Applying U.S. Central's Loss Across the Retail Corporates

What better way to mark my 100th post than with an update to my favorite table ever, the one that shows the expected impairments of MCS and PIC across the corporate network.

This table needs updating by virtue of U.S. Central's announcement of a $308 million loss for Q3, the result of an additional $320 million in writedowns on its ailing book of private-label mortgage securities. Here are the results. As always click on the image to enlarge.



(Note that I've eliminated the columns where I showed capital ratios for the various institutions. It didn't add a lot of value, and it was tedious to calculate given the ratios' reliance on daily average net asset values.)

The highlights:

There used to be seven corporates with reserves. Now there are just five. This latest round of U.S. Central losses will require Eascorp and CenCorp to begin depleting members' capital. None of the five with retained earnings (Corporate One, Corporate Central, Corporate America, VolCorp and Iowa) would see their reserves depleted even if U.S. Central ends up completely impairing its MCS.

Paid-in capital at Southeast, SunCorp, First Carolina, First Corp, Kentucky and Kansas is now fully depleted because of U.S. Central's Q3 losses. These six corporates' PIC was partially depleted as of the end of Q2, but it's all gone now.

The three retail corporates most at risk for insolvency remain, in order (and in my opinion) Constitution, Members United and Southwest.

Constitution has $13.3 million in capital, $2 million in remaining capital at risk at U.S. Central, and its next round of OTTI will likely produce a loss of roughly $10 million. I have a hard time imagining this ending happily for Constitution.

Members United has $233 million in capital with $23 million in exposure to U.S. Central. Its next OTTI should be much less than $200 million (my guess is somewhere between $50 and $75 million), but with a few more quarters of similar investment impairments, insolvency would be on the horizon.

The same is true for Southwest, which has $200 million in capital and $19 million in remaining exposure to U.S. Central. Its next round of OTTI will likely also be a quarter to a third of its remaining capital, which means--like Members United--Southwest has less than a year of solvency at the current burn rate.

WesCorp's Excellent Adventure in the Capital Market

Given that I post about every two weeks these days, it's probably not difficult to recall the theme of my last effort. Namely, that U.S. Central's debt issuance illustrated the uncompetitive rates being paid to natural person credit union's on their NCUA-guaranteed corporate certificates.

I suggested--crudely--based on a survey of corporate CD offerings, that corporates were underpaying their members by about 10 bps.

It turns out this was only partially accurate. Based on today's sale of $1.5 billion in three-year notes by WesCorp, it appears that corporates are in fact paying a decent market rate on their certificates but that U.S. Central got taken to the cleaners by its new bondholders, to the tune of a 10 basis point gift.

How so? First, it's not quite accurate to compare the yield on the U.S. Central three-year note (1.92%) to the yield on today's WesCorp offering (1.79%) and say that U.S. Central overpaid by 13 bps. The underlying level of interest rates changed over the last two weeks. Instead, the appropriate basis of comparison is to look at the spreads of the two offerings, either to the Treasury curve or to the swap curve.

The U.S. Central three-year note was priced at +47 bps to Treasuries and +5 bps to swaps. The WesCorp note today priced at +35.9 bps to Treasuries and, supposedly, at -5 bps to swaps, although it appears that it priced at a lower yield than this, like swaps less 6 or 7 bps. The exact spread isn't especially important--we have enough information to conclude that the WesCorp members got a deal some 10 to 12 bps better than the U.S. Central members did. This is pretty clear evidence that the U.S. Central offering of two weeks ago was underpriced (i.e. offered too much yield) by almost the same amount I contended the corporates were underpaying their members.

Why does any of this matter? Well, I will concede that the most compelling argument when it comes to defending corporates paying low rates on CD's is the idea of a 'zero-sum game.' The fundamental notion behind this argument is that natural person credit unions are paying the entire bill for this debacle anyway. If corporates pay high CD rates, the final price tag for the stabilization will be higher (since the corporates are less capable of earning their way out of the hole). Similarly, by paying low rates, the corporates might be able to make extra income which lowers the stabilization expense. The total cost to the industry is fixed irrespective of CD rates.

This is a solid argument, and the only real complaint I have with it is that the NCUA should offer an incentive other than fear to get natural person credit unions to invest in the corporates. If that means paying higher CD rates, then at least the NPCU's who funded the corporates through this turmoil come out relatively better off than those NPCU's who have forsaken the corporates. Paying low CD rates implicitly favors those who invest outside of the corporates. The extra yield these credit unions earn on their investments more than offsets the fractionally higher stabilization expense they'll face.

