Description

Long-Dated GMAC Unsecured Bonds – May 8, 2005

It seems not a day goes by that GM’s woes are not bandied about in the press. The secular decline of its North American auto business in recent decades had been truly staggering (25-26% share YTD vs low 40s% 20 years ago), blunted only by temporary cyclical peaks and temporary supply-demand imbalances of certain hit vehicles (like large SUVs in the late 90s). The UAW essentially runs the corporation, and not surprisingly, the company more closely resembles a socialist enterprise than a modern American corporation operated to maximize value to its shareholders.

Problems Galore at GM
The future at GM does not look particularly promising. New vehicles have thus far not been hits, and the GMT-900 light truck platform remains to be seen. The union seems unlikely to provide any major concessions before the current contract ends in Sept 2007. Healthcare costs continue to spiral out of control and contribute to a $1500 / vehicle structural cost disadvantage on an ongoing basis, not to mention the $60B+ unfunded OPEB liability. Transplants continue to expand capacity in North America (particularly in light trucks, the last remaining profit center for GM). The industry suffers from chronic excess capacity due to barriers to exit (union contracts, municipalities providing tax breaks to save jobs, etc). Market share losses only exacerbate the lopsided ratio of retirees:actives, a vestige of the old days when GM produced the majority of cars for the North American market. S&P cut both GM and GMAC by 2 notches, firmly into junk territory, a quicker fall to BB than many were anticipating. etc etc. The press has done a good job of articulating all the headwinds being faced by GM, so I will go into all of them again here.

The Opportunity
Nevertheless, I believe that long-dated GMAC unsecured bonds have become attractive, yielding approximately 10% to maturity and nearly 10% on a current yield basis, and priced in the mid-high 70s cents on the dollar. Specifically, there are 4 retail debt issues (dubbed PINES) aggregating $3B in face value that trade like preferred stock on the NYSE with $25 face value, but which are bona fide unsecured bonds (tickers GKM, GMA, GJM, GOM). There are also large “conventional” long-dated bond issues such as the 8% of 2031s (a $4B issue).

Given the overwhelming sheer dollar amount of GMAC, GM, Ford, and Ford Credit debt that has been cut to junk, there has been significant forced selling. Spreads on GMAC debt now resemble CCC levels. This is a classic example of technical factors carrying the day over the short term and creating opportunity for other investors to exploit. But in contrast to many similar situations, this opportunity is highly liquid.

While short-dated GMAC unsecureds are also attractive, I prefer the long-dated bonds because their lower dollar price and longer duration provide both greater protection in a bankruptcy and more upside if Ch 11 is avoided (albeit greater mark-to-market risk in the short term).


Parent Bonds Speculative
While parent GM automotive bonds yield even more and have a slightly lower dollar price (8.375% of ’33 at a bit over 73; ~11.6% YTM), I consider them to be quite risky and believe it makes a world of sense to pay up a bit for the safety of GMAC. If GM eventually files Ch 11, I think eventual GM unsecured recoveries would be severely diluted by the sheer magnitude of retiree and trade claims that would arise. With just $32B of auto debt compared to $61B PV of unfunded OPEB (which should continue to rise as VEBA cash is drawn down, healthcare costs continue to skyrocket, and “interest cost” far outpaces VEBA asset returns) and $24B of A/P at yearend, GM unsecureds would be a minority of the reorg equity. And this is before any potential pension claims.

High Leverage Helps
GMAC bonds, on the other hand, should provide far higher recoveries in a bankruptcy.
First and foremost, by virtue of GMAC being a finance company with highly liquid assets and a balance sheet with tremendous financial leverage, the value of the assets must sustain a truly enormous hit before unsecureds start to recover less than their cost basis in the mid-high 70s cents on the dollar. Not only would the book value have to be written down from $22.4B at yearend to $0, it would have to be written down to $0 many times over.

