I have had a number of posts on mark-to-market accounting in the recent past. Most recently, the posts have suggested that accounting is going to be favourable to banks and their quest to present a well-capitalized face to the world (see the post “JPMorgan’s $29 Billion windfall”).
A reader who deals with accounting issues has seen fit to send me an in-depth reply to a recent article on the issue, “What the stress tests reveal about Obama’s thinking on banks.” Below is his excellent commentary in full. This is a very technical analysis, so if math and accounting is not your thing…
I unfortunately disagree with your assertion of my opinion, but I do hope you will consider my positions on your thoughts, which I will try to outline from here, addressing them in the chronological order of your post. I do apologize for the length and slight rambling (I have attempted to break it into sections), but you did give me a chance to try and summarize my thoughts and think about the other various aspects which I had not previously considered, so thank you for that. I am warning you though, this is not short, it came out to almost 4,000 words. I hope it makes sense.
I understand what Kyle is saying: FAS 157 guidelines will have no impact on reported earnings.
Foremost, I want to be clear that this is not what I am attempting to say. I am saying that it will impact reported earnings for exactly the reason you state, but the overall financial situation of the entity does not change (subject to some hypotheticals presented below). My point is simply that the writedown is broken into two places, the earnings/income statement portion, and the OCI portion reflected in Stockholder’s Equity portion. Furthermore, this is not due to the 157-4 change, but due to FSP FAS 115-2 and FAS 124-2, Recognition and Presentation of Other-Than-Temporary Impairments, which was issued concurrently with FSP 157-4. So, if you’re looking at the Balance Sheet, the effect is null. Instead of Retained Earnings being lower by the amount of the change, OCI is lower by the amount of the change. The end result is lower equity by the same amount. The 157-4 revision does not lessen the total writedown in this case, it lessens the portion reflected in earnings, and increases the portion reflected in OCI. The net effect is the same.
I. HTM vs. AFS
A bank only has to attribute its actions to 157-4 if it is amending prior accounting to reflect a change in asset designation to ‘holding to maturity.’ However, if it marks assets today as ‘holding to maturity’ and then is later forced to write down those assets, these writedowns will not be attributed to changes in the FAS 157 guidelines.
I believe this statement is somewhat incorrect, but please allow me to expand on that.
First, the 157-4 change does not require a firm to acknowledge a change from AFS/T to HTM. That has always been a GAAP requirement, and would be reflected somewhere in the F/S, either as a footnote, or as a change in the Significant Accounting Changes portion of the Notes to F/S. The FSP 157-4 change primarily reclassifies when price quotes from inactive markets should be considered for fair value changes. That does not allow a firm to change the asset from AFS/T to HTM, it just means that an asset that was designated as Level 2 is now designated as Level 3 due to use of different inputs. This is a change in accounting estimate, and would be disclosed on any quarterly statement. (I will attempt to provide an example with WFC later on.)
It is absolutely true that if a firm decided to move said asset from AFS/T to HTM, any future write down would not be attributed to a FSP 157-4 change, but I am unclear as to your point in the text quoted above. If the security is moved to HTM as in the situation you presented, it is carried at the FV as of the change, and is then carried at amortized cost, so any market value price changes would not be reflected subsequent to the reclassification. A firm trying to utilize a change via 157-4 to write-up previously marked down assets would lose all ability to exercise that right if they moved the security to HTM, be locking in heavy losses, and I think have a very difficult time getting their auditors to go along with it in the first place.
Furthermore, if they wanted to kill off any fair value losses from this type of accounting treatment, I would think they would have done it in like 2007, before suffering all those previous writedowns, or at any point in time since then. Again, emphasizing this point, FSP 157-4 does not make moving securities from AFS/T to HTM any more likely or feasible from an accounting standpoint. As I said before, I suspect their auditors would not let them just off and dump their entire portfolios like that to begin with, for obvious reasons.
I think people (not necessarily you, I’m generalizing for the sake of generalizing) have a misconception that firms have these gigantic holdings of HTM securities in their portfolios, so moving stuff in and out of them at will is an easy way to hide stuff, and that is simply untrue. (Note: I picked these examples pretty much at random, but tried to get a big money center, a regional, and an I-Bank in there for diversification purposes.) Regions has $45 Million total in HTM Securities [LINK]. JPM has $31 Million total in HTM Securities[LINK]. State Street, on the other hand, has a non-trivial $15 Billion worth of HTM Securities (pre-consolidation, noted below), and it is primarily made up of ABS and CMOs [LINK].Kind of scary.