However, the 'zero-sum game' argument breaks down terribly when, in fact, corporates are paying their members a fair rate while U.S. Central is paying its external investors abnormally high yields. Now it's the worst of both worlds. Not only is the final stabilization bill higher (due to U.S. Central's higher funding costs relative to CD's), but the extra investment income is accruing to entities outside of the credit union industry.

(One other common argument made on this topic is that it is inappropriate to compare U.S. Central note rates to retail corporate CD rates. A better comparison would be to know the wholesale CD rate, i.e. what a retail corporate earns when it buys a CD from U.S. Central. This is true, but it also illustrates an important point that this spread between the the wholesale and retail corporate CD rates is a deadweight loss. If we rid ourselves of the two-tiered corporate market, this extra yield would go where it belongs, to the natural person credit unions.

==================================================

One last thought: why did the WesCorp offering come at a 10-12 bps premium (lower yield) to the U.S. Central offering of two weeks ago? The most logical argument is that investors were now familiar with the NCUA-guaranteed bonds. Where they might have shied away from the U.S. Central bonds (indeed, the U.S. Central 3-year note reportedly had mediocre interest while the WesCorp offering was 2.5 times oversubscribed), investors now understand the TCCULGP and are willing to accept a fairer yield.

But what if the NCUA got the message that the sweetheart yields available to investors outside of the credit union industry were an outrage? Could NCUA have dictated that this bond had to be priced at a spread more comparable to where guaranteed corporate CD's are offered? Probably not, but it's good grist for the conspiracy theorists' mills.

U.S. Central Issues Debt, Extent of Corporates' Lowball CD Rates Revealed

First, let me say that it's absolutely shameful that neither U.S. Central nor the NCUA would publicize the issuance of $4 billion in debt by an institution that we all own. (Alright, we don't completely own U.S. Central yet since it still has a tiny amount of MCS, but give it a couple more quarters and it'll officially be a ward of the industry....) I learned about the bond issuance from one of my readers who learned about it from Bank of America-Merrill Lynch. When news about the corporate credit union network is coming from the investment banks, you know we've entered bizarro world.

Moreover, what little communication that did come from U.S. Central and NCUA (a letter from U.S. Central CEO Francois Henriquez to corporate CEO's and a note from the NCUA to the trade publications) offered a rather cryptic explanation of the yields on the newly-issued debt. Specifically, these letters said that U.S. Central issued $500 million of a two-year floating rate note at 3-month LIBOR flat, $1.5 billion of a 2-year fixed rate note at swaps + 0 bps, and $2 billion of a 3-year fixed rate note a swaps + 5 bps. Clear as mud, right?

Let's dig a little deeper into the two fixed rate issues. (Thanks to the reader that sent the Bloomberg info my way...)

USCENT 1.25 10-19-11 was issued at an original price of 99.949, which gives a yield of 1.276%. This yield is equivalent to the two-year swap rate, so the NCUA / U.S. Central explanation makes sense. U.S. Central sold $1.5 billion of this issue, which was fully backed by the NCUSIF, and I hear that this issue was quite popular and oversubscribed. U.S. Central must pay the NCUA an additional 10 bps for guaranteeing this note.

The second fixed-rate note was USCENT 1.90 10-19-12, which settled at a price of 99.936 or a yield of 1.922%. This was indeed equivalent to the three-year swap rate plus 5 basis points. This was a $2 billion issue that apparently wasn't quite as oversubscribed as the 2-year note but still sold fairly well. U.S. Central must pay NCUA 15 bps for the guarantee on this one.

Want to know why U.S. Central and NCUA quote the yields on these things versus swaps instead of coming right out and telling us what they yielded? Because they offered significantly better yields to investors than what corporates are paying to natural person credit unions on guaranteed CD's.

The investors that bought the two-year note (presumably banks and/or big money managers--I haven't heard from any credit union that they had the opportunity to buy these notes as new issues) got a yield of 1.276%. Here's a sampling of what some of the larger corporates are paying on 2-year CD's (which have the same credit rating, but poorer secondary liquidity so theoretically should have a slightly higher yield than the notes).

• WesCorp - 1.26%
• Members United - 1.23%
• SWCorp - 1.23%
• First Carolina - 1.19%
• Mid-Atlantic - 1.15%
• SE Corp - 1.13%
• CenCorp - 1.13%

Let me say it one more time: the natural person credit unions, who have done a fantastic job of supplying liquidity to the corporate network when it would have been easy to walk away, are earning substantially LESS yield on their corporate investments than banks and money managers who came to the party this week. I hope you find this as outrageous as I do.

