Susanne Craig had an article in DealBook today about regulatory capital relief trades and a basic thing about regulatory capital relief trades is that they always allow you to say this:
One Citigroup executive with knowledge of this trade but not authorized to speak on the record said it was structured to reduce Citigroup’s exposure to shipping loans. The fact that it reduced the amount of capital the bank had to hold was “an added bonus.”
"Bonus" is, as everyone knows, a word with different meanings in and out of the financial industry. Most people say "bonus" to mean a relatively small and unexpected gift that you're pleased to receive but weren't, like, counting on or organizing your life around. For bankers, the "bonus" is the whole point of what you're doing.
The Citigroup trade that Craig refers to is a bespoke tranche trade with Blackstone on a pool of shipping loans; we discussed it a little in February. Per her numbers Blackstone has exposure to the 0-12% tranche and has an expected return of 15%, while Citi "was able to reduce by roughly 90 percent the amount of capital it needed to set aside to cover losses on the shipping portfolio." So Citi has reduced its notional exposure by 12%, and its capital exposure by 90%. A 7.5:1 ratio of bonus to base ... seems fair, I guess?1 A little 2007, maybe, but then these are bespoke synthetic CDO trades, so much about them is a little 2007.
Anyway Craig provides various other examples and you can think your thoughts about whether this should bother you. I like shenanigans so I am blithely unperturbed, but the "AIG" word is thrown around pretty freely so you may disagree. If you are perturbed, you might be enamored of the Brown-Vitter proposal to get rid of risk-based capital measurements, which would cut off a lot of banks' abilities to get capital relief by selling risky slices like this. That said! It would also create new and exciting ways for banks to get capital relief by just taking on riskier exposures.2Relevant:
[T]he capital measure used by regulators will, over time, come to be outperformed by a measure that the regulators are not using. So, if you are using standard capital, risk-based capital measures will better predict bank risk, and conversely.
Everybody knows now that risk-based capital does a worse job of predicting bank risk than does brute leverage, but they forget the conversely. Give Brown-Vitter some time and you'll be right back in the soup.
Anyway, that's not the point, this is the point:
A number of American investment firms like Spring Hill Capital Partners have been trying to find investors for these deals. Glenn Blasius, another Lehman alumni, said he was raising money for the Ovid Regulatory Capital Relief Fund, which will invest in these trades.
NO JUST NO.
Look here is the basic secret of selling financial products. It's all simple boring stuff. If you want to reduce your exposure to shipping loans, make fewer shipping loans. If you want to hedge your equity risk, sell some shares. You can reduce your delta to anything with complex derivatives, or you can just buy fewer deltas. Buying fewer deltas is always more efficient.3
Pretty much any more complicated structure that you glop on top of your trade is designed to obfuscate something, and to make up for the economic inefficiency of the structure with some sort of tax, accounting, regulatory, public relations, something benefit. Here you can read a brief and genuinely delightful history of the development of credit trading, and some of the things that various credit products were built to obfuscate seem perfectly harmless - genuine risk shifting without offending clients, for instance.4 Others seem perhaps more harmful - ratings agency arbitrages, for instance.5 You can take your own counsel on regulatory capital arbitrage trades.
But you gotta obfuscate it. The whole game is to point to an economic trade, and show that there's a risk/reward tradeoff that makes it seem economically rational for an economically motivated actor to pursue. You don't do accounting trades with no non-accounting purpose, or do tax trades with no non-tax purpose, or explicitly build a business around tax avoidance for that matter. You don't put "Optimizing regulatory capital" in the subject line of an email. You put "reducing shipping loan risk" in the subject. "The regulatory capital benefit is just an added bonus," you say, with a knowing wink if you're feeling frisky.
You don't, for the love of God, do your regulatory capital relief trades with a fund called "Ovid Regulatory Capital Relief Fund." Honestly, it might as well be called "Ovid Trade With Us And Go To Prison Fund." You do your regulatory capital relief trades with nice responsible pension funds. Maybe Blackstone. Something vague like "Spring Hill" is fine. In a pinch, AIG.
1.Okay: you know, and I know, that I'm being unfair here for the joke. You can fake out some math here; let's say that these loans have an expected one-year default rate of 1% with a standard deviation of 1% (sort of BB-ish, historically). Just super lazily say that this is a five-year trade and pretend that the expected 5-year default rate is 5% (??) with a standard deviation of 2% (goes with sqrt(T) don'tcha know). That means that 12% defaults is like 3+ standard deviations away, or <1% likely, so in some sense Citi has gotten rid of more than 90% of the risk, not less. Real math is left as an exercise for the reader.
2.E.g.: just buy that risky 12% slice, instead of selling it. Does it have 90% of the risk of the whole portfolio? Umm no we calculated above that it has 99% of the risk. Whatever, it certainly doesn't have just 12% of the risk. But it does to Brown-Vitter! Etc. Informally, you can turn $100 of any asset into $10 of that asset, 10x levered, off balance sheet, and if you have no risk weighting you will.
3.I realize that this is wrong, or wrong-ish, and the bias of a former corporate derivative structurer, but: it's right in most contexts relevant here.
4.CDS was basically designed to obfuscate client relationships:
JP Morgan wanted to be able to hedge their loan books. Ultimately that was their primary motivation initially. They wanted to be able to lend money and demand the borrowers use them for lots of cross-sold services. The problem was that borrowers wanted “unfunded revolvers”. These were a pain to manage as they didn’t pay much and required banks to be able to fund on demand. The loans couldn’t be transferred outside the banking system, and any bank that owned a chunk of the loan, also demanded ancillary business from the borrower. The solution, was to lend more, crowd other banks out, become the dominant lender, able to extract maximum extra value, and to do that, they needed to hedge. That was their interest in credit derivatives.
Tchir goes on though to discuss genuine efficiency reasons for CDS (basically fungibility), so it doesn't fully support my thesis above. Again, recommended.
5.Though YMMV on "exploiting" vs. "correcting" rating agency errors:
[Leveraged loans] “are the most misrated mispriced investment in fixed income”. Because they were high yield, many investors couldn’t own them. Because rating agencies were restricted to giving 1 or 2 notches of rating values (unless fully defeased with treasuries) the value of the collateral wasn’t priced in. This was a huge divergence between the “structured product” groups that relied on “expected loss” and the corporate rating groups which had their own methodology.
Et voilà, CLOs.