However, all three banks present (and are required to present) these HTM portfolios at amortized cost [their carrying value], and then at their comparable fair value in the notes. Accordingly, in the event that they had the ability to, and subsequently did, move any material amount from AFS to HTM to hide further writedowns from 157-4 adjustments to fair value measurements, aside from the fact that they would have to explicitly disclose the move, you would see an enormous jump in those balances to begin with and it would be immediately obvious to anyone looking at their F/S. Again, this would not be a result of 157-4 explicitly, it would have to come from another interpretation or a new FSP regarding this. As I stated earlier, if they intended to do this, they probably would have done it already. I tried to outline why I thought this would be unlikely from potential consolidations in the previous section.
On the other hand, an accounting change which did give further clarification or manoeuvrability in this sense [of moving to HTM from AFS] would have enormous material effects on recognized [but not disclosed, which could be affected by FSP 157-4, as outlined next] losses to be sure, and would probably render current balance sheet reporting all but entirely useless, in my opinion. The point I am trying to make is that the entire premise of any firm being able to hide changes due to 157-4 in a HTM reclassification rests on the ability to move the security from AFS/T to HTM, which has not occurred at all thus far, despite obviously having plenty of available motivation to do so. If that indeed does become a common practice, it will be due to another accounting change, because 157-4 does not address this issue in any respect, and as I just stated, would be an unbelievable fold on the part of FASB. I hope that made clear why I do not think firms will have the ability to hide losses related to FSP 157-4 by simply moving them to HTM from AFS.
II. Writedowns of Unconsolidated OBS Securities
Given that only one firm (that I am aware of) actually re-marked-up assets in applying FSP 157-4 during Q1, the only way it could give rise to further markups (or the ability to reclassify under FSP 157-4 at all and thus lessen future writedowns) would be if markets currently considered active became considered inactive, as this is the reason that WFC chose to re-mark their assets, and is the crux of FSP 157-4. It is unclear to me, however, to what degree this would result in a write-up though, and to what degree any number of markets that are currently considered active have the possibility of becoming inactive. I do not know of any specific accounting literature to specify one way or the other. To what degree does the security get revalued in the event of a market becoming inactive? If WFC is the standard setter for this treatment, it could definitely be bad news, but it’s hard to know the answer without knowing the assumptions versus observable market prices for various assets, how much those might diverge between now and the time of any change, and what exactly an auditor requires in order for a market to explicitly become inactive.
In my view, it is future credit card, jumbo loan and CRE exposure which will be most affected by this. These are areas where you should expect heavy pressure from securitized assets on bank balance sheets due to deterioration in income from credit card receivables,prime mortgage loans, and commercial real estate loans. What mark-to-market guidelines effectively mean is that banks will not have to reduce capital by nearly as much as had they not marked these assets as hold to maturity.
I definitely agree with you on this premise, Ed, aside from the HTM part. Per the previous section, I do not think firms moving currently held as AFS/T securities to HTM is a likely scenario, outside of another fundamental accounting change, and certainly not directly due to FSP 157-4. I would, however, expect that those areas which you just highlighted to start seeing very heavy pressure starting in Q2. You’re already seeing this effect, as per State Street’s decision to consolidate some conduit assets last week.
But, going back to my previous point about inactive markets, in the event that those markets became inactive, there is a chance that those securities could be marked down less than what they would have been subsequent to FSP 157-4. But in order for this to even be relevant from a HTM standpoint (AFS standpoint next), they would also have to consolidate those assets back onto their Balance Sheets first as per STT, which almost unilaterally would result in immediately realizing previously unrecognized losses, unless they revalued them upwards before consolidation. But again, they report the change in fair value on a quarterly basis, so you would know if they did indeed consolidate them at a higher value, and I think that is highly unlikely per the reasoning below.
This is just an opinion, but I find it difficult to believe that an auditor would allow a company to mark up an unconsolidated asset exclusively based on 157-4, then consolidate it, and avoid having to recognize any of the losses due to changes in fair value. At the very least, they would have to recognize some OTTI, which is likely a material amount. Most of the assets in those conduits are already valued on a level 3 basis to begin with, as I doubt they have very many observable inputs. The only example of this that I have is STT, which explicitly states in the 8-K that all of the assets they consolidated were carried on level 3 valuation basis, and that of course resulted in a realized after tax loss of $3.6 Billion. If the treatment STT gave to its conduits is the norm (the conduits are almost all abs of various sorts), then it appears unlikely that any firm would take the above mentioned approach of a write-up pre-consolidation.
