 45 day period public companies have to issue financial statements. It would take a judge, jury and lawyers several years to resolve this dispute. Why she thinks auditors can do it in the span of six weeks is beyond my understanding.

Let’s not forget that AIG Financial Products stopped entering new trades 18 months ago and they are still employing hundreds of traders just to unwind those trades. They will continue doing that for a couple of years. Francine thinks those matters should have all been settled by the auditor and that traders aren’t necessary. That’s a bit like saying that while the trial was going on Atticus Finch should have also integrated the schools. Or maybe Lindbergh shouldn’t have landed in France but should have instead crossed the Atlantic and then flown straight to the moon. Those are all great suggestions but for the fact that they are (a) impossible and (b) not their job.