By way of (overly) simple illustration, GMAC had $324B of assets on balance sheet at yearend. It had $92B of secured debt (mostly on balance sheet securitizations), $7B of insurance loss reserves and unearned premiums, and $177B of unsecured debt. At 76 cents on the dollar, the unsecured debt is created at roughly $135B. Then assume secured debt and insurance loss reserves, etc are paid in full, $99B. $99B + $135B = $234B of value needed for unsecureds to recover at 76 cents. $324B - $234B = $90B. This implies the $324B of assets must sustain a $90B hit before unsecureds lose money.

And of course, your cost basis falls over time as you collect coupons prior to a bankruptcy. Plus, in a bankruptcy, post-petition interest does not accrue on unsecured debt – building up cash and further buttressing recovery values.

OPEB
Importantly, the OPEB liabilities do not carry over to GMAC. They are solely the liability of the parent (and perhaps the government to some extent if the UAW ever gets its way with Washington, though I’m not holding my breath).

Pension
Pension is a different story. Pension liabilities extend on a joint and several basis to all members of the parent’s “control group”, which includes any subsidiary that is at least 80% owned by the parent. So GMAC could be on the hook. And while the US pensions are currently fully funded, with $90B of assets, a bear market that causes an extended period of poor returns could create a sizeable funding gap.

Projected pension liabilities could also increase if GM is successful in getting workers to take early retirement, as GM would now be paying a pension for more years and starting those payments sooner. That said, its hourly workforce is already much older than its competitors (average age now apparently in the low 50s), so much of the workforce is already approaching the 30 years of service at which time most retire anyway. And GM’s ongoing operating cash flow would increase (retiree pension payments are less than active wages) at the same time.

Moreover, it seems exceedingly unlikely that the PBGC would be in any hurry to terminate GM’s pensions given its own present funding gap, and the tremendous administrative burden of running the largest plan in the country. (The PBGC does not have the infrastructure in place to do so). In the unlikely event the plan were terminated, I believe the potential claim on GMAC would be limited to 30% of GMAC’s net worth (as defined by the court). That number should pale in comparison to GMAC’s unsecured debt given the balance sheet leverage at GMAC, and it would be a pari unsecured claim. Finally, 20.1% or more of GMAC could be sold or spun off to remove GMAC from the control group and thus eliminate any potential pension liability altogether.

UAW Concessions?
Much has been made of the UAW and the difficulty in getting concessions. I acknowledge that GM is run for the benefit of labor, dealers, etc, with shareholders dead last. The UAW should therefore capture 100% of all economic rents and leave shareholders none. (That is why I could not envision investing in GM common over the long term). But, the converse is also true. To the extent the UAW finds itself in a position of capturing greater than 100% of economic rents, it would be rational for it to give just enough concessions to bring things back “in balance”.

I can’t say I expect the UAW to be rational, as it is a political organization run by leaders that are not eager to be voted out of office. Nevertheless, I do think the stage is being set for the possibility of material concessions in 2007. Legacy costs and the need to downsize are all over the news. Everyone can see what happened to the legacy steel and airline companies and their workers.

Recall that Ford got cuts from the UAW and plant closures in order to stabilize when it last hit the skids around 2002.

Industry Cyclicality
The auto industry as a whole is notoriously cyclical. But the cyclicality of individual players within it is even more pronounced. Just in the past 5 years or so, conventional wisdom/sentiment on Ford has gone from top of the heap, returning billions in cash to shareholders with Nasser at the helm; to hitting the skids several years later; to being considered “clearly” ahead of GM today on the strength of its renewed F-150. GM was in terrible shape in the early 1990s. But it then recovered on the strength of demand for its light trucks. Chrysler narrowly avoided bankruptcy at least twice, rescued the first time largely on the strength of its minivan.