Of course, STT only has to recognize the OTTI on those assets from now on since they just suffered the realized $3.6 Billion A-T loss on the conduit in order to bring it onto the B/S, but they weren’t recognizing any impairment prior to consolidation [stating the obvious, I know]. Those are toxic assets somewhat de-toxified, to the extent that you now know exactly what kind of losses were lurking Off-Balance Sheet, and what FV might still be unrecognized per the disclosures.
So I guess if you were being really “proactive” about it as an organization, you would re-consolidate now while the market is jubilant, take the current loss including fair value markdown, (which may be a lot less than it will be in a year, although that may already be more than substantial in some cases, and would absolutely require massive capital raising for most firms that decided to do that; see STT, which had to do a CS offering and a debt raising to keep them above water from the consolidation) then only recognize OTTI and avoid having to mark down any fair value changes in the future. No one else has done this thus far that I know of [aside from C last year, which I am sure you recall], and they would still have to disclose the difference between fair value and carrying value on those now consolidated-HTM securities. Again, if firms started consolidating left and right prior to those assets really nosediving (or prior to December, when they likely won’t have a choice due to FIN 46(R), where there is currently a debate as to whether to bring those on at FV or BV), it would be pretty obvious. In the event they have to bring those OBS vehicles on at fair value, we will probably have another legitimate meltdown, as I’m sure you’re well aware.
Anyway, if they did consolidate some OBS vehicles pre-November (at amortized cost and net of any OTTI or FV write down) and then just had billions of unrecognized losses sitting around that everyone knew about as per the disclosures in the notes, I can’t imagine they would be able to raise any capital. This is also primarily why they will probably be up a creek in November, when they have to bring those assets on book, either at fair value or net of OTTI (as I don’t see the above mentioned scenario happening). It would be the same effect as if they had been allowed to just throw everything into HTM 3 years ago, and then only had to recognize the OTTI since, but had to disclose the massive fair value losses in the notes. The B/S would be pristine, with a nasty disclose of “Fair Value losses on HTM portfolio of whatever billion dollar figure you would like to use.” The FSP 157-4 change could allow the FV presented on those disclosures to be different going forward, but it would really be irrelevant from a recording standpoint [but not a reporting standpoint]. It would be sneaky, and I’m unclear as to whether they have to or will have to disclose that information in the future. I suspect they will.
.III Effects on AFS Securities
From the other side of the table, and I apologize if this is what you were initially referring to, if a firm isn’t subject to consolidation of an obs vehicle, but instead is a holder of a securitized receivable carried as an AFS/T security, they could start revaluing them on a Level 3 basis as per the FSP 157-4. It would not hold true that they would be able to hide these potential losses, because, as previously stated, that would require a transfer from level 2 to level 3, which needs to be disclosed along with the relevant accounting interpretation for that treatment, a la 157-4, and highly unlikely that they would be allowed to just move it into HTM as per Part I of this email.
I admit there is a gray area here and room for chicanery to an unknown extent going forward [as I tried to note above]. Depending on the asset class backing the security, they may still rely on L2 inputs, so incorporating L3 inputs might reduce the writedown significantly, or even result in a writeup. Again, if this is the crux of your argument, I do not disagree, and I apologize for the long letter. Still, those are not losses they can necessarily hide, and I think its impossible to accurately know what kind of writedown to expect in one situation versus another as for the reasons noted above (lacking information about L3-related assumptions vs. the here-to-fore observable market prices). They could be huge, they could also be minimal, there’s really no way to know as per above. I’ll defer to your expertise on that, and if you’ve got specific projections for writedowns that you expect (as relates to your comment about the future losses to be “less than what you expect”) pre-157 change for CC/CMBS/etc., I would love to see them just as a gauge against what actually ends up happening.
I have not seen any documentation of any expected losses on those areas for the upcoming quarter on a firm-by-firm basis, just expected declines in underlying assets on an industry-wide basis from various ratings agencies. I am also skeptical of this happening to the degree that it has not been adopted and applied retroactively to previous writedowns. In any case though, you would know that they were trying to do reassess losses because of 157-4, as it would be reflected in the change from L2 to L3 valuations, and not be able to be hidden by moving from AFS to HTM, as previously tried to outline.
