The SIV Bailout Plan: Does the Math Work Even for Citi? (Revised)

A reader question got me to work through a back of the envelope calculation of what the SIV rescue plan, the so-called Master Enhanced Liquidity Conduit, would buy for its chief beneficiary, Citigroup. What I came up with gives cause for pause.

It’s one thing to know in a general way that a proposal is inadequate, another to work through the details to see how far it falls short (for background on this plan, see here, here, and here).

Note that I have revised this post since it first ran last night. There was an error in the analysis, which has been corrected, and a reader comment pointed to a second change in the analysis (it turns out that Citi’s funding need for November is maturing medium term notes, not commercial paper as I had assumed). That leads to a more elaborate set of assumptions per below. The conclusion is still not pretty, but now the results look better for Citi, but still woefully insufficient for the market as a whole.

Any comments, corrections, or improvements upon this quick-and-dirty estimate very much appreciated. This post is what they call in consulting a forcing device: you make your best stab at something and let the client take pot shots, with the goal of getting a much better approximation.

Based on what I had found on the Internet, it seemed that the typical capital structure for an SIV was 1% equity, 4-5% mezzanine debt, the rest commercial paper. I’ve seen numbers like that in other blogs, so they may have gone to similar sources.

However, that information appears to be out of date. One reader told me:

SIVs meanwhile are dynamically managed with strict but varied terms, levered and credit enhanced by subordinated and junior “tranches” financed by Medium Term Notes and CP.

Last summary I saw had financing 67% MTN and 25% CP remainder capital.

Another said:

3 Citi funds seen (~55% of total) have 63% MTN, 31% CP, 4.6% rest mezz and capital. Also have no market value capital tests.

Other SIVs may have a range from all CP to any mix I guess.

The number given above was a Feb 2007 S&P average for all SIVs.

These capital strucutres put the size of the MLEC in the same order of magnitude as Cit’s likely commercial paper funding needs. If you instead assumed their SIVs were funded roghly 95% with commercial paper, the MLEC would be a band-aid on a gunshot wound. So let’s play with the reader-submitted information.

The size of the market for SIVs has been pegged at $400 billion. The Citi funds in aggregate have (had) $100 billion in assets. They sold $20 billion, so now they have a $80 billion liability structure to worry about (note some sources have acknowledged the sales but still put the total size of the Citi program now at $100 billion, but we’ll stick with the more widely reported $80 billion.

The most recent reports say that Citi will take only half of the MLEC, so we’ll say $40 billion. Commercial paper goes out to 270 days. But asset-backed commercial paper, in particular SIVs, was frozen out as of the first week of August, so let’s say we have only 210 days of outstanding CP maturities (rounding up), and they are roughly evenly staggered.

So assume that a portion of Citi’s MTNs are part of the package. With Citi’s MTNs at 60-65%, or $48 to $52 billion, it’s not hard to imagine that $15 billion or more are coming due in November.

So the math says this program works nicely for Citi unless its MTNs maturing in November are much greater than $15 billion OR it has additional MTN maturities before the markets come to accept ABCP again.

Remember, per above, that by May of 2008 all the pre-August 2007 commercial paper will have matured.

But how does this program now help the rest of the market? Of the $75 billion for the MLEC, roughly $37.5 billion is going to other players who have a total of $300 billion of SIVs. If you again assume that 25-30% is funded by commercial paper, their needs are $75 to $90 billion. So for them, the program is an assist, not a rescue. So far that is consistent with the general reaction, that this program helps Citi more than the market.

But consider further: many other players, like Citi, must have some MTNs maturing in the next six months too. And if Citi needs turn to the MLEC for that, it means investors aren’t receptive to MTNs issued by SIVs either. So the funding needs of the other SIVs are surely greater than the CP requirement by some undetermined amount ($40 billion? $80 billion? Who knows?)

So the more you work through the numbers, the more that this program, as far as the market is concerned, looks like a band-aid over a gunshot wound.

Consider further: it appears that at least the Treasury believes that the MLEC will produce a return to normalcy. But in the LTCM crisis, it took a full a year after its was resolved for swap spreads to return to their former levels. And that was a vastly smaller rescue program.

So what happens if investors continue to be unwilling to buy commercial paper issued by SIVs? Citi is putting the industry through hoops to get the MLEC done. Yet this may or may not be a sufficient solution for Citi, depending on how large its MTN needs are, and it clearly isn’t much help to the market as a whole.

How will it play out when in a mere few months, when banks are continuing to tell investors that the SIV problem is not resolved, they have to dissolve the SIVs, and take substantial losses? This problem is not going away soon, and the prospect of a continued drumbeat of bad news on the SIV front will deter conservative investors from stepping forward.