I realize that it may be different this time, especially as GM continues to lose market share and the retiree:active ratio gets further out of balance. Not to mention that the last bastion of Big 3 profitability (light trucks) is firmly under attack from the transplants. And that GM is no leader at present in what appear to be the new hot segments – CUVs and hybrids. Nonetheless, reversion to the mean can be a powerful force.

In any event, I think Kirk Kerkorian and other buyers of GM common require far more conviction in a turnaround than do buyers of GMAC bonds at a large discount to par.

Liquidity at Parent
Excluding GMAC, GM auto has $16B of unrestricted cash on the balance sheet. It also has $20B of assets in the VEBA trust. Management explained on the call that up to 18 months worth of retiree cash healthcare cost can be removed from the VEBA at any given time (and used for anything at that time). So although all $20B cannot be removed at once, under any reasonable scenario, all $20B could be removed over the course of several years, thus providing another $20B of liquidity.

GM has stated that barring major unforeseen changes, no cash pension contributions to the US pensions should be required for the rest of the decade. VEBA cash balances should cover cash retiree healthcare for the rest of the decade. So at least in the intermediate term, cash calls on operating cash flow due to legacy costs are minimal. Furthermore, GAAP OPEB expense is significantly in excess of cash OPEB benefits that are being paid (even ignoring the fact that VEBA could pay cash OPEB instead of paying it out of operating cash flow). Likewise, a large GAAP pension expense is still being recognized even though no cash contributions are foreseen for the rest of the decade.

Plus, the $32B of automotive debt has an average maturity of 19 years, with only $3B maturing in the next 5 years. In short, while auto companies can be notoriously cyclical, burn tremendous amounts of cash at times, and destroy shareholder value over long periods of time, it is hard to see why GM auto should face any liquidity issues over the next several years (barring the “mother of all strikes”).

If even GM auto (the “weakest link”) can lumber along for at least the next several years, there is no reason why GMAC should not continue to pay its bond coupons as they come due.

Negative Pledge
While GMAC is able to securitize its receivables (they are considered to be “sales” of assets), the GMAC unsecureds have a negative pledge. Some have suggested that GMAC’s banks may require collateral in the future (they are presently unsecured). Given the negative pledge, I believe this would require them to share that security pro rata with the unsecureds, which should prevent this from happening given the size of the unsecureds relative to bank debt. More likely in my opinion, the banks may implement more stringent financial covenants, which would also benefit the unsecureds. (Currently, the only financial covenant is a limitation on leverage, which thus limits the amount of dividends that can be upstreamed to the parent).

Operating Agreement
Fortunately, GMAC does not guarantee GM debt. In addition, an “operating agreement” was put in place in October 2001 (filed as an 8-K, also attached to the end of GM’s 2004 10-K) to protect GMAC from GM. All transactions must be done at arms length. GMAC will not guarantee GM debt. GMAC’s capital must be maintained at commercially reasonable and adequate levels given the size, mix, and quality of its assets. Separate books, records, financial statements, bank accounts must continue to be maintained. Cash cannot be commingled. Etc etc. While all this is great on paper, and I certainly prefer to have it than not, and GMAC insists that it is enforceable by all impacted stakeholders, lawyers I’ve spoken to say its legal enforceability is questionable.

Subvention
When GMAC takes on 0% incentivized loans, GMAC does not bear the credit risk from GM. For instance, for a $20,000 loan at 0%, GMAC buys it from the dealer at a discount (say for $18,000) and GM auto pays the dealer the $2,000 incentive on day 1. GMAC is not awaiting “make-whole” payments from GM over the life of the loan. In contrast, the way it works at Ford is that Ford Credit buys the loan at par for $20,000 and then receives periodic payments from Ford Motor for the difference between market interest rates and 0%.