.IV Example of FSP 157-4 Effect (to Date)
I will use WFC (because they are the only example) to demonstrate what the 157-4 change does mean, how it has been used thus far, and to further provide evidence of why moving an asset to HTM from AFS/T is not related to FSP 157-4 and is purely a function of valuation change inside the FV hierarchy. Here is the copy/paste from the WFC 10-Q (bold emphasis mine):
FSP FAS 157-4 addresses measuring fair value under FAS 157 in situations where markets are inactive and transactions are not orderly. The FSP acknowledges that in these circumstances quoted prices may not be determinative of fair value. The FSP emphasizes, however, that even if there has been a significant decrease in the volume and level of activity for an asset or liability and regardless of the valuation technique(s) used, the objective of a fair value measurement has not changed. Prior to issuance of this FSP, FAS 157 had been interpreted by many companies, including Wells Fargo, to emphasize that fair value must be measured based on the most recently available quoted market prices, even for markets that have experienced a significant decline in the volume and level of activity relative to normal conditions and therefore could have increased frequency of transactions that are not orderly. Under the provisions of the FSP, price quotes for assets or liabilities in inactive markets may require adjustment due to uncertainty as to whether the underlying transactions are orderly. For inactive markets, we note there is little information, if any, to evaluate if individual transactions are orderly. Accordingly, we are required to estimate, based upon all available facts and circumstances, the degree to which orderly transactions are occurring. The FSP does not prescribe a specific method for adjusting transaction or quoted prices, however, it does provide guidance for determining how much weight to give transaction or quoted prices. Price quotes based upon transactions that are not orderly are not considered to be determinative of fair value and should be given little, if any, weight in measuring fair value. Price quotes based upon transactions that are orderly shall be considered in determining fair value and the weight given is based upon the facts and circumstances. If sufficient information is not available to determine if price quotes are based upon orderly transactions, less weight should be given to the price quote relative to other transactions that are known to be orderly.
The provisions of FSP FAS 157-4 are effective in second quarter 2009; however, as permitted under the pronouncement, we early adopted in first quarter 2009. Adoption of this pronouncement resulted in an increase in the valuation of securities available for sale of $4.5 billion ($2.8 billion after tax), which is included in other comprehensive income, and trading assets of $18 million, which is reflected in earnings.
As you can see, FSP 157-4 provides no relief to companies in their discretion of deciding whether or not to carry any asset at fair value or HTM, it only gives them discretion at deciding what prices to use in determining FV. The overall lack of revision industry-wide suggests that WFC’s claim that “many companies” were using the lowest common denominator as fair value is likely overstated in my opinion. Now, you could certainly suggest that firms are lying about their self-admitted belief that they do not anticipate the change to have any future effects (as is presented in the notes under the statement of significant accounting principles); but in the event that it did affect the aforementioned currently active and soon to be potentially inactive markets, they would still be required to disclose any change in estimate, which is what this is, as this is a requirement of GAAP. It is located under the “Significant Changes in Accounting Estimates” portion of the Notes.
Here is the table attached to the previously quoted portion of the WFC F/S:
The following table provides the detail of the first quarter 2009 $4.5 billion (pre tax) increase in fair value of securities available for sale under FSP FAS 157-4.
Residential $ 2,311 Commercial 1,329 Collateralized debt obligations 492 Other (1) 394
(1) Primarily consists of home equity asset-backed securities and credit card-backed securities.
That example was supposed to show what the effects of FSP 157-4 have been, and what any changes going forward would resemble, from a reporting standpoint. That resulted in a change in the OCI portion of the Stockholders’ Equity statement. It is not a change from AFS to HTM.
.V Stress Tests and Summary
I do not disagree with anything you said regarding the stress tests; I think you are absolutely correct. I do think that TCE will be effected by FSP 157-4 to the extent that a firm can move an asset to L3 and recognize less of a writedown; I do not think they will be able to move it to HTM and hide the fact that any potential decreased writedowns from FSP 157-4. I think the biggest issue financial firms face is the upcoming consolidation rules of OBS vehicles, as well as the writedowns of CMBS/CC/Prime securities. I think the potential effects of FSP 157-4 on those securities is inconclusive at this point, but I do think there is room for potential abuse as per the above examples. Essentially, I think most firms (despite the claims of WFC above) have already been discounting the observable inputs affected by the change of FSP 157-4, which is why you have seen little to no restatement of prior writedowns. If that is indeed the case, I see no reason why that same accounting logic has not already been applied to the other potential areas mentioned, thus my feeling that the actual language issued in FSP 157-4 is null.
Again, it’s not that I don’t think FV accounting matters, because it definitely does. I just don’t think FSP 157-4 specifically matters. I apologize for the length, but I hope it makes sense.