(I'll analyze some of the positive benefits of having non-credit union funding in the system over the weekend. I'm not opposed to the idea of inducing external money into the system, but I am opposed to corporate credit unions taking advantage of their members.)

I'll close tonight with an anecdote from the San Diego NCUA town hall (which is available online here). OCCU director Scott Hunt told a story about credit unions who have come to him and said that (I paraphrase) we can earn extra yield by investing outside the corporate network, to which Hunt replies, yes you can, but recognize the higher cost that the system faces if NCUA is forced to liquidate troubled investments due to a lack of systemic liquidity. Hunt makes it sound like there are two choices, either accept what the corporates are paying you, or withdraw your money and prepare to meet your maker via a systemwide collapse. He doesn't mention a third way: ask the corporates to pay you a fair rate for your investment.

A Town Hall Retrospective

Now that the three NCUA townhalls are in the books, it's worth reviewing the highlights of the meetings. It appears the most fireworks happened at the San Diego meeting last Monday, where, according to reports from the CU Times (links here, here, here and here) and readers' feedback, four interesting developments occurred:

1. The proposed NCUA regulations on reforming the corporates (formally, Part 704 of the Rules and Regs) will be released on November 19th. The proposal may require the corporates to ask members for paid-in capital before providing services to members. Moreover, the proposed regs may seriously handcuff the corporates' portfolio managers from adding as much interest rate risk to their books. The sources of excess spread for a corporate investment portfolio are (generally) interest rate risk (including buying securities with embedded options), liquidity risk (buying securities with less actively traded secondary markets) and credit risk. If anything, the one risk you'd like to leave them is interest rate risk. It's the easiest to manage (CapCorp notwithstanding--we've all gotten a lot smarter since then) with derivative overlays, and you have to give the corporates some ability to earn excess spread. If they can't earn reasonable yields on their investments, they won't be able to offer competitive deposit rates, and they'll be forced to increase fees on other services.

2. NCUA predicts the failures of large credit unions in 2010. This is not surprising in itself, but it's interesting that the NCUA would discuss this in a public forum. The failures in 2009 have generally been of small to midsized institutions, but failures are increasingly likely to affect larger institutions in 2010. Chairman Debbie Matz indicated a strong positive relationship between failures and exposure to business and indirect lending. (I haven't written about too many non-corporate issues on this blog, but I feel the trade groups' push for relaxed business lending caps will have the unintended consequence of giving credit unions more rope with which to hang themselves. Yes, some credit unions can profitably and responsibly make business loans. Many--most?--others can't, and charge-offs can pile up a lot faster in a book of business loans than with typical CU lending activity. But I digress....)

3. Scott Hunt, the director of the NCUA's Office of Corporate Credit Unions, told San Diego attendees that WesCorp will almost certainly record large additional other-than-temporary impairments (OTTI) when they close the books on Q3. This isn't surprising--the collateral backing WesCorp investments grows more troubled every day and recent months have seen the realization that loss severities on option-ARM loans will be staggering. According to CU Times, Hunt characterized the chances of WesCorp's investments performing better than currently estimated as "highly unlikely," I'm not sure if this is simply Hunt's opinion or if this notion is backed by other external analysis, but it certainly reflects poorly on the Clayton valuation process. If the investments were properly valued and impaired, the likelihood of over- or underperforming credit loss estimates would be about 50/50. Hunt apparently believes that the Clayton estimates are way too optimistic.

4. Finally, Hunt seemed to slip up and indicate that some ring-fencing of WesCorp's troubled assets was in the works. One attendee told me that Hunt mentioned WesCorp issuing NCUSIF-backed bonds, and this was followed by Hunt refusing to provide any detail. Another person present at the meeting said Hunt mentioned splitting the bad assets from the corporates, perhaps sometime in Spring 2010. I am a big fan of the "good corporate / bad corporate" model, assuming you can find investors willing to finance the bad assets. And I think it makes a tremendous amount of sense to have WesCorp be the Bad Corporate. They are a basket case and so hopelessly undercapitalized that no one in his right mind believes them to be a going concern. If you are able to add a lot of debt financing to WesCorp's balance sheet (even if the NCUA has to guarantee it), you can strip WesCorp of depository authority and still not be in a position where you have to sell the bad assets. (Think of the proceeds of the external debt issuance being used to redeem member deposits at WesCorp, and then closing the book on WesCorp as a depository institution.)