Could GM Pillage GMAC?
Notwithstanding the operating agreement, there is the risk that the parent pillages GMAC. While dividends are limited by the leverage covenant in the GMAC bank debt, GM might be able to take cash out of GMAC in exchange for unsecured intercompany notes. While this is one of the risks that worries me most, I again take comfort from the sheer size of the GMAC balance sheet. GM auto may hemorrhage cash, but even if it were to pull out $5B per year from GMAC via intercompany notes (above and beyond the dividends of $2B+ per year and tax sharing payments of ~$1.5B per year it already receives), it would take many years before it would put a meaningful dent in the recovery of GMAC unsecureds. Moreover, I suspect that once that began happening, bondholders, rating agencies, etc would have plenty of time to resort to a public outcry before too much damage had been done.

Comparison to Xerox
Nevertheless, I took a look back at the Xerox / Xerox Credit situation to see if I could draw any comparisons or learn any lessons, as Xerox essentially took cash out of Xerox Credit in exchange for interco notes. True, Xerox Credit bondholders have thus far not been impaired. Still, I take even greater comfort from the facts surrounding GMAC for several reasons.

First, the turnaround plan at Xerox included major asset sales, outsourcing of future equipment financing to third parties such as GE, and thus the orderly liquidation of Xerox Credit. Perhaps the reason Xerox was able to take out cash and stuff Credit with unsecured interco notes is that they did not need to return to the debt market to raise new debt at Xerox Credit. Applying this to GMAC, the question becomes whether a 3rd party could easily take over future vehicle financing; or if the 80%+ share of dealer floorplan financing and ~50% share of retail financing is of a sheer size as to make that difficult to accomplish. Xerox Credit's porfolio was ~$5 billion, obviously a drop in the bucket relative to GMAC's portfolio (closer to $120 billion on balance sheet excluding the mortgage business). I suspect that would make it a lot harder for GM to find a 3rd party to take over financing. If a 3rd party cannot easily step in and take over future financing, then this could be a key difference and source of protection for GMAC unsecureds.

Next, Xerox Credit was essentially a shell. It had no real infrastructure or employees (only a handful of execs who were also execs of the parent, and those execs received no compensation from Credit for their work). Credit could not even service its portfolio -- it reimbursed Xerox parent for doing that. In contrast, GMAC is clearly not a shell. It services its own portfolio, has its own infrastructure, has some execs/board members that are execs of GM, but also some that are not, etc. Xerox Credit’s financial statements were also pretty bare bones and indicative of a shell -- it might be a stretch to say that Xerox Credit unsecureds were relying on Credit's standalone creditworthiness when it could not even service its own portfolio. Furthermore, Credit could not sell, transfer, or assign its finance receivables without prior approval of the parent. So not only couldn’t Credit service its portfolio, Credit could not realize the value of its assets by selling them either. Therefore, Xerox Credit unsecureds were implicitly relying not just on the creditworthiness of Credit, but also on the creditworthiness of Xerox parent. Again, this is not the case with GMAC where GMAC can service its own portfolio and is free to sell its receivables.

Taking a look at the Xerox / Xerox Credit Support Agreement and Operating Agreement, I saw no mention of transactions needing to be done at arms-length. (All that seemed to be mentioned was that Xerox parent would make payments to Credit to ensure Credit had fixed charge coverage of 1.25x and that parent would be required to own 100% of Xerox Credit at all times). While the GMAC operating agreement may not necessarily be enforceable by third parties such as GMAC unsecureds, at least it publicly states that intercompany transactions must be done on arms-length terms. And at least it doesn’t restrict the complete or partial sale of GMAC to outside investors.

Also, all the original underlying finance receivables were originated by Xerox parent and thus were contracts between customers and Xerox parent, not Xerox Credit. Xerox Credit purchased its entire portfolio from Xerox parent after the parent had originated the contracts. In contrast, I believe that GMAC receivables are contracts between customers and GMAC or customers and third party dealers -- I do not think GM is a party to those receivables. And again, when Xerox Credit routinely purchased finance receivables for its portfolio from Xerox parent, the purchase price of receivables "shall be such as the parties shall agree" – there is no mention of arms-length terms. Though technically this applies to Credit's purchase of customer finance receivables from Xerox parent for its portfolio -- this does not explicitly refer to interco notes -- the absence of any mention of arms-length terms nonetheless seems telling.