In the spring of next year, even if the MLEC gets done, we will be in almost exactly the same spot we are today, unless we have son-of-MLEC. But even if the MLEC gets done, the assets remaining in the SIVs that sold a portion to the MLEC will be de facto nuclear waste. If you could find it at all, the cost of enough credit enhancement to make that stuff salable to conservative buyers would be too costly to make it a viable option. There may not be enough capacity among the natural buyers, the bottom fishers, for all the lower-quality SIV assets to find buyers.

What this says to me is that even if the MLC gets done, we won’t know if we have gotten through this credit crisis till mid-late spring, when the last of the pre-August 2007 ABCP will have either been rolled because the market is again operating reasonably well, or other funding has been arranged. And per the discussion above, the prognosis is not good. That also suggests we have many rough months ahead for equities as well.

Print Friendly, PDF & Email


  1. Anonymous

    The issue for SIVs is not so much CPs but the MTN maturities. CP holders will probably roll on but with shorter term maturities. A bunch of MTNs (for many SIVs) come due in November and December. If the MLEC plan takes 90 days to materialize, sponsors need to find someway to plug the funding gap.

  2. Anonymous

    I am not sure about the 400bn. Methinks that would describe the whole SIV market. Citi it seems to me had closer to 100bn in sponsored SIVs. (They may have more ABCP conduit exposure).

    If they have paid down 15-20bn CP which seems to gel with reasonable estimate of a 20% reduction in assets at their SIVs (and that the CP matures faster than supporting MTNs) and they have rolled the remainder (and “termed out” the liabilities as a consequence) they still have about 20-25bn CP and 50+bn MTNs outstanding.

    If we project this (all pure guesswork) onto a 375bn market total and adjust for higher CP in troubled SIVs and SIV lites we get about 50bn CP and 20bn MTNs paid down, on a 20% asset shrinkage.

    Looking at this it strikes one that in terms of CP shrinkage SIVs have made a significant but not the major contribution to 370bn? ABCP market declines.

  3. David Pearson

    The process of “recognition postponement” that you describe is a normal one in credit cycles. Hope of a rebound dominates this early stage, and aging inventory is not a big concern. The market welcomes opacity and delay as it greases the move towards higher prices.

    The transition to the next stage is a subtle psychological one. Little by little, opacity begins to feed anxiety rather than relieve it. Its okay to accumulate stale inventory when prices may rise, but not when you expect them to fall. The market begins to thirst for inventory clearance as a salve. The urge shows up as a suppressed desire, and then turns into a raging impulse.

    The response to Paulsen’s plan shows the transition is underway. His dogged advocacy indicates he still believes opacity is a salve, and its always dangerous when a policy maker misreads the situation.

    Once the transition begins, its impossible to predict how quickly we proceed to the “raging impulse” stage. The process accelerates with each policy-maker (or financial institution) attempt to prescribe opacity.

  4. Anonymous

    JP Morgan noted mid 2006 the following split in SIV investor base:

    Money Mkt Fund………70%
    Securities Lenders……20%
    Inv Mgr…………………..5%

    Lets assume this has stayed the same and turn holdings to $ using $375bn total.

    Money Mkt Fund…..262.50
    Securities Lenders….75.00
    Inv Mgr………………..18.75

    Lets do the same for geography and then for assets as given by an FT breakout from Bloomberg/Moodys.


    Asset Type
    Corp Financial……..161.25
    Mortgage Backed…..86.25
    Credit Cards…………18.75
    Student loans……….15.00
    Other ABS…………….7.50

    May be unconnected and maybe not over at CR these two new RegW and 23a exemptions were noted for Barclays and RBS est. amounts $20 and 10bn:

    … in addition to those granted for you guessed it MLEC troika Citi, JP Morgan, B of A, and mooted newbie…Deutschebank.


  5. Anonymous

    RegW 23A exemption amount limits

    Citibank ………25bn
    JP Morgan……25bn
    RBS………….. 10bn

  6. Anonymous

    Most expansive RegW and 23a exemptions stated purposes:

    … “to extend credit to market participants in need of short-term liquidity to finance their holdings of certain
    residential and commercial mortgage loans and mortgage-backed securities, commercial paper and other structured products”

  7. bob


    wherein Nouriel Rubini demolishes this shell game.

    “Can this financial engineering work? If the problem was only one of pure illiquidity and a generalized bank run then that solution may in principle work … But the reality is different in this case as the assets of the SIVs are not only sound (i.e. AAA or AA) but illiquid assets. Many of these assets are toxic MBS, CDO tranches and other ABS assets of dubious quality; thus, there is a problem of credit or solvency on top of illiquidity.”

    Some sort of solution that tries to to avoid devaluing the good with the bad, or devaluing the bad even worse than it should be may be called for – but it must be one that is transparent – and one that leaves none of the fools and/or criminals that committed this atrocity in place.

  8. Anonymous

    Citi has already said it has buyers secured for all its SIV CP for the next year. Problem doesn’t seem to be a funding issue, but I suspect more a problem with sliding NAVs.

Comments are closed.