There are still a few problems with this plan, including how you transfer the bad assets at U.S. Central and others to WesCorp without realizing the unrealized losses. And the demise of WesCorp would create a huge vacuum for natural person credit unions on the west coast. This second issue is probably easier to solve than the former, but if U.S. Central also goes away (or also becomes a Bad Corporate) as part of the corporate restructuring, then even the issue of transferring securities between institutions isn't a huge deal.

The letter They don't want you to read...

Did you notice that U.S. Central has published a revised annual report where the letter from Jim Nance was simply removed? Didn't the Advisory Board of Directors sign off on the annual report before its release? Was what Nance wrote really so shocking that it had to be totally stricken from the record? If anything, I have a more sympathetic view of Nance simply because the NCUA saw fit to expunge his words.

Incidentally, the more I hear about Nance's exit, the more I think this was purely a personality issue. If his dismissal truly amounted to strategy differences, the likely conflict was that Nance was much more interested in selling securities out of U.S. Central's portfolio than were the on-site NCUA guys. Yes, it's plausible that this difference escalated to a fireable offense, but it's much easier to see a situation where the guys running the firm simply didn't get along.

Anyway, for those of us not down with U.S. Central's revisionist history, here is the text of Nance's original letter (click on each image to enlarge):

Sayonara, Nance

I don't really have anything to add on the topic of James Nance's firing--I've pretty much heard the exact same thing that the trades are reporting. But I did want to open up a new thread for any comments and announce a temporary detente in my war on journalism!

I will add one thing: regardless of whether it's Jim Nance, Francois Henriquez or a suit-to-be-named-later sitting in the big office at U.S. Central, the NCUA is the boss. "Divergent views about how to move U.S. Central forward" is the politically-correct way of saying that Nance wasn't onboard with the NCUA's plans for U.S Central.

We'll have to wait till we see the NCUA's prescription for the corporates to have a clearer idea of the conflict, but (speculation alert!) it wouldn't surprise me if the NCUA is preparing for life without U.S. Central while Nance expected to manage a going concern.

CU Journal Strikes Again

Before I get too deep into the Credit Union Journal's latest example of playing fast and loose with the facts, let me disclose one change I made to my previous post. I have removed the screen shots that I took from cujournal.com of the stories I referenced in the last post. I have left the links in place so that those with subscriptions to the Journal can still directly access the stories. My removal of these images was done without any prodding from CU Journal. (Although Frank Diekmann and I have been in contact since my last post, this issue never came up.) Instead, one of my readers pointed out that I was potentially infringing on CU Journal's copyrights by posting these images. I agree, and since I like staying on the high road as much as possible, I made the unilateral decision to remove these images.

To review, between last Friday and this Wednesday, CU Journal had run a series of four articles on U.S. Central's 2008 audit results that significantly misrepresented the actual financial condition of the institution. Assertions were made that all membership capital would have to be impaired at U.S. Central because of the audited 2008 results, that the accumulated losses at U.S. Central exceeded the institution's capital, and that $6 billion in losses had been racked up by U.S. Central over the previous 18 months. The only way you can believe any of these contentions is if you either dismiss the guidance that the NCUA has provided on the subject of capital impairment or ignore the reversal of $3.7 billion of the losses due to the early adoption of FSP 115-2.

After seeing nothing on the Journal's site yesterday about U.S. Central, I was hopeful that they had seen the confusion their reporting was causing and decided to give it a rest. Turns out I was too optimistic.

CU Journal is back with another inaccurate story with today's piece, "NCUA Predicts More Corporate Losses to Trickle Down from U.S. Central FCU." (subscription required)

Since I'm not going to reprint their article, allow me to paraphrase and excerpt a couple passages. The back story is that the NCUA's Office of Corporate Credit Unions has issued a letter to retail corporates giving them some guidance on what to do in the aftermath of U.S. Central's annual report.

The letter's author, Mark Treichel, acting director of the OCCU, in fact says that all corporates will have to record "substantial impairment" due to U.S. Central's losses. However, one passage in the CU Journal article stands out:

But NCUA’s letter indicates that last week’s report showing $4.9 billion of losses for U.S. Central last year will mean additional write-downs on the remaining $453 million of U.S. Central MCS they hold on their books.