Funding Cost Increases
It goes without saying that a finance company cannot sustain itself profitably over time if its borrowing costs are 10%. The current situation, therefore, is not tenable for the long term. I think there are several ways the situation may resolve itself.

As many have suggested, future funding may largely shift towards ABS and away from unsecured borrowing. Next, parts of GMAC could be monetized. Private equity firms are studying the Commercial Mortgage business. The Residential Mortgage business is in the process of being ring-fenced and potentially spun off. Much of the balance sheet growth, in fact, derives from the ever-growing mortgage business and not automotive financing.

Moreover, the balance sheet could potentially be shrunk to generate cash. Some would dismiss this idea out of hand as the auto business relies on GMAC to provide financing to new vehicle buyers. However, as GM continues to lose share, there are less GM vehicles that need to be financed. Perhaps more importantly, GMAC’s share of both wholesale floorplan financing (>80%) and retail financing (~50%) are both around all-time highs. Decreasing share of such financing would allow for the balance sheet to contract. The last time GM ran into serious trouble in the early 90s, GMAC shrank its balance sheet for several years. Ford Credit has just come off several years of shrinking its balance sheet, and it has paid $8B of dividends to Ford in just the past 2 years alone!

Structural Subordination
Considerable attention of late has been focused on the effects of a funding shift to ABS, whereby unsecureds would effectively become structurally subordinated over time. This is a definite risk, but I think it has become overblown. When you actually run the numbers in a recovery analysis, recoveries for unsecureds are surprisingly robust. Moreover, the bulk of ABS relate to the mortgage business, which would of course vanish if that business gets sold or spun off.

Furthermore, such a shift to ABS will take time. It will not happen overnight. And as time passes, bond coupon payments are received. So your cost basis is falling at the same time, protecting you against the effects of structural subordination. After 5 years, for instance, your cost basis is below 40 cents on the dollar.

GMAC is well run, has tremendous liquidity of its own, and should have time to weather the current storm. Moreover, with high quality and liquid assets that run off more quickly than its debt over the next 3 years, it should not have a problem meeting its debt maturities for the foreseeable future.

Upside Potential
While I’m comfortable with the downside and see minimal risk of permanent loss of capital from present price levels, the upside potential is decent and you get approximately 10% in current yield to wait. For instance, the GMAC 8% of 2031 are quoted at 78.75 last. If they were to trade up to par over 2 years (at that point they would yield 8%, still a spread of 337 bps), that would provide 21.25 points of capital appreciation in addition to 16 points of interest payments, north of 21% compounded over 2 years.

As for a potential home-run scenario, the GMAC 8% of 2031s are not callable. If GE ever were to buy GMAC, these bonds could easily trade to 130 (they’d still be yielding 5.77%, 115 bps spread, at 130). (Note that the retail PINES bonds, on the other hand, are callable, so their potential upside above par is capped.)

Risks
Parent pillages GMAC by taking out cash and stuffing it with intercompany notes.

Structural subordination as funding mix shifts to ABS.

Substantive consolidation – exceedingly unlikely according to lawyers I’ve spoken to.

Consumer meltdown.

Mark-to-market risk.

Could trade perennially cheap -- lack of HY index buyers and closet-indexers relative to the amount of GMAC debt outstanding.

Catalyst

-UAW concessions with the new contract in 2007.

-Sale / spin-off of GMAC mortgage and/or insurance business.

-Third party investor buys a stake in GMAC.

-Bonds complete their migration out of investment grade hands.

-Contractual interest payments at ~10% current yield and no near term liquidity issues to interrupt that flow of payments provide a continuous catalyst to realizing value, making additional catalysts less critical than in other situations.