To be fair, "additional" is a very nebulous word and this story doesn't specifically say, like previous articles have, that all remaining MCS would be impaired. I believe the implication is made that big impairments are coming, but this merely my reading of the article. However, the bigger issue is that the NCUA's letter indicates nothing of the sort. This statement is completely without factual basis.

Compare the CU Journal story to the article posted on NAFCU's website. The NAFCU story is largely the same, but without the assertion that U.S. Central will have to impair "additional" MCS because of the annual report.

Here is the actual letter that Treichel sent to corporates:


And here is his August 18, 2009 letter referenced in this week's correspondence.


The key section of the NCUA letter is this:

OCCU understands the taking of PIC & MCA at the end of 2008 to cover losses that exceed retained earnings with a corresponding cumulative effect adjustment in 2009 "recapturing" some of those earnings results in an inequity in the depletion of 2008 regulatory capital, i.e., more capital is depleted than is appropriate. Accordingly, it is the opinion of OCCU that a corporate need not deplete PIC & MCA capital in 2008 to the extent that an OTTI cumulative effect adjustment in 2009, if known and considered at 2008 yearend, would not have resulted in the depletion of capital balances. Other than this one-time exception, regulatory capital should be depleted as described in Letter to Credit Unions 09-CU-10 dated May 2009.

There's a lot of jargon here, but this essentially says that you can ignore the really big 2008 year-end OTTI if you know that there is a reversal (or recapturing) of some of that OTTI in 2009. In other words, the U.S. Central loss that should be applied to capital in 2008 isn't $4.9 billion, it's $1.2 billion ($4.9 billion 2008 OTTI less the $3.7 billion reversal in 2009).

More importantly, there is nothing in either of these letters that "indicates that last week’s report showing $4.9 billion of losses for U.S. Central last year will mean additional write-downs on the remaining $453 million of U.S. Central MCS they hold on their books," as the CU Journal reported.

This isn't the first time we've heard about this accounting treatment for U.S. Central's large year-end OTTI. U.S. Central spells it out even more clearly in their annual report, on pages 14 and 15:

As of Dec. 31, 2008, U.S. Central had an accumulated deficit that was greater than the combined total of PIC I, PIC II and MCS. However, the NCUA did not require U.S. Central to fully deplete all PIC and MCS accounts as of Dec. 31, 2008, for the reason discussed below.

In April 2009, the Financial Accounting Standards Board (FASB) issued FASB Staff Position 115‐2 / 124‐2 (FSP 115‐2), which changed the requirements for OTTI recognition. The NCUA opined that the Dec. 31, 2008, depletion of PIC and MCS should be determined as if the Jan. 1, 2009, reversal of non‐credit losses ($3.7 billion) had occurred one day earlier. As a result, PIC II of $450.0 million was fully depleted, and PIC I was depleted by $104.4 million as of Dec. 31, 2008.

In the first two quarters of 2009, U.S. Central recorded additional OTTI charges of $1.1 billion, which resulted in significant net losses and additional depletion of member capital accounts. Through the first half of 2009, the remaining $195.6 million balance of PIC I was fully depleted, and MCS balances were depleted by $789.4 million.

Let me make two additional points:

1. There is one chance that CU Journal's reporting could be correct, and that would be if auditors refused to sign off on corporates' financials that incorporated the NCUA's opinion on when to recognize OTTI when calculating capital impairment. Indeed, if auditors forced the retail corporates to write down MCS based purely on U.S. Central's year-end results, then capital depletion would be larger.

But without providing evidence that auditors are balking at the NCUA's accounting guidance, then this series of articles rests on very shaky factual ground.

2. More OTTI is coming down the road. U.S. Central told its members as much on conference calls this week. This isn't surprising given continued economic weakness and the technical way that OTTI is calculated (it's the present value of future credit losses and even if loss estimates were unchanged in Q3 versus Q2, the OTTI would increase as the present value of future credit losses increased).

This forthcoming round of OTTI has nothing to do with what CU Journal is reporting. They have reported that all of U.S. Central's MCS is impaired because of 2008 audit results. Yes, all of the MCS could end up impaired some time in the future, but it would be because of additional credit losses, not because of the 2008 audited financials.

Why am I so irritated with you, CU Journal? Because I know that people read these stories with the expectation that they can believe what they read.

They can't. You have ceded all credibility on the subject of covering the corporates' losses. You are disgracing your profession. I can no longer chalk up your misinformation up to ignorance--it's crossed over into willful misrepresentation. And I would encourage NCUA and others to stop treating CU Journal as a news source until the magazine demonstrates it can publish